Party-Voter Linkage in Africa pptx

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 subdiscipl

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 http://dnb.d-nb.de.

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 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.

pub-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

Th is work was submitted in fulfi 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

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

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

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

BGC Bloc des Centristes Gạnde

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

FSD/BJ

GA Front pour le Socialisme et la Démocratie/Benno Jubël Great Alliance

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

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

NIMD Netherland Institute for Multiparty Democracy

NIP National Independence Party

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

PIT Parti de l’Indépendence et du Travail

PLS Parti Libéral Sénégalais

PNDC Provisional National Defense Council

Abbreviations and Acronyms 11

PPC Parti pour le Progrès et la Citoyenneté

PRA Parti du Regroupement Africain

PSA Parti Socialiste Authentique

PSAS Parti Sénégalais d'Action Socialiste

PSD/JB Parti Social-Démocrate/Jant-Bi

PSR Parti Socialiste Républicain

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

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 230

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 20th 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)

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 survey4 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

Oceania

Western Europe

Asia Central/Eastern Europe

Central/South America

North America/Carrib.

Middle East

Africa

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

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

http://ec.europa.eu/public_opinion/standard_en.htm

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:

ƒ 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;

ƒ 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

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:

ƒ hardly distinguishable programs which in most cases are meaningless to policies

ƒ hardly countable members

ƒ weak bureaucratic organisation

ƒ informal relations, partly based on clientelistic relations and patronage, predominate the party structures

ƒ strong personalism

ƒ high degree of factionalism

ƒ weak funding

ƒ lack of internal democracy

ƒ weak formal linkages to society

Party Research in Africa: Findings and Problems 25

ƒ 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 literature7 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:

ƒ 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

ƒ African parties do not seem to fulfil the functions that are ascribed to them (Emminghaus 2002; Randall/Svåsand 2001)

ƒ 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)

ƒ Political parties have only weak organisational linkages to civil society groups (Widner 1997; Randall 2007)

ƒ 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 law9, 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)

Party Research in Africa: Findings and Problems 27

Another definition is offered by Frank J Sorauf for whom a political party

is

an agency for the organization of political power characterized by exclusively political functions, by a stable structure and inclusive membership, and by the ability to dominate the contesting of elections (Sorauf 1964: 13)

It is obvious that a “stable structure and inclusive membership” is seldom found

in Africa and can therefore not be a part of the definition

This short review highlights some of the difficulties in dealing with political parties in Africa These concerns are not entirely new Sartori, who

offered a comprehensive framework for analysis in his book Parties and Party Systems (2005 [1976]), was faced with the same problem: parties in Africa did

not fit neatly into his models His metaphor of the ‘African Labyrinth’ (ibid.: 248) marks his capitulation before what he called the fluidity of African polities

He concluded that

the polities lacking structural differentiation and consolidation cannot be incorporated under the Western categories, nor can they provide categories for the West (ibid.: 266)

The very same problem continues to be highly relevant today, and questions such as “Are political parties in Africa different?” reappear in the discussion

We are not short of examples:

ƒ In 2004, Gero Erdmann bemoaned a ‘Western bias’ in party research

ƒ In a similar vein, Matthias Basedau (2002: 312) noted that possibly all of the descriptive and analytic concepts of party research are only partly applicable to Africa

ƒ In the introduction to African Political Parties, Salih (2003: 3) asks the

question whether African parties differ from other parties

ƒ Peter Burnell (2007) named one of his articles “Political Parties in Africa: Different, Functional and Dynamic?”10

ƒ A session of a conference held by the German Friedrich-Ebert-Foundation

2005 in Accra/Ghana was entitled “Do political parties in Africa differ from parties in other regions?” (see Basedau 2005)

10 Burnells article was a reply to Erdmanns above mentioned article “Party Research: The Western European Bias and the ‘African Labyrinth’”

The fact that African parties deviate empirically from the classical theoretical assumptions concerns not only area specialists but also poses a challenge to the whole field of party research In the next chapter we will see if and how this challenge has been taken up

