Single-Party Rule in a Multiparty Age: Tanzania in Comparative Perspective

Home , Abeid Karume, Daily News (Tanzania), Demographics of Tanzania, Ibrahim Lipumba, Jangwani, Joel Barkan

SINGLE-PARTY RULE IN A MULTIPARTY AGE: TANZANIA IN COMPARATIVE PERSPECTIVE

A Dissertation Submitted to the Temple University Graduate Board

in Partial Fulfillment of the Requirement of the Degree DOCTOR OF PHILOSOPHY

by Richard L. Whitehead August, 2009

ABSTRACT

Title: Single-Party Rule in a Multiparty Age: Tanzania in Comparative Perspective

Candidate's Name: Richard L. Whitehead Degree: Doctor of Philosophy Temple University, 2009 Doctoral Advisory Committee Chair: Richard Deeg

As international pressure for multiparty reforms swept Africa during the early 1990s, long- time incumbent, such as UNIP in Zambia, KANU in Kenya, and the MCP in Malawi, were simultaneously challenged by widespread domestic demands for multiparty reforms. Only ten years later, after succumbing to reform demands, many long-time incumbents were out of office after holding competitive multiparty elections. My research seeks an explanation for why this pattern did not emerge in Tanzanian, where the domestic push for multiparty change was weak, and, despite the occurrence of three multiparty elections, the CCM continues to win with sizable election margins.

As identified in research on semi-authoritarian rule, the post-reform pattern for incumbency maintenance in countries like Togo, Gabon, and Cameroon included strong doses of repression, manipulation and patronage as tactics for surviving in office under to multiparty elections.

Comparatively speaking however, governance by the CCM did not fit the typical post-Cold-War semi-authoritarian pattern of governance either. In Tanzania, coercion and manipulation appears less rampant, while patronage, as a constant across nearly every African regime, cannot explain the overwhelming mass support the CCM continues to enjoy today.

iii

Rather than relying on explanations based on repression and patronage alone, I locate the basis of post-reform CCM dominance in a historical process whereby a particularly unique array of social and economic policies promulgated during single-party rule culminated in comparatively affable social relations at the onset of multiparty reform. In Tanzania, this post- independence policy mix included stemming the growth of vast regional wealth differentials, a rejection of ethnicity as a basis for organizing collective action, and the construction of a relatively coherent national identity. By contrast, in most other African cases, policies under single-party rule acted to reinforce many of those economic and ethnic divisions inherited at independence. These divisions in turn, acted as material and moral capital for organizing dissent against incumbency, and the consolidation of opposition parties following political reform.

ACKNOWLEDGMENTS

I first want to give recognition to some of the folks and institutions in East Africa that helped make this project a reality. I give special thanks to Professor Rwekaza Mukandala, from

Research and Education for Democracy in Tanzania (REDET), for his unique insights into politics in Tanzania generally, and the CCM specifically, as well as his help in fieldwork arrangements and collect data. Professor Ruth Shayo, from the Institute of Development Studies

(IDS) at the University of Dar es Salaam, also deserves special mention here, as she organized a seminar at the IDS, where I was able to present my work in progress to some thoughtful and perceptive students. More generally, I would like to thank all the folks at REDET, IDS, and the

Media Institute of Southern Africa in Tanzania, along with the staff at the Institute for Education and Democracy and the Kenya Human Rights Commission for their warm welcome, hospitality, and assistance. Finally, I give thanks all those party leaders that were kind enough to take the time to meet with me as I prodded for interviews.

I also pay tribute to my friends and colleagues at the University of Bergen (UiB) and Chr.

Michelsen Institute (CMI). I am especially appreciative of the continual assistance from

Professor Lars Svåsand, of the Comparative Politics Department at UiB, who helped shape so much of my graduate work on political parties in the developing world. I am likewise appreciative to Professor Thorvald Gran, from the Institute for Public Administration and

Organizational Theory, also at UiB, for his remarkable assistance in grappling with the research questions for the dissertation.

The members of my dissertation committee also deserve special acknowledgements.

Professor Richard Deeg, at the Department of Political Science at Temple University, and Tony

Lucero, at the Henry M. Jackson School of International Studies at the University of Seattle, encouraged and challenged me to excel in my studies from start to finish. I likewise want to thank Professor Elke Zuern at Sarah Lawrence College, for her invaluable insights into social movements along with her methodological critiques and suggestions. Finally, I want to pay tribute to Professor Niyi Akinnaso, from the Anthropology Department at Temple University, for jumping in at the last minute to offer brilliant feedback as an external reader for my dissertation defense.

Finally, I reserve the most solemn gratitude to my wife, Wenche Snekkevik, and son,

Marius Snekkevik, for tolerating my complicated work schedule while researching and writing this piece. I could never have done this without the support from you both. My thanks also goes to Wenche for assisting me with some of the tables, figures, and proof reading. I am forever grateful.