2.3 New Challenges: Neo-democracies and Party Change

For some time it seemed that political parties were becoming redundant and that other actors, for example new social movements, could perform their functions equally as well, if not better (see, among others, Merkl/Lawson 1988) Despite the gloomy scenarios of party failure or decline, political parties have proven to

be indispensable institutions and in all democracies, and the field of party research has seen a substantial reawakening in the recent years Two factors have contributed to this:

ƒ the changing role of parties in established democracies (‘party change’), and

ƒ the proliferation of parties in formerly authoritarian regimes in the course

of the ‘Third Wave of Democratization‘

These developments confront theory building with new problems and call for a reassessment of old concepts Interestingly, political parties in the new democracies in Asia, Eastern Europe, Africa, Latin America, and the Middle East share some striking similarities They all appear to have weak formal structures and ideologies (see for example Carothers 2006: 6-7) Furthermore, they have fewer institutionalised contacts with civil society groups than their counterparts in Western Europe (Betz/Erdmann/Köllner 2002: 2) In short, they all lack the element of ‘mass politics’ that has long been believed constitutive to party development Thus, empirical findings from new democracies challenge the conventional wisdom about political parties

Simultaneously, political parties in established democracies are undergoing

a process of transformation They are losing members, becoming less strict in their ideological orientations, becoming gradually more disentangled from civil society groups, and as a consequence, becoming more volatile (see, for example, von Beyme 2000) Party development in Western Europe is often believed to have taken place in the following stages (see von Beyme 2000: 27; Katz/Mair 1995: 8-12):

ƒ elite parties in the late 19th and early 20th century;

New Challenges: Neo-democracies and Party Change 29

ƒ mass parties from 1880 to 1960;

ƒ catch-all or people’s parties from 1945 onwards,

ƒ and cartel parties from 1970 onwards

The elite parties in the 19th century did not have an extensive organisational apparatus because the voters and the people in positions of power belonged to the same social strata and were often personally interconnected (Katz/Mair: 1995: 9; Duverger 1959: 5) The formation of mass parties was therefore inextricably linked to the extension of suffrage and the need to attract large numbers of new voters (Duverger 1959: 2) Mass parties with their strong emphasis on the role of party activists were initially the organisational form of the working class and disenfranchised segments of the society They usually had

an extensive organisational network, a strong ideological orientation and relied

on their members for funding and campaign activities In the struggle for votes, parties on the right followed their example Duverger called this process a

“contagion from the left” (ibid.: 6) In the first half of the 20th century, parties represented well-defined social groups, and politics was about “the competition, conflict and cooperation of these groups” (Katz/Mair 1995: 6) Kirchheimer (1966: 184-186) argued that after the Second World War the ideological differences between the major political parties in Western Europe were decreasing and parties were moving away from their position as representatives

of traditional social milieus They became more professional, more elitist and primarily oriented toward vote-maximisation Kirchheimer called this type of party the ‘catch-all’ or ‘people’s party’ He was avowedly critical of this development because it adversely affected the integration and representation function of political parties (ibid.: 199-200) In many aspects, the type of the cartel party, which was proposed by Richard Katz and Peter Mair in 1995, is a continuation of the catch-all type The authors hold that parties are moving further away from civil society and subsequently “become a part of the state apparatus itself” (1995: 14) The decline in participation makes parties more dependent on state funding, and their role as governors and law makers allows them to determine the rules for these subventions themselves (ibid.: 15) This leads to the formation of a cartel in which the spoils are distributed among the existing parties while newcomers are kept outside