TABLE OF CONTENTS

ABSTRACT...... iii

ACKNOWLEDGMENTS...... v

LIST OF TABLES ...... xii

LIST OF FIGURES...... xvi

CHAPTER

1. CONTENDING ACCOUNTS OF REGIME TENURE ...... 1

Probing the Basis for Incumbency ...... 5

Explaining Party Dominance...... 10

2. TOWARD A THEORY OF REGIME TENURE ...... 19

Toward a Theory of the Micro-Macro Making of Party Dominance ...... 21

Ideational and Material Resources and Macro-Structural Divergence...... 27

Micro-level Actions in the Formation of Party Dominance...... 32

3. RESEARCH METHODOLOGY ...... 48

Data Analysis: From ‘Discrete Causality’ to Casual Process...... 50

Case Selection: Finding Robust Comparisons ...... 58

4. REJECTING DISSENT – REPRESSION IN TANZANIA...... 62

Repression and Dominant Parties...... 64

Batons, Guns, and Prisons: Repression Under Single-Party Rule...... 66

Repression after Multiparty Reform...... 90

Conclusion...... 101

vii

5. SOCIAL-STRUCTURAL CONDITIONS AND SINGLE-PARTY RULE ...... 103

The Material and Ideational Substance of Political Contestation...... 105

TANU Dominance in Comparative Perspective ...... 109

The Colonial Background ...... 115

Political Factions and the Rising Party Leadership Stratum ...... 123

The Structural Underpinnings of the Rising Leadership Stratum ...... 125

Race, Class, and the Crisis on the Isles ...... 133

Conclusion...... 144

6. AFRICAN SOCIALISM, IDENTITY AND EQUALITY ...... 146

Collaboration and Coercion: The African Post-Colonial State ...... 148

Collaboration and the Post-Colonial State in Tanzania...... 153

Consent and the Post-Colonial State in Tanzania...... 156

Language Policy...... 166

Education for Self-Reliance ...... 170

Rural Development...... 173

Nationalization ...... 183

Party/State Leadership...... 189

Conclusion...... 193

7. AID AND THE REFORM STIMULUS ...... 196

Reform Stimulus ...... 199

The CCM and the Rise of the Bureaucratic Bourgeoisie ...... 206

New Complications ...... 210

8. THE MATERIAL BASIS FOR ORGANIZING CONTENTION...... 213

Economic Development and Capitalization ...... 214

viii

Class and Political Divisions...... 219

Conclusion...... 236

9. INEQUALITY AS A BASIS FOR CONTENTION...... 238

The CCM’s Populist Adversary ...... 240

Economic Inequality...... 243

Inequality and Political Orientations...... 250

Conclusion...... 252

10. ETHNICITY AS A BASIS FOR CONTETION...... 254

Ethnicity in Theory...... 255

Ethnicity in Tanzania...... 257

Data on Ethnic Saliency ...... 259

Ethnicity and Political Contention ...... 262

Conclusion...... 277

11. RELIGION AS A BASIS FOR CONTENTION ...... 279

A Survey of Religion in Tanzania...... 280

Saliency of Religion ...... 285

Religion and Political Contention ...... 289

Conclusion...... 297

12. NATIONAL IDENTITY AND THE CONVERGENCE OF SOCIAL FORCES ...... 299

The Exceptional Nation...... 301

Vanguard of National Unity...... 307

Branding the CCM ...... 314

Colonial Stooges and Arab Conspirators ...... 319

The Child and the Petty Squabbler...... 320

The Crook, the Thug, and the Devil...... 322

Machetes and Murder...... 323

The Genocidal Agent...... 324

CCM Guardianship...... 326

Caricature Diffusion...... 328

Conclusion...... 334

13. TANZANIA IN COMPARATIVE PERSPECTIVE ...... 335

Argument in Review ...... 336

The CCM in Comparative Perspective...... 341

Generalizing to Non-African Cases ...... 361

14. CONCLUSION ...... 371

New Challenges for the CCM ...... 372

Insinuations and Implications...... 382

Democratic Transitions and Consolidation ...... 382

African Socialism...... 386

Clientelism ...... 393

Two Models of Party Dominance ...... 397

REFERENCES CITED ...... 399

APPENDIXES

A. DATA FROM COUNTRY REPORTS ON HUMAN RIGHTS...... 442

B. ELECTION TURNOUT FOR SELECTED COUNTRIES...... 454

C. REGIONAL RESULTS FOR THE 1995 PRESIDENTIAL ELECTION ...... 456

D. REGIONAL RESULTS FOR THE 1995 PARLIAMENTARY ELECTIONS ...... 463

E. REGIONAL RESULTS FOR THE 2000 PRESIDENTIAL ELECTION...... 470

F. REGIONAL RESULTS FOR THE 2000 PARLIAMENTARY ELECTIONS...... 477

G. REGIONAL RESULTS FOR THE 2005 PRESIDENTIAL ELECTION ...... 483

H. REGIONAL RESULTS FOR THE 2005 PARLIAMENTARY ELECTIONS ...... 490

I. PROVINCIAL RESULTS FOR THE 1992 KENYAN PRESIDENTIAL ELECTION ...... 497

J. PROVINCIAL RESULTS FOR THE 2002 KENYAN PRESIDENTIAL ELECTION...... 500

K. REGIONAL RESULTS FOR THE 1994 MALAWI PRESIDENTIAL ELECTION ...... 503

LIST OF TABLES

Table Page

1. TABLE 3.01: CASE COMPARISON SCHEMATIC...... 61

2. TABLE 4.01: QUESTIONING DURING CAMPAIGN – MOROGORO URBAN, 1985...... 83

3. TABLE 4.02: SAMPLE TURNOUT FOR THE 1990 ELECTION ...... 89

4. TABLE 4.03: REASONS FOR REGISTERING TO VOTE...... 92

5. TABLE 4.04: EVALUATING THE UPCOMING 2005 ELECTIONS ...... 92

6. TABLE 5.01: RESULTS OF THE 1960 LEGISLATIVE COUNCIL ELECTIONS IN TANGANYIKA ...... 110

7. TABLE 5.02: RESULTS OF THE 1961 LEGISLATIVE COUNCIL ELECTIONS IN KENYA ...... 113

8. TABLE 5.03: RESULTS OF THE 1968 NATIONAL ASSEMBLY ELECTIONS IN ZAMBIA ...... 114

9. TABLE 5.04: PRE-INDEPENDENCE ELECTION RESULTS ON ZANZIBAR...... 140

10. TABLE 6.01: ETHNIC REPRESENTATION IN GOVERNMENT ...... 167

11. TABLE 6.02: GROWTH OF INCOME INEQUALITY IN TANZANIA ...... 177

12. TABLE 6.03: NUMBER OF VILLAGES BY REGION...... 180

13. TABLE 7.01: FOREIGN FINANCING DURING SINGLE-PARTY ERA ...... 200

14. TABLE 7.02: EXTERNAL DEBT – PERCENTILE RANK ...... 201

15. TABLE 8.01: OCCUPATION OF FOUNDING MEMBERS...... 220

16. TABLE 8.02: IMPACT OF NATIONAL GOVERNMENT ON DAILY LIFE ...... 222

xii

17. TABLE 8.03: IMPACT OF LOCAL GOVERNMENT ON DAILY LIFE...... 222

18. TABLE 8.04: “ZONAL” VARIATION IN INCUMBENT SUPPORT ...... 226

19. TABLE 8.05: EDUCATION OF FOUNDING MEMBERS ...... 227

20. TABLE 8.06: EDUCATION VERSUS PARTY IDENTIFICATION ...... 227