The fourth type is still controversial,11 and there is little agreement on terminology or the extent of change, but it is generally assumed that party change is an adaptation of political parties to a changing social environment One set of explanatory factors centres on the weakening of social cleavages, which leads to a dealignment of parties and voters (Betz/Erdmann/Köllner 2004: 12) The reasons for this are manifold Gunther and Montero (2002: 8) name secularisation, the expansion of the middle classes, greater participation of women, migration, higher degrees of personal resources among citizens, postmaterial values that conflict with traditional party ideologies, and better informed citizens independent from the guidance of secondary associations and opinion leaders as various motivating factors Alternative or complementary explanations stress the changing role of the media: mass-communication opens new channels for direct access between citizens and leaders that need not pass through traditional partisan channels (Gunther/Montero 2003: 8; see also Poguntke 2003: 11; Alemann 2002)

Although these tendencies are observable in most Western countries, party change is a highly uneven process and there are still many unresolved problems Because of different understandings of party membership, the extent of partisan erosion in the various countries is difficult to compare (Norris 2002: 116) Another question is whether party membership has become meaningless and whether the decline in membership is an irreversible phenomenon (Wiesendahl 2006) Norris assumes that the main decline could have been in the more peripheral supporters, while the core workers may persist (2002: 111) Furthermore, the decrease of party membership is unevenly spread across parties in the same country Social-democratic parties, which are the historical model for the mass party paradigm, have been hit much harder than conservative parties Conservative or Christian parties initially resembled the catch-all type, while parties on the left had to adapt to this situation (von Beyme 2000: 34) That is why the story of party change is very much the story of the change of social-democratic parties in Europe However, the crisis of the German Social Democratic Party (SPD), the series of electoral defeats in federal elections and the ongoing erosion of membership is not only a result of the erosion of the traditional working class milieu, but is also very much related to the performance of the party Under Chancellor Gerhard Schröder, the party moved away from its traditional labour policies and reoriented itself towards a

11 For a critique of the cartel party model, see Koole (1996) Alternatively, contemporary parties could also be described as “professionalised voters’ parties”, as it was proposed by Panebianco (1988)

New Challenges: Neo-democracies and Party Change 31

‘new middle’ that embraced more market-friendly and neo-liberal positions.12 In the following years, the package of social cutbacks and labour market reforms that was introduced alienated substantial proportions of social-democratic core voters while the party was unable to make inroads into new voter segments This gave rise to Die Linke, a party on the left of the SPD, which actually promotes the ‘old’ social-democratic values and which is especially successful among the unemployed and people with a low income.13 These evolutions pose

a challenge to the notion of ‘catch-allism’ First, it seems that the left-right divide is still structuring according to the preferences of a considerable proportion of voters in Western Europe Secondly, the process of ameliorating living conditions and voters’ preference for post-material values seems to have been exaggerated In Germany, social inequalities are increasing again, accelerated by the neo-liberal policies under the Schröder-government14, and materialist values are reappearing This fits into the findings of recent studies, which suggest that ideological change in political parties is overestimated (Poguntke 2003: 14; Gunther/Diamond 2001: 191)

While there are many controversies and open questions, it is interesting to note that party change is often described using negative terminology, almost as a

deviation from what parties ought to be This line of argument began with

Kirchheimer and cumulated in the various scenarios of parties in crisis and decline Although party research has not produced a general theory (von Beyme 1988: 211 and 2000: 15), the debate on party change cannot to be disassociated from the notion of the mass party as an ideal type of party One of the reasons might be that some of the most influential scholars in the field originated from Western European countries Even parties in the USA do not easily fit into the dominating categories and are often viewed as “puzzling phenomena by scholars from abroad” (Janda 1993: 164) Richard Katz gives us a summary:

12 This view was expressed 1999 in the so-called ‘Schröder/Blair-paper’ by the then-German Chancellor and the then-British Prime Minister Anthony Blair Schröder and Blair were in the forefront of a reformulation of social-democratic policies

13 The SPD is primarily losing the trust of voters with low income and unemployed people According to the analysis by the opinion research institute Forschungsgruppe Wahlen only 19% of unemployed voted for the SPD, but 31% for Die Linke in the German general elections in 2009 See: http://www.forschungsgruppe.de/Aktuelles/Bundestagswahl_2009/Newsl_BTW09.pdf, (accessed: 12 October 2009) For an interactive tool on German party affiliation according to various social indicators see http://www.bpb.de/fsd/werwaehltwas/bpp6s0 (accessed: 12 October 2009)