21. TABLE 8.07: EDUCATION STATUS ACCORDING TO PARTY MEMBERSHIP ...... 229

22. TABLE 8.08: PARTY MEMBERSHIP ACCORDING TO EDUCATIONAL STATUS ...... 230

23. TABLE 8.09: VARIATION IN PARTY SUPPORT ACCORDING TO OCCUPATION...... 235

24. TABLE 9.01: HOUSEHOLD INCOME INEQUALITY AT INDEPENDENCE...... 245

25. TABLE 9.02: TODAY’S ECONOMIC INEQUALITY...... 247

26. TABLE 9.03: GINI COEFFICIENT FOR INCOME...... 249

27. TABLE 9.04: YOUR OWN STANDARD OF LIVING COMPARED TO OTHERS ...... 249

28. TABLE 9.05: RATINGS OF GOVERNMENT EFFORTS AT ADDRESSING INCOME INEQUALITY...... 250

29. TABLE 9.06: CORRELATION COEFFICIENTS FOR SIGNIFICANT ECONOMIC INDICATORS ...... 251

30. TABLE 10.01: OPINIONS OF INTER-CLEAVAGE MARRIAGE ...... 260

31. TABLE 10.02: OWN ETHNIC GROUP’S REALTIVE ECONOMIC CONDITIONS...... 260

32. TABLE 10.03: OWN ETHNIC GROUP’S RELATIVE POLITICAL INFLUENCE...... 261

33. TABLE 10.04: OWN ETHNIC GROUP TREATED UNFAIRLY...... 261

xiii

34. TABLE 10.05: ETHNIC VERSUS NATIONAL IDENTITY...... 262

35. TABLE 10.06: PRESIDENTIAL ELECTION CONSTITUENCY RESULTS IN KILIMANJARO ...... 267

36. TABLE 10.07: PRESIDENTIAL ELECTION CONSTITUENCY RESULTS IN SHINYANGA...... 268

37. TABLE 10.08: LANGUAGE ACCORDING TO POLITICAL AFFILIATION ...... 271

38. TABLE 10.09: POLITICAL AFFILIATION ACCORDING TO LANGUAGE...... 272

39. TABLE 10.10: LANGUAGE ACCORDING TO POLITICAL AFFILIATION (KENYA)...... 273

40. TABLE 10.11: LANGUAGE ACCORDING TO POLITICAL AFFILIATION (MALAWI)...... 275

41. TABLE 10.12: LANUGAGE ACCORDING TO POLITICAL AFFILIATION (ZAMBIA)...... 276

42. TABLE 11.01: DISCUSS POLITICS ACCORDING TO RELIGION ...... 285

43. TABLE 11.02: EDUCATION ACCORDING TO RELIGION...... 287

44. TABLE 11.03: REACTION TO DAUGHTER MARRYING A MAN FROM ANOTHER RELIGION...... 288

45. TABLE 11.04: RELIGIOUS BACKGROUND OF FOUNDING MEMBERS...... 291

46. TABLE 11.05: RELIGIOUS FAITH ACCORDING TO POLITICAL AFFILIATION ...... 292

47. TABLE 11.06: RELIGIOUS FAITH ACCORDING TO CANDIDATE SUPPORT ...... 292

48. TABLE 11.07: CANDIDATE SUPPORT ACCORDING TO RELIGIOUS FAITH ...... 293

49. TABLE 11.08: POLITICAL AFFILIATION ACCORDING TO RELIGION ...... 293 xiv

50. TABLE 11.09: RELIGION ACCORDING TO POLITICAL AFFILIATION ...... 294

51. TABLE 11.10: RELIGION ACCORDING TO PARTY AFFILIATION IN KENYA...... 295

52. TABLE 11.11: RELIGION AND PARTY AFFILIATION IN MALAWI ...... 296

53. TABLE 12.01: CONTINUATION OF THE SINGLE-PARTY SYSTE,...... 310

54. TABLE 12:02: ADOPTION OF A MULTIPARTY SYSTEM...... 310

55. TABLE 12.03: SHOULD CHANGES BE IMMEDIATE? ...... 311

56. TABLE 12.04: BENEFITS OF THE MULTIPARTY SYSTEM...... 311

57. TABLE 12.05: POSSIBLE EFFECTS OF MULTIPARTY SYSTEM ...... 312

58. TABLE 12.06: PARTY PREFERENCE...... 312

59. TABLE 12.07: MEASURES OF CAMPAIGN PENETRATION...... 329

60. TABLE 12.08: VIEW OF VOTERS AND LEADERS ...... 331

61. TABLE 12.09: APPROVAL OF ONE-PARTY RULE...... 332

62. TABLE 12.10: POLITICAL PARTIES ARE DIVISIVE VS. MANY PARTIES NEEDED...... 333

LIST OF FIGURES

Figure Page

1. FIGURE 1.01: BASIC MODEL FOR CONDITIONING POST- REFORM MOBILIZATION...... 7

2. FIGURE 1.02: COMPARATIVE SHARE OF VOTES IN PRESIDENTIAL ELECTIONS ...... 11

3. FIGURE 1.03: SHARE OF VOTES IN PARLIAMENTARY ELECTIONS ...... 12

4. FIGURE 1.04: SHARE OF VOTES IN PRESIDENTIAL ELECTIONS ...... 12

5. FIGURE 2.01: MACRO-MICRO FRAMEWORK ...... 26

6. FIGURE 2.02: POLARIZATION CONCEPTUALIZED...... 29

7. FIGURE 2.03: MACRO-STRUCTURAL DIVERGENCE AND CLIENTELISTIC CONTROL ...... 31

8. FIGURE 2.04: POLICY IMPLICATIONS OF MACRO- STRUCTURAL DIVERGENCE ...... 39

9. FIGURE 2.05: MULTIPARTY CONSEQUENCES OF POLICIES DURING SINGLE-PARTY RULE ...... 41

10. FIGURE 4.01: POLITICAL RIGHTS ...... 74

11. FIGURE 4.02: CIVIL LIBERTIES...... 75

12. FIGURE 4.03: POLITICAL RIGHTS FOLLOWING REFORM ...... 93

13. FIGURE 4.04: CIVIL LIBERTIES FOLLOWING REFORM...... 94

14. FIGURE 4.05: POLITICALLY RELATED HUMAN RIGHTS VIOLATIONS...... 96

15. FIGURE 5.01: PERCENTAGE OF SEATS WON BY LARGEST PARTY IN EARLY ELECTIONS...... 111

xvi

16. FIGURE 5.02: THE ZANZIBAR ISLANDS...... 135

17. FIGURE 7.01: GROSS DOMESTIC PRODUCT IN SELECTED COUNTRIES...... 197

18. FIGURE 7.02: ANNUAL PERCENTAGE CHANGE IN GDP PER CAPITA...... 197

19. FIGURE 10.01: REGIONS OF TANZANIA ...... 266

20. FIGURE 7.01: GROSS DOMESTIC PRODUCT IN SELECTED COUNTRIES...... 197

xvii

CHAPTER ONE CONTENDING ACCOUNTS OF REGIME TENURE

“Results! Why, man, I’ve gotten a lot of results! I know several thousand things that won’t work.”

Thomas Edison, circa 1900

It has been widely argued within the comparative politics literature that single-party, military, and personal rule throughout Latin America, Asia, and Africa were regime types characteristic of the structural conditions that prevailed amid the power-balancing politics of the

Cold War. Whether referring to American-backed rule in Zaire by Mobutu, or Ghana’s especially cozy relations with the Soviet Union during the 1960s, authoritarian restrictions on political participation and competition was often overlooked or encouraged by contending superpowers simply because stable but exclusive rule acted to contain the spread of revolutionary or reactionary forces with the potential for upsetting powerful geopolitical interests. At the same time, within the corridors of the academy, there was a general consensus among modernization theorists and communist vanguard thinkers that authoritarian rule was necessary for steering potentially divided and underdeveloped societies down the path of capitalist development or socialist revolution respectively. In short, the configurations of intellectual world views and the distribution of material capabilities at the international stage constituted the critical causal forces for sustaining authoritarian rule throughout much of the

“third world”1.