14 Ironically, only a few years after the SPD announced its orientation towards a “new middle”, the German Institute for Economic Research (DIW) published a study on the shrinking middle class and rising income inequality in Germany An English version of the report is available on the website of the DIW: http://www.diw.de/documents/publikationen/73/84186/diw_wr_2008-4.pdf (accessed: 8 September 2008)

[There is] a strong tendency (…) to think of parties both empirically and normatively

in terms of the European mass party of integration (Katz/Müller 1997: 172)

Obviously, the notion of the mass party has also served as the basis for a characterisation of political parties in Africa and elsewhere in non-European countries as ‘different’ or ‘weak’

Taken together, party change in Western democracies and the weaknesses

of non-European parties indicate a similar tendency On this basis, Philip Schmitter suggests that the “crisis of representation/intermediation through partisan channels is generic” (Schmitter 1999: 491-492) While it remains doubtful, whether, as Schmitter argues, other channels of collective representation (interest groups, social movements) could replace political parties, the question of the relationship between the processes in old and new democracies is surely of some interest (see also Betz/Erdmann/Köllner 2002: 2) Likewise, Carothers (2006: 7) asks whether

the widespread weakness of and unhappiness with parties is a sign of some deeper evolution in global politics away from parties altogether

Diamond and Gunther (2001: xii) locate the disaffection with parties in a broader context of disaffection with democratic institutions They state that in Japan, North America and Europe, confidence in the government is in decline and people are cynical about their representative institutions (2001: ix) Thus, is the worldwide weakening of parties a loss of their ability to serve as a transmission belt between state and society and to connect citizens to the political process? If this holds true, then there will be serious implications for democracy as well The answer to this question largely depends on the understanding of democracy itself If democracy is thought of as a participatory system, in which political parties are a means for the representation of citizens’ interests, then these developments are indeed worrying Alternatively, political parties can be regarded as gate-keepers that lend legitimacy to a given system by filtering and aggregating societal demands before they can potentially threaten the whole system (Wiesendahl 1980: 202) In this version, parties are instruments for the containment of interest conflicts A third reading would suggest that parties are only brokers of group interests in a political market According to this view, the primary function of political parties is to get elected and allocate positions of power to their leaders As a result, interest articulation

is just a by-product of vote-maximizing strategies (see Wiesendahl 1980: 206) For the last two conceptions, a disentanglement of parties from their social milieus and well-defined groups is not an immediate problem to the functioning

of democracy

New Challenges: Neo-democracies and Party Change 33

Katz and Mair (1995: 6 and 21-22) remark that the type of the cartel party corresponds with an understanding of democracy as a political market However, the logic of the political market has its own inherent tendency of self-destruction As the same authors state:

electoral democracy may be seen as a means by which the rulers control the ruled, rather than the other way around (ibid.: 22)

Instances of misconduct of politicians in office are to be found across the world, and they feed the sense that “politicians are a class unto themselves” (Diamond/Gunther 2001: xii), who do not care about the real problems of their people Decreasing electoral participation is in fact a sign of a disenchantment with politics, and while neo-liberal economic policies are reinforcing social inequalities, underprivileged groups may not find their concerns represented in the contemporary parties (see also Norris 2002: 107) Although to varying degrees, there is a parallel between new and old democracies: political parties are suggesting an illusionary pluralism, while in reality they are instruments of elite circulation In the long run, this could delegitimise existing systems and give rise to parties like those of the right in Europe mentioned by Katz/Mair (1995: 24), which can escalate protest against established politics into xenophobic attitudes Alternatively, new issue-based parties could enter the stage and renew ideological competition

What is almost sure is that the cartel party is not the final stage of party development Neither the properties nor the functions of political parties are fixed once and for all, and the real types of parties are very much determined by their social and political environment Empirical findings from the new democracies are especially valuable in this regard: they help to free party research from elements which were thought to be universally applicable, but which are in reality bound to specific developments in Western Europe For example, research on single parties in the Third and Second World in the 1960s revealed that electoral competition cannot be a component of a definition for political parties (Janda 1999: 165; see also Randall 1988: 4) An inclusion of non-Western societies into the research agenda would provide thought-provoking new insights that could alter our understanding and definition of a political party Certainly, that would necessitate a flow-back of research results from such areas into the main field of political science Unfortunately, so far this has not always been the case To cite just one example, in a German introduction to comparative political science, edited by Berg-Schlosser and Müller-Rommel (2003), Thomas Poguntke wrote a chapter on the state of the art

in “International Comparative Party Research” (emphasis added), which fails to

mention parties outside Europe and the United States with even a single word

While this is still symptomatic, there are also signs of an increasing awareness and sensitivity to the problems relating to political parties in Non-Western societies In their essay on party types, Richard Gunther and Larry Diamond (2001: 4) notice that existing models of political parties do not adequately capture the full range of variation in party types and suggest a universal typology that is explicitly inclusive of Third Wave democracies.15

Especially noteworthy is the journal Party Politics which was founded in 1995

and which contains contributions about parties in all regions of the world In

2005, a special issue on Africa was published

Whether the formulation of a general theory of political parties will ever be possible is doubtful Gunther/Montero (2001: 21) fear that this search “may prove to be as fruitless as the search for the Holy Grail”, but they are optimistic that a formulation of more compelling middle range theories will be possible (ibid.: 4)

2.4 Democratisation, Hybridisation and Political Parties

Party research cannot be separated from the broader topic of democratisation and its problems in Africa The early 1990s saw an outburst of studies dealing with the democratic prospects of the African continent, but enthusiasm dwindled soon and gave way to more pessimistic evaluations In fact, the balance sheet of democratisation is very mixed: there are successful cases as well as cases where authoritarian rule still prevails While democracy in Africa has fallen short of overly optimistic expectations, there has surely been some advancement Gyimah-Boadi (2004: 6-8) reminds us of the positive outcomes:

ƒ democratic reforms have helped to make the African state less autocratic

ƒ media censorship has been relaxed

ƒ new constitutionalism and new liberal constitutions in a number of countries

ƒ civil societies are gaining greater sophistication and are moving away from crude anti-government and anti-state confrontations toward building consensus, fostering moderation and enriching the policy process

ƒ the legal and political status of parliaments has improved

15 For more on the applicability of this typology to African parties, see Erdmann (2002)

Democratisation, Hybridisation and Political Parties 35

ƒ democratic developments have helped to expand the political space and enhance opportunities for citizen participation

ƒ human rights advocacy and elaboration of protection have been expanding The Freedom House-Index16 also indicates an upward trend towards greater freedom (see Figure 3 below) In 1988, only two African countries were classified as ‘free’, while thirty belonged to the ‘not free’ category In 2008, the number of ‘free’ cases has risen to 11, while the number of ‘not free’ countries has halved from 30 to 15 cases

Figure 3: Freedom House Ratings for Africa, Selected Years

Source: Freedom House, various years (www.freedomhouse.org)

Yet, Freedom House reveals another striking development: Although almost all African countries hold multiparty elections, the majority are still only ‘partly free’ In these cases, authoritarian, non-democratic practices coexist with formally democratic procedures For example, many countries hold regular and even free and fair elections, but certain civil liberties are severely limited This has spurred discussions about the quality of democracy – not only in Africa but also in other democratising countries Larry Diamond has pointed out a “gray zone between fullfledged democracy and outright dictatorship” (Diamond 2002: 23) In a first reaction, democracy theorists have tried to capture these developments in numbers of diminished subtypes, such as illiberal democracy, pseudo democracy, façade democracy and so on Collier and Levitsky (1997: 431) have counted more than 550 notions of “democracy with adjectives”; meaning that there are many more subtypes than countries being analysed (ibid.:) Again, the assessment of democratic quality is determined by the understanding of democracy itself As is the case with many other concepts, democracy is not unequivocally defined and is not free from normative connotations Today, many scholars resort to procedural, minimalistic conceptions which exclude social rights One of the most commonly used definitions is Dahls’ ‘polyarchy’ which focuses on free and fair elections and popular participation.17 While free and fair elections are a necessary feature of democratic regimes, a procedural definition bears the danger of ignoring informal practices that could distort democratic structures Writing on Central America, Terry Lynn Karl (1995: 80) noticed:

Gains in the electoral arena have not been accompanied by the establishment of civilian control over the military or the rule of law Elections are often free and fair, yet important sectors remain politically and economically disenfranchised

Karl referred to these regimes as “hybrid regimes that combine elements of both authoritarianism and democracy” (ibid.: 73) This terminus has been taken up by other authors, for example by Friedbert Rüb (2002), who tried to establish analytic criteria for hybrid regimes For Africa, Gero Erdmann (2002) has linked the notion of hybridity to neopatrimonial rule In neopatrimonial regimes, an informal system of clientelistic relations penetrates a formal bureaucratic system and renders democratic institutions ineffective (see also chapter 5.3 in this thesis)

17 Dahl developed his theory in his books “A Preface to Democratic Theory” (1956); “Polyarchy” (1971) and “Democracy and its Critics” (1989)

Democratisation, Hybridisation and Political Parties 37

While there is little agreement on terminology in democracy studies, it is becoming clear that the imperfectly democratic regimes in Africa and other parts of the world are no longer in transition Thomas Carothers has compellingly criticised the assumption that democratisation would follow the stages of opening, breakthrough and consolidation (2002: 7) Neither did the successful transitional countries follow this pattern (ibid.: 15), nor did democracy in Europe evolve in this fashion (Berman 2007: 37) The disillusionment about the quality of progress in Third Wave countries has revived the debate on the preconditions of democracy In the past, authors like Lipset, Przeworski, or Dahl had argued that nations with a relatively high socio-economic development were much more likely to sustain democracy (see Schmidt 2000: 438-460 for a summary) Others held that democracy was historically tied to the existence of certain social classes Moreover, liberal

thinkers and critics from the left often insisted on the essential role of a

bourgeoisie (Beckmann 1989: 84) In short, this view is expressed in the famous statement by Barrington Moore: “no bourgeoisie – no democracy” (Moore 1966: 418) Elsenhans argues that, in contrast to industrialised countries, the elites in developing countries are ‘state classes’ with little economic power and

no source of accumulation that is independent from the state (Elsenhans 1997)

As a consequence, important socioeconomic conditions for democracy and an autonomous civil society are lacking (Elsenhans 1999: 261) Other factors which are believed to be conducive to democracy are ethnic homogeneity, certain cultural values, the separation of powers, and a democracy-friendly international environment (see Schmidt 2000: 446-447) Initially, the outbreak

of the Third Wave of Democratisation was a blow to preconditionists as it seemed to support the universalistic view that democracy could occur in all kinds of settings (Berman 2007: 28-29) For Africa, Bratton and van de Walle (1997: 129-130) have underlined that there was no direct correlation between economic development and democratisation: democratic change has occurred in very poor as well as in relatively wealthy countries But, as van de Walle (2002: 66) also admits, a stable democracy is difficult to establish under the condition

of low living standards, longstanding authoritarian traditions, little administrative capacity, an array of ethno-linguistic divisions and disastrous economic records Similarly, Sandbrook (2000: 23) calls attention to the hostile environment in which democratisation in Africa takes place Just recently, the debate on the necessary conditions for democracy has been renewed in various

articles in the Journal of Democracy Carothers (2007: 24) contends that levels

of economic development, concentration of sources of national wealth, based divisions, historical experience with political pluralism and nondemocratic neighbourhoods function as facilitators or nonfacilitators of

identity-democratic success Other authors stress the role of effective institutions Mansfield and Snyder (2005) hold that an effective state, the rule of law, organised parties, and news media must develop before democratisation can occur (see also Berman 2007: 30)