1 While the term “third world” might be taken as offensive to some, other terms, such as the “global south” or “developing” and “underdeveloped” each elicit potentially offensive symbols or are simply misleading descriptors. Throughout the remainder of this work, I will generally refer to those countries with historically low volumes of GDP as underdeveloped or third world, while recognizing the caveats associated with each term. However, since the early 1990s, the number of these so called “hard authoritarian regimes,” where legal provisions prohibited political participation and competition, has dramatically declined (Diamond, 2002). Due to a rising unipolar world, which elicited demands for expanding political competition and participation in the third world, and a corresponding economic crisis during the 1980s, the international political conditions which sustained personal, military, and single-party rule more or less eroded. These changes opened new opportunities for entrepreneurial politicians to challenge incumbents and provided the moral justification for mobilizing dissent against authoritarian, albeit with varying levels of contestation from one country to the next.

In step with the erosion of these economical and political conditions, over the past two decades the comparative literature on political regimes has morphed into two competing interpretations of what was actually taking place in the post-Cold War world. First, the conclusion of military rule in a number of Latin American countries during the 1980s, the collapse of communism in Eastern Europe during the late 80s and early 90s, and the adoption of multiparty reforms in Africa throughout the 1990s, breathed new life into the democratization literature that identified the emergence of a “third wave” of transitions to democratic rule.

Broadly speaking, the third wave literature identifies variables, such as massive third-world debt, increased international pressure from powerful financial institutions and first-world foreign ministries, as well as emboldened domestic social movements, as the key contributors toward the emergence of regimes on a trajectory toward electoral and liberal democracy. Names such as Lech Wałęsa and Václav Havel were touted by western liberal thinkers and policy- makers as pioneers paving the way for the a “global resurgence of democracy”.

As if to mock the democratization theorists, by the end of the 1990s, it became increasingly obvious to many observers that the initial enthusiasm behind the third-wave thesis may have been based on an overzealous faith in the power of liberal ideals and the unrealistic assumption that the process of reforming or terminating hard authoritarian rule was synonymous with the process of democratization. While a number of reformed countries continued to be governed by long-time incumbent regimes, and while new regimes proved only marginally less authoritarian than the incumbents they replaced, third wave theorists began talking of “stalled transitions” or

“guided democratization”. Another school of thought rejected the tendency to conceptualize the bulk of the regime changes during the 1980s and 90s as a transition to democracy at all. Instead, in domestic milieus characterized by fragmented power structures revolving around subnational sentiments and the absence of the rule of law (Linz & Stepan 1996; Hadenius, 2001; Mansfield

& Snyder, 2007), new multiparty elections were culminating in new forms of stable nondemocratic rule strategically adapted to international changes that demanded more political competition and participation. Rather than undergoing transitions to democratic rule, what was instead underway was the emergence of softer, yet durable forms of authoritarian rule.

Variously referred to as “electoral authoritarianism” (Schedler, 2002), “soft authoritarianism”

(Means, 1996), or “virtual democracy” (Joseph, 1999), these ‘softer’ forms consist of regularly held and sometimes relatively fraud free multiparty elections. However, these regimes manage to survive by periodically harassing the oppositionists and defending an institutional environment which remains enormously partial to the incumbent (Levitsky & Way, 2002).

The effect of the shifting international climate for regimes throughout the third world is the backdrop for the main research concern in this piece. Why do some incumbents survive the transition from harder to softer forms of authoritarian rule, while other incumbents fail? Stated

3 differently, why did the emergence of new opportunity structures, including international reform pressures and multiparty elections, result in varying abilities for political entrepreneurs to mobilize dissent against single-party rule and cultivate support during multiparty elections? In some cases, these entrepreneurs were able to unseat long-time incumbents, in other cases they were not.

Probing the Basis for Incumbency

The principal research question here does not directly seek to explain the causes of semi- authoritarian rule into the post-cold war era per se. Instead, in its most narrow form, the question

I ask is why some regimes or political parties emerge from the upheavals of multiparty reform without facing serious electoral challenges? The first answer to this question is an obvious one: aside from half-hearted commitment to multiparty reforms, genuine political reforms are blocked by incumbents through repeated attempts at thwarting the formation and effectiveness of opposition parties or social movements by harassing dissenters, maintaining an effective command over the media, or blatantly rigging multiparty elections. To be sure, the failure of the opposition in Togo to unseat the incumbent Rassemblement du Peuple Togolais (RPT) is without a doubt in large part due to the use of military force, election rigging, and constitutional manipulation on the part of former president Étienne Eyadéma and his son president Faure

Gnassingbé. Similarly, in Gabon President Omar Bongo Ondimba and the Parti Démocratique

Gabonais (PDG) continued to dominate elections since the 1991 constitutional reforms by engaging in questionable ballot practices. Alexander Lukashenko’s rule in Belarus has persisted amid a fragmented array of opposition parties and blatantly unfair elections.

Indeed, it would be difficult to classify these regimes as having demonstrated any real commitment to multiparty reform in the first place. However, what about those cases where governing elites have more clearly engaged in multiparty reforms by adopting multiparty constitutions, less frequently using force against opponents, and conducting relatively fraud free elections, all while continuing to govern in a climate where opposition parties and movements fail to organize and draw large crowds of support? Why do these regimes, perhaps best represented by the Chama Cha Mapinduzi (CCM) in Tanzania, continue to face opposition movements that are so disorganized, so numerically weak, and so incapable of challenging incumbent rule during election times? The answers to this question are far different than those answers provided for the RPT in Togo, the PDG in Gabon, or Lukashenko’s tenure in Belarus.

While coercion and rule manipulation forms the basis of incumbency in nearly every polity, as will be demonstrated later in this piece, widespread election-rigging in Tanzania is relatively restricted, as is widespread coercion, harassment, and intimidation.

In this exposition, my primary task is to explain the casual basis for dominant parties that cultivate favorable records on permitting political competition and participation. For the sake of measurement expedience, the level of tolerance toward political competition and participation is defined here by levels of civil liberties and political rights. At the abstract level, I argue that the causal underpinnings of party dominance requires an investigation into the historical self- reinforcing interaction between state policy on one hand, and the generative social structural context which sustains particular regime coalitions on the other hand. As alluded to above, my primary case study is the CCM in Tanzania. At the same time, I select a handful of contrasting secondary case studies, including Malawi, Zambia, and Kenya.

Figure 1.01 below provides a more specific, yet still general overview of the explanatory framework to be used throughout this exposition. State policies under single-party rule in

Tanzania, as listed in the left-hand side of the figure, included a comparatively strong emphasis on economic and political equality, a coherent national identity which spanned every corner of the country, and state control over most sources of economic accumulation. In short, Tanzania was to be the model African Socialist state, a term that will be described in more detail later.

These policies in turn yielded a macro-structural context, as specified in the center of the figure, characterized by a narrow spectrum of divergence between social forces, which in turn reduced the probability of mobilizing opposition to incumbency at the onset of multiparty reform.

Figure 1.01: Basic Model for Conditioning Post-Reform Mobilization

More specifically, the emphasis on political and economic equality, national unity, and state economic leadership during single-party rule in Tanzania had two important consequences for the level of what I call macro-structural divergence, or historical patterns by which social entities differentiate from one another. First, state economic leadership reproduced an economy defined by weak capitalist class structures. Secondly, with an emphasis on national cohesion and

6 egalitarianism during single-party rule, subnational loyalties around ethnicity, religion, and class failed to fester to the degree found elsewhere in Africa.

At the time of multiparty reform, this macro-structural environment, characterized by weak capitalist classes and highly diffuse and non-polarized subnational loyalties had two highly interrelated consequences for regime tenure. First, the CCM’s internal clientele relations were not as susceptible to the centrifugal forces that resulted in regime disintegration in other African cases. Secondly, those political entrepreneurs surviving outside the spatial reach of the regime’s clientelistic tenticals found an almost impossible environment by which to mobilize material resources and subnational loyalties around an opposition party agenda. In a word, the macro- structural environment at the time of multiparty reform ensured that contestation to the CCM would remain weak, allowing the regime to survive multiparty elections without engaging in blatant civil liberties and political rights violations. In the terms described in figure 1.01, the narrow spectrum of macro-structural divergence translated into a multiparty environment whereby the possibilities mobilizing ideational and material2 resources outside the party-state nexus where comparative minimal.

This is in stark contrast to that which played out in Kenya since the time of independence.

Here, lines of economic and political exclusion, which corresponded with ethnicity, intensified structural divergence, acting as an important impetus for mobilizing political contestation along ethnic lines. During the 1980s, when Jomo Kenyatta’s successor, Daniel Moi, tried to coercively reverse previous patterns of ethnic inclusion and exclusion, those negatively impacted political

2 Ideational and material resources are more thoroughly discussed in the next chapter. Suffice it here to say that the ideational realm implicates the values and norms that define what is to be resisted or accepted (Fangyin, 2007). Hence, I speak of ideas here as those factors which motivate people into social action and provide the moral framework for defining the essence of political struggles. By contrast, the material realm here refers to those factors which give ideas casual force in the political realm, i.e. the capacity to resist or institute domination. In reality, the two are “interwoven” and “interdependent” (Hay, 2001, p. 7).

7 entrepreneurs responded by organizing contestation in an environment conducive to mobilizing ideational and material resources outside the party-state nexus. This highly competitive environment characterized Kenya’s volatile transition to multiparty rule, botched multiparty elections, and ultimately constituted the basis for dislodging the Kenya African National

Union’s (KANU) rule in 2002.

The scope of this exposition takes place on two levels. First, the lion’s share of the text within these pages specifically addresses regime durability in Tanzania. Therefore, on one hand this piece can be interpreted as a study on the unique political processes in this country alone.

On the other hand, by selecting Mills Method of similarity and difference as my principal comparative tool, I am able to state some general propositions about regime survival beyond the specifics of the Tanzanian case. On the basis of finding similarly situated regimes that nevertheless resulted in a mix of outcomes, I select Kenya, Zambia, and Malawi as the most dissimilar cases. I also illustrate the differences in multiparty political contestation between the

Tanzania Mainland and Zanzibar to make my case. In the final chapter I talk at length about

Botswana as the most similar case.

On yet another level, the scope of this project reflects my ambitions of broadly contributing to the literature on party dominance generally. To be sure, the framework sketched in the following chapter will have value beyond these cases. Throughout other parts of Africa during single-party rule, the failure of leaders to pay attention to the role that resource inequalities and class formation had on ethnic identity did not bode well for forging party dominance durable enough to survive the twin pressures of international demands for good governance on one hand, and domestic challenges during multiparty elections on the other hand. Beyond Africa, my analysis will, at the very least have partial relevance to other cases of party dominance. In

Mexico, inferences about the bases to PRI’s long tenure are difficult to make without references to the historical reproduction of its peasant base through periodic appeals to the spirit of the

Revolution. In Malaysia, we see the limitations faced by the incumbent party, the United Malays

National Organization (UMNO), when trying to liberalize the political process, while at the same time, trying to maintain a hold on state power in a context where strong structural divergence is pulling at the seams of political stability. Before explicating more specifically the model described in figure 1.01, it is essential to 1) reveal the comparative level of dominance in

Tanzania and 2) highlight some of the contending ways in which party dominance is described within the comparative politics literature.

Explaining Party Dominance

The level of party dominance in Tanzania is described in the three following figures. Figure

1.02 defines the multiparty election success of the CCM in comparative perspective. The highly divided group of opposition parties in first multiparty elections in Kenya, while failing to remove Moi and KANU from office, managed to attract enough support to reduce incumbent victory to a mere plurality rather than majority. In Malawi, after a referendum in 1993 revealed an overwhelming support for multiparty change, opposition parties contested against President

Hastings Banda and the Malawi Congress Party (MCP) in the elections of 1994. The results of the election placed Bakili Muluzi and his United Democratic Front (UDF) in control of the executive and legislative branches. Perhaps the most dramatic post-reform election in East

Africa took place in Zambia in 1991. Here, the long-time incumbent, President Kenneth Kaunda and the United National Independence Party (UNIP), faced a stiff challenge from the Movement

9 for Multiparty Democracy (MMD) and Frederick Chiluba, who overwhelmingly defeated

Kaunda and reduced UNIP to a small group of opposition members of parliament.

Figure 1.02: Comparative Share of Votes in Presidential Elections

Source: African Elections Database, 2007k, m, s, w

In the first post-reform multiparty election in Tanzania, some of the opposition parties did perform well in certain regions of the country. In the cash crop region of Kilimanjaro, the

NCCR-Mageuzi secured over 54 percent of the votes in the parliamentary election. The Civic

United Front (CUF) drew from a strong following on Zanzibar, CHADEMA did reasonably well in Kigoma, and the United Democratic Party (UDP) had a respectable showing in the party leader’s home region of Shinyanga. However, as the NCCR-Mageuzi’s presidential aspirant,

Augustine Mrema himself pointed out in recent years, “in 1995 elections we in the opposition had better performance as compared to 2000. And this happens because Tanzanians themselves do not have political will to take the ruling party out of power” (Stephen, 2005). Indeed, in the subsequent multiparty elections of 2000 and 2005, the CCM has eclipsed the opposition altogether in nearly every region. Only in the Zanzibar regions of Pemba North and Pemba

South did more voters prefer opposition to the CCM, most of which supported CUF. As figures

1.03 and 1.04 illustrate, overall support for the opposition parties in the presidential and parliamentary contests has decline with each election cycle.

Figure 1.03: Share of Votes in Parliamentary Elections

Source: National Election Commission, 1997, 2001, 2006

Figure 1.04: Share of Votes in Presidential Elections

Source: National Election Commission, 1997, 2001, 2006

An obvious place to begin when looking to explain election success of the CCM following multiparty reform is to focus on a collection of macro-structural conditions which are theoretically and empirically tied to the length of regime tenure. For starters, does Tanzania possess a natural resource profile that could act as a significant basis for rent collection, thus minimizing the need to tax the population (Beblawi, 1990; Smith, 2004)? The answer is an emphatic no. Since independence, Tanzania’s economy has been largely a peasant one, and only recently begun to focus on expanding the mining sector. Furthermore, while international aid is a significant source of income for nearly every regime on the African Continent, the adjustment programs which often accompany aid have effectively meant that aid inflows also correspond with declining food and fuel subsidies, increased cost-sharing in healthcare and education, and currency devaluations, all of which translate into increasing rather than decreasing the financial hardship of everyday life. Therefore, the rent-seeking hypothesis can be dismissed outright.

Other conditions might include ‘crisis indicators’, such as the rate of economic growth, inflation, and unemployment. When economic depression and high unemployment persists and inflation is rampant, widespread popular discontent may take to the streets. Rising economic stagnation and escalating political protests will consequently entice some regime insiders to consider reforms in effort to placate contention. Much of the contemporary literature on the relationship between economic crisis variables and regime change suggest that regime breakdown occurs when an economic crisis generates a legitimacy crisis that, in turn, undermines the elite bargains underpinning authoritarian regimes (O’Donnell & Schmitter,

1986; Haggard & Kaufman, 1995). Brazil's rapid economic decay in 1982 for example, which took place after years of inflation and the accumulation of one of the world’s highest foreign debt rates, provided an opening for street demonstrations against the military regime.

Recognizing the need to stem the rising tide of discontent, General Figueiredo presented the public with a compromise electoral college election. However, the incumbent regime itself was unable to maintain unity during the subsequent 1984 election, as many of the discontented electors simply voted for the opposition candidate, Tancredo Neves. Similar splits and popular protests took place in Argentina during the early 1980s, when a series of successive military rulers were unable to stabilize the economy after the massive currency devaluation of 1981.

After the Falkland Island invasion debacle, the military was utterly divided and discredited, resulting in the 1983 election, where Raúl Alfonsín emerged as the victor.

One would be hard pressed to argue that economic conditions are not central to the legitimacy of political regimes of any sort. Watching ones savings or earnings dwindle due to drastic increases in prices or being unable to provide basic needs for one’s family due to a declining job market is enough to upset even the most reserved person. To be sure,

12 combinations of economic weakness and the imposition of structural adjustment measures, including the introduction of school fees, labor retrenchment, and currency devaluation, have played havoc for a number of regimes throughout Africa. Such scenarios featured prominently in the removal of UNIP in Zambia, as unions were emboldened by reduction in salaries, subsidies, and jobs. The growth of Abdoulaye Wade’s opposition Parti Démocratique Socialiste

(PDS) in Senegal was largely a response to economic weakness. The “tactical retreat” of

Kérékou in Benin was due to the presence of energized protests against the collapse of the banking system (Nugent, 2004).

There are however, two inadequacies when linking crisis indicators to rising political protests and an incumbent’s eventual demise. First, the presence of grievances from economic hardship might not necessarily translate into political parties with the wherewithal to compete in multiparty elections. Undoubtedly, rising protests may force regimes to make some painful political concessions. The ability of entrepreneurial politicians to take advantage of the new opportunities from concessions will however, depend on a host of other factors, most notably the availability of financial resources and polarized identities.

Secondly, it is not even clear that political protests arising out of economic discontent necessarily leads to splits within the ruling regime. In Cameroon for example, the

Rassemblement Démocratique du Peuple Camerounais (RDPC), led by the highly aloof Paul

Biya, managed to remain relatively cohesive across several multiparty elections, despite intense periods of rising unemployment, declining wages, and escalating political protests. Perhaps the most notable case of prolonged tenure through periods of economic decline is seen in Zimbabwe under Robert Mugabe who, while blatantly repressing opposition, continues to preside over a

13 relatively cohesive regime and enjoys widespread support throughout a number of societal sections.

According to a number of well-known scholars, two institutional conditions will intervene between the state of the economy and the cohesiveness of the regime. First, Geddes (1999) demonstrates that, when compared to the more corporatist nature of military regimes in Latin

America, which are highly susceptible to regime splits when faced with a sour economy, single- party regimes are less affected by bad economic performance since they have the institutional arrangements for channeling discontent through things like elections and access to public officials. Secondly, as Bratton & van de Walle (1997) have extensively articulated, the neo- patrimonial3 regimes commonly found throughout Africa tend to display a remarkable degree of elite cohesiveness because the material benefits that do exists, are often distributed exclusively to regime insiders, at the expense of oppositionists and economic development generally. These perspectives suggest that, under single-party neo-patrimonial rule, economic decline may simply fail to generate those significant splits within the ruling elite that doomed many authoritarian regimes throughout Latin America.

Rather than simply splitting the regime between hardliners and reformers, single-party neo- patrimonial rule may be more significantly threatened by those individuals or groups excluded from the security of patronage, and thus more exposed to the hardships of economic decline

(Bratton & van de Walle, 1997). Under such circumstances, popular resentment and contention outside the spatial research of the regime’s clientele networks may grow into organized protests

3 The term neo-patrimonial rule here refers a “hybrid political system” where the rule based on individual “personal prestige and power” is melded with “rational-legal institutions” (Bratton & van De Walle, 1997, p. 62). While I refer to this concept at various points throughout this piece, I do so only to highlight the usage of the concept within scholarly works. At the same time, I do not subscribe to the typical explanations that attribute neo- patrimonial rule to the blending of traditional and legal-rational authority. In my view, the practices of unaccountability, despotism, and personal rule which are often attributed to neo-patrimonial rule are more clearly tied to patterns established under colonial rule rather than pre-colonial rule.

14 and social movements calling political reforms and possibly regime change. When paired with pressures from international financial institutions and donors, incumbents are forced to concede to popular demands by amending constitutions and scheduling multiparty elections.

While Bratton and van de Walle’s thesis describes the effects of neo-patrimonial rule on regime cohesion and societal protests, the neo-patrimonial framework cannot suffice as a singular explanation for the longevity of party dominance in Tanzania. For starters, neo- patrimonial rule is a concept equally applicable to nearly every country in sub-Saharan Africa, including the multitude of cases where multiparty reforms were followed by the demise of incumbency. With these varying outcomes in mind, the approach adopted in here treats neo- patrimonial rule as a constant. While single-party neo-patrimonial regimes in Zambia, Malawi, and Kenya failed to survive multiparty contestation, the CCM in Tanzania has shown remarkable resilience. Why do similarly situated regime types experience divergent outcomes?

Furthermore, by adopting the “retrospective voting” theory (Fiorina, 1981), we can see that the economic growth in the years of, and prior to, presidential elections in some of the cases discussed in this piece do not correlate with elite fragmentation or regime turnover. While

Tanzania has had some remarkable economic growth since the mid-1990s, like most African counties, Tanzania experienced a severe economic hardship in the late 1980s and early 1990s.

When looking at GDP growth rates around the time of multiparty change, we are left asking why the CCM overwhelmingly secured the 1995 elections, a year that followed three years of negative economic growth. If crisis indicators are important casual explanations, negative growth should have resulted in widespread protests and support for Tanzania’s opposition parties. Instead, the lead-up to multiparty reform and subsequent multiparty elections can be defined as comparatively uneventful and unchallenging for the incumbent.

Rather than addressing economic decline per se, perhaps there are civil society variables responsible for thwarting the emergence of organized dissent. To be sure, the girth of civil society literature rightfully recognizes the importance of organized citizenry in the push for democratization (Harbeson, 1994; Ndegwa, 1996; Gyimah-Boadi, 1997; Bauer, 1999) and the management of public policy in Africa (Baccaro, 2006). Where collective social life is uncoerced, self supporting, and confined by a legal order or set of shared rules, civil society organizations (CSOs) can act as a significant counterweight against would be unaccountable politicians (Diamond, Kirk-Greene, & Oyediran, 1997; Paffenholz & Spurk, 2006)).

The civil society literature offers up some potentially useful insights into explaining the lack of popular protests and organized opposition against party dominance. For example, it is reasonable to suggest that greater levels of civil society “vibrancy” or “capacity” should also lead to a greater potential for citizens to organize political contention, and consequently, a greater potential for grievances to be channeled into viable social movements and political parties. In turn, vibrancy or capacity depends on a host of structural factors. For example, where social capital is weak, political cultures submissive, and economic resources scarce (Almond &

Verba, 1963; Putnam, 1993), CSOs are either absent, ill-equipped, or not committed to challenging authoritarianism. Such conditions persist most notably in the contexts of economic underdevelopment, natural resource dependent economies, and pervasive authoritarian rule, which dominate all forms of association life through clientelism and repression (Lipset, 1981;

Beblawi, 1990; Messiant, 2001).

There is at the same time a great deal of confusion over what exactly constitutes a CSO.

While some scholars have a highly inclusive catalogue of CSOs, more strict conceptions maintain that in order to count as a CSO, the organization must display some commitment

16 toward democratic governance. This includes things like tolerance, inclusiveness, and dedication toward advancing the public good (Nzomo, 2003, p. 181). Coinciding with this conceptual confusion, the concept of civil society has lost nearly all analytical rigor over the past decade (Nugent, 2004). For example, do religious organizations count as CSOs? In terms of values, religious organizations can be quite conservative, lacking commitments to the values of tolerance and reciprocity, which are seen as central to democratic governance. In terms of autonomy, religious organizations often times survived in Africa by taking non-confrontational stances with governments, or even collaborated with some of the most notoriously heinous regimes. Yet, mainstream churches, such as the Anglican, Presbyterian, and Catholics, “became much more assertive” once political opportunities allowed them to do so (Nugent, 2004). Where is the line between being an agent of the state versus being an autonomous entity able to challenge the state? Does lobbying or acting passively compromise this autonomy?

Since the primary focus of this research concerns the longevity of party dominance rather than democratic governance, concern over the democratic qualities of societal organizations would undoubtedly be misplaced. Instead, what is more important here is the ability and willingness of social organizations to act as causal forces in an arena of political contestation.

Are there societal organizations in positions to take advantage of the new opportunity structures brought on by international pressures and multiparty elections? The answer to this question will be shaped by a number of factors identified by the civil society scholars, namely the level of state repression, the quantity and quality of economic development, and the intensity of societal identity structures. These factors offer up a logical point for shifting into a more detailed discussion of the theoretical framework which guides the overall course of this research.

CHAPTER TWO TOWARD A THEORY OF REGIME TENURE

“Whenever a theory appears to you as the only possible one, take this as a sign that you have neither understood the theory nor the problem which it was intended to solve.”

Karl Popper, 1972

The “agent-structure” problem is found in nearly every social science genre, as academics try time and again to capture the possibilities and limits of human action. David Dessler (1989) defines the essence of the agency-structure problem in the following passage:

This problem is, strictly speaking, a philosophical one. It emerges from two

uncontentious truths about social life: first, that human agency is the only moving

force behind the actions, events, and outcomes of the social world; and the

second, that human agency can be realized only in concrete historical

circumstances that condition the possibilities for action and influence its course

(p. 443).

Agency, or the choices and actions made and taken by individuals, along with structures, or the constraining or constitutive milieu in which action takes place, are social and political phenomena that merit attention when trying to explicate the bases of party dominance. Starting with agency, the choices made and the actions taken by incumbents and oppositionists will have specific implications for their successes of failures in current and future political battles. An opposition party for example, may strategically choose to boycott an election in order to draw attention to electoral malpractice. At the same time, by not participating in an election the party might also weaken its institutional position for fighting future political battles. Similarly, the decision to repress political protests by deploying the police or the military may have the effect of stemming the spread of discontent today, but might also harden opposition resolve in the future.

In the real world, agency always takes place under the influence of structural forces that are generally unalterable at the time of action. In some cases, structures like institutional rules or levels of economic development will shape actor incentives or limit the real options available to actors. Furthermore, structures not only act as constraints, but also constitute the very interests that actors pursue. Shared beliefs and values which accentuate ethnic or religious distinctions for example, will constitute the motivations of political actors and whether or not they oppose or support the incumbent party.

The goal of this chapter is to explain the theoretical framework that guides the interpretation of the empirical data presented in chapters four through twelve. The theoretical perspective offered here is informed by two objections to the limited research scope offered up in the literature on semi-authoritarian rule. First, the semi-authoritarian genre is trying to explain the emergence of a new regime type from a highly static, a historic approach, often failing to account for the casual effects of the exercise of power during earlier periods of hard- authoritarian rule. Secondly, while scholars in the semi-authoritarian genre emphasize the importance of subtle repression as an incumbency maintenance tactic adapted to the watchful eyes of international actors demanding good governance, semi-authoritarian writers fail to systematically examine the most subtle forms of dominance – domination through the construction and deployment of socially salient ideas and symbols, which ultimately reduces the necessity for coercing or manipulating to stay in office. This failure is quite surprising since the semi-authoritarian literature establishes regime typologies along an authoritarian-democratic continuum. Yet the literature fails to account for some of the most important elements sustaining

19 stable contemporary democratic rule – collective beliefs about the efficacy of competitive multiparty elections in the delivery of the public good, the ability of checks and balances and coalition government to check against overly ambitious human nature, and the objectivity of the rule of law. Why not give space for the role of collective beliefs when accounting for softer forms of authoritarian rule?

Toward a Theory of the Micro-Macro Making of Party Dominance

The approach adopted in this work explains the persistence of party dominance in Tanzania by adopting a historical framework that can account for the ways in which ideas, when paired with material capacity, provide opportunities for social action. The first statement undergirding this framework maintains that actors’ preferences are not a given in reality, but are constructed in the process of interacting with other actors. Interest groups, politicians and sometimes corporations engage in advertising campaigns not simply to inform the public about their positions or products, but to try to convince agents into supporting or purchasing by changing the way they think about themselves and the world around them. More specifically, I argue that in order to fully appreciate the politics of dominant party systems, one must account for the ways in which opposition entrepreneurs and incumbents attempt to employ morals, values, and beliefs, what I refer to as ideational structures, as basis for mobilizing people to oppose incumbency or to defend it (Jenkins, 1981; Kaufman, 1997; Polletta & Jasper, 2001).

Ideational structures not only speak to the ability to mobilize a popular following, but also speak to extent that political regimes are able to survive multiparty elections without routinely resorting to election rigging and the services of state security agents. Ignoring for example, the importance that symbols of national unity and national identity might play in sustaining regime

20 tenure is a rejection of a reality visible in everyday life. Interest groups, politicians, and sometimes corporations do not only describe the technical details of platforms and products in advertising campaigns. In fact, this is rarely done. Instead, by advertizing, interest groups, politicians, and corporations try to create perceived value in platforms and products by relating them to values found in collectively held beliefs, such as the virtues of the invisible hand, the just nature of electoral processes, or the social ostracism of being overweight. Where ruling parties are able to maintain loyalties through similar tactics, it translates into reductions in the need to deploy coercion as a means for reproducing tenure.

The second and most basic thesis of this framework maintains that actions in the past, most notably state policies and programs, constitutes a structural milieu which guides present political action. Structures in this sense are derived from uncertainties about the consequences of changes in regime (Pierson, 2004), the growth of an “institutional matrix” that develops in response to the particular nature of the regime (North, 1990), or powerful interests that are vested in the current institutional order (Levi, 1997). A quote from Pierson (2004) neatly sums up the definition of the effects of historically constructed structures; “once established, patterns of political mobilization, the institutional ‘rules of the game,” and even citizens’ basic way of thinking about the political world will often generate self-reinforcing dynamics” (p. 10) which may sustain the regime even after the initial causes that gave rise to it have long since vanished

(Stinchcombe, 1968; Pierson, 2004).

Indeed, there is something to the world around us that tends to stabilize political action over time. Car enthusiasts often remain loyal to particular brands, even when doing so, on the surface, appears to be a violation of rational self-interest. Conservatives generally vote for conservative parties throughout their entire lives, irrespective of their life’s economic

21 circumstances. In Tanzania, peasants continue to support a regime that at one time, forced them from their land, made them sell their crops at below market value, and prevented them from forming or supporting a political alternative.

Commonly referred to as path-dependence (Hacker, 2002; Pierson, 2004), the historical framework presented in this piece argues that developments born out during regime formation in

Tanzania throughout the 1950s and 1960s institutionalized ideological and bureaucratic interests which precluded certain post-colonial policy options, such as development patterns based on private sector cash crop exports or politics based upon collaboration with ethnic associations and chiefs. More specifically, policies pursued under the auspices of African Socialism1, which included commitments to political and economic equality, national cohesion, and state ownership of production, created powerful structural legacies which named a society whereby economic resources, ethnic identities, and popular discontent failed to provide a sufficient basis for popular protests and the mobilization of material and ideation resources by would be political entrepreneurs.

So, the nature of multiparty contention is conceptualized here as arising out of a series of actions that take place over a prolonged period of time. I classify the level of action as micro- level phenomenon, which implicates choices and maneuvers made by individual political agents.

At this micro-level, metaphors such as ‘fluidity of action’ or ‘actor orientation’ (Silverman,

1970, p. 222) might be used to describe political processes. By actors, I refer to political elites or political entrepreneurs. At times, I will deploy the term elites, while at other times I will refer to party factions, ethnic and religious groups, and classes under the assumption that groups have leaders that mobilize group members and engage in political battles with other elites.

1 The term African Socialism will be subject to a more rigorous definition in the pages to come. Suffice it to say here that African socialism, as applied to Tanzania, was akin to a sort of utopian socialism rather than scientific socialism.

In an effort to simplify the analysis, I first apply this micro-level concept to the policy undertakings of political leaders during the era of single-party rule. The question is how do the various political pursuits of regime leaders impact the longevity of regime dominance? More specifically, what was the array of social and economic policies undertaken by political elites during the single-party era? How did these policy pursuits impact the array of existing social groups and their own political orientations?

Secondly, in addition to a micro-level analysis as the method for conceptualizing policy pursuits during single-party rule, I also apply this micro-level concept to actions under the contemporary multiparty framework in Tanzania. The discussion of political action in the multiparty setting, will look quite a bit like resource mobilization theory, which rejects the notion that structural conditions, such as deprivation (Gurr, 1971), are sufficient conditions for the rise of opposition movements. For resource mobilization theory, grievances can be considered only as a sort of precondition, one given causal force by well-placed rational and instrumental actors with the ability and willingness to construct social movements (Jenkins,

1983; Tarrow, 1998). Without well-placed actors, grievances might never surface or when they do, they may take the form of an uncontrollable crowd with no coherent agenda (Canel, 1997).

As Canel (1997) notes, “resource mobilization theory focuses on how groups organize to pursue their ends by mobilizing and managing resources” (p. 146). Hence, questions at the micro-level will refer to the tactics that political actors employ for challenging incumbency or defending it.

In short, at the micro-level, the intensity of the CCM’s dominance in multiparty Tanzania is defined by the referencing incumbent and opposition campaign tactics, including the usage of slogans, symbols, finances, and institutions. At the same time, the nature of the policies enacted and enforced during the single-party era has enormous impact on present interactions between

23 the CCM and the opposition parties. In order to appreciate how policies during the single-party era are important factors for understanding CCM dominance today, we must move on to define the macro-structural-level framework.

The macro-structural level is where patterns or regularities of behavior over time are found

(Silverman, 1970; Zucker, 1988). These patterns are often described as “socialized” or

“ritualized” outgrowth of micro-level actions (Wendt, 1999). However, over time, these patterns are more than just products of micro-level action. As cited earlier, structures at the macro-level in turn provide constraining and constitutive the framework in which future political battles at the micro-level will play out. This, agency and structure are not reducible to one another

(Giddens, 1984; Guzzini & Daase, 2006).

Again, while largely a theory of micro-level action, resource mobilization theory has something important to say about the significance of macro-structures as causal mechanisms for the rise of social movements. First are the changes in opportunity structures, or conditions not directly controlled by political entrepreneur, which open-up new possibilities for action

(Tarrow, 1998). In this piece, shifting opportunity structures includes 1) the transition from the

Cold War to post-Cold War international climate which consequently increased pressures for democratic governance and human rights, and 2) the new opportunities for political party building after multiparty reform. Secondly, resource mobilization theory suggests that the ability of actors to mobilize resources, be it economic, cultural, or symbolic resources, in order to take advantage of new opportunities will ultimately shape the rise of social movements and their institutional carriers, political parties. Hence, “affluence and prosperity” are causal bases for the rise of social movements, giving way to well-financed political entrepreneurs and donors to political efforts (McCarthy and Zald, 1977; Canel, 1997). Likewise, existing social identities,

24 whether religious, ethnic, or class may constitute the necessary political space for developing oppositional sentiments (Polletta & Jasper, 2001). Taken together, social milieus where political entrepreneurs are able to rally together financial and ideational resources are more likely to produce vibrant and sustained political parties versus milieus where material resources are scarce and sectional identities weak.

Figure 2.01: Macro-Micro Framework

Figure 2.01 above defines