From this perspective questions of which democracy promoting functions

political parties can or should perform can be asked The latter view also

suggests that strong and effective parties are functional for stable democracies Appealing as this point may be – the relationship between the performance of political parties and democratic quality is not entirely clear Under autocratic and repressive regimes, it is unlikely that organised parties will develop before democratisation, precisely because authoritarian leaders often restrict the building of alternative organisations The first parties that appeared in the course

of political liberalisation in Africa were rather loosely organised opposition coalitions that pushed for their legalisation Many autocrats managed to control the speed of liberalisation and used the legalisation of political parties as a tool against the opposition, as was done by Jerry Rawlings In Ghana, the ban on political parties was lifted on May 1st 1992, presidential elections were scheduled for November and legislatives for December, leaving no time for oppositional groupings to transform themselves into well-organised parties (see also Jeffries/Thomas 1993: 339) This leads us to a general point of critique: if effective parties are necessary for democracy, how and when should they be built in the course of democratisation?

The relationship between party system properties and democracy is equally unclear Party systems with moderate fragmentation, high institutionalisation, and low polarisation are believed to be most favourable for democracy (Basedau 2007: 106) Although the findings in Africa are partly in line with theory, there are deviant cases that do not follow the expected pattern (ibid.: 130) Moreover, two thirds of non-authoritarian elections nevertheless produced dominant parties (Erdmann/Basedau 2007: 8)

While weak parties may have negative consequences for interest representation, they are not the only institutions with shortcomings (Carothers 2006: 11) Representation also requires a citizenry capable of expressing collective interests and a state capable of responding to these interests (ibid.: 12) This suggests that while democratic parties could be regarded as a necessary condition for democracy, preconditions for the development of efficient parties as may also exist Randall and Svåsand (2001: 15) name urbanisation, educational level and general improvement of living conditions as factors which enhance the development of efficient and democratic parties High levels of urbanisation and intact infrastructures facilitate communication with party sympathisers and voters, whereas improved education levels and higher

African Studies between Universalism and Exceptionalism 39

living standards raise both voter awareness and their ability to contribute financially to political parties (ibid.: 16) Furthermore, the continued persistence

of non-democratic practices limits the efficiency of opposition parties in various ways Election campaigns are often an unlevel playing field in which incumbents “obstruct, weaken, harass and divide the opposition” (Olukoshi 1998: 29) Incumbent parties frequently try to determine the rules of electoral competition to their own advantage, they deny the opposition access to the media, make use of state resources for their own election campaigns, try to manipulate voters’ registers, or introduce constitutional changes in order to keep potentially dangerous oppositional forces away from power Under these circumstances, the growth of a strong parliamentary opposition which can hold ruling parties responsible is seriously endangered For this reason, high expectations in the democracy-promoting capacities of African political parties have to be warned against

The relationship between political parties and democratic quality seems to

be – at least in Africa and possibly in other new democracies – a egg-problem: parties should enhance and improve democracy, but at the same time, the hybrid environment of authoritarian and democratic practices undermines the very organisational capacities of political parties Here we have come full circle: ‘weak’ parties correspond with ‘weak’ democracies in nearly all Third Wave-countries There is not enough space here to sort out causes and

chicken-and-effects, but the striking fact is that conventional wisdom of party research and

democracy studies, which is derived from Western European examples, can not

be easily reconciled with these developments

Is it simply a problem of perspective, do we have to change they way we are looking at things and abandon Western conceptions? Before we can draw conclusions for a framework of analysis, it makes sense to take a closer look at the Western bias and otherness of Africa, because these problems lie at the core

of Area Studies

2.5 African Studies between Universalism and Exceptionalism

2.5.1 Research Paradigms in African Studies

In the chapters 2.2 and 2.4 it was noted that Africa is often described in negative, general terms: African parties are weak, African democracies are defective or illiberal, the African state is weak or has even failed, the whole

continent is economically underdeveloped, and so on and so forth This is not to

say that all authors use these terms in such a pejorative fashion, it is rather the expression of a specific discourse which has almost become the exclusive form

of describing Africa We simply have no other words available than those with a derogatory undertone Mbembe put it:

While we now feel we know nearly everything that African states, societies, and

economies are not, we still know absolutely nothing about what they actually are (Mbembe

2001: 9)

Two related questions arise from the assumption that Africa is different First, stating that something is different requires a basis for comparison: different

from what? Secondly, what justifies us speaking of Africa as one entity?

In a book with the telling title, The African Exception (2005: 17), Patrick

Chabal has underlined that, ideal-typically, there are two different approaches in African Studies:

The first considers that Africa ought to be studied according to the theories of (economic and political) development that apply, or are thought to apply, to the rest of the world ( ) The second holds that Africa is a special case, an area of the world different from the rest and that it needs to be studied separately

In the field of party research, we find an insightful example In 1999, Gero Erdmann proposed to subsume political parties in Africa under the type neopatrimonial party, thus relating them to the area-specific debate on neopatrimonial states In 2002, he withdrew his area-specific type in favour of the inclusion of African parties into the universal typology of Gunther/Diamond, which was first published in 2001 This example demonstrates the necessity of debating both area-specific and universal conceptions, as both approaches have strong and weak points

2.5.2 Universalism

Universalists claim that all regions of the world can be analysed under the same universal categories In practice, we are quite often confronted with exceptions, namely when reality deviates from theoretical assumptions As we have seen, this is the case with democratic development and political parties in Africa Even if we leave the clearly negative connotations of diminished subtypes of democracy aside, we can only identify an aberration or exception if we take an

African Studies between Universalism and Exceptionalism 41

assumed condition of normalcy as the basis for comparison But how do we know what is normal?

Very often, the reality of democracies or institutions like political parties in Africa is pitted against an ideal of Western democracy This has taken the form

of an ahistorical, normative discourse close to modernisation theory Olukoshi (1998: 15) reminds us that:

There is, thus, no such thing as a ‘full’ or ‘pure’ democracy since the democratic process is constantly being renewed on local and global scales Static notions/models of democracy against which all other processes of democratization are measured, therefore, serve little other than an ideological-political end Yet, it is such static notions of democracy that underlie much of the attempt to understand the contemporary African experience of transition politics

Here, one could raise the objection that in Africa, democracy is not seen as a

“Northern idea being foisted on an unwilling South” (Bratton/Mattes 2001: 107) Based on the Afrobarometer Opinion Survey, Bratton and Mattes claim that the understanding of democracy in Africa is universal rather than culturally specific (ibid.: 110) As evidence they cite that 69% of the respondents refer to a procedural definition of democracy, while only 17% refer to substantive outcomes like peace and unity, social and economic development, and equality and justice (ibid) While there is nothing to say against the intention of the authors to highlight the approval of democratic procedures among African peoples, the characterisation of this as universalist is a bit odd The procedural

definition of democracy is currently the most common – but it is still one

particular reading of a contested concept The text of Bratton and Mattes is somewhat typical for the tendency in contemporary democracy discourse to equate universalism with liberal democracy or polyarchy, irrespective of the fact that universalist claims could also be made on the basis of alternative definitions

of democracy (see also Hartmann 2007: 413) Competing understandings of democracy include for example social rights and substantial participation in all stages of decision-making (see, among others, Robinson 1996) It is therefore not surprising that those Africans who expect democracy to deliver substantive goods as economic equality or social justice are also the least satisfied (ibid.: 118)

Proponents of procedural understandings often vindicate that the exclusion

of social progress makes their approach universal (see also Saul 1997: 342) However, the reduction of democracy to procedures disassociates the concept from its social origin and reconciles it with neo-liberal reform The dominant discourse conceptualises economic and political liberalisation together as good governance Yet, the equation of capitalism and democracy is misleading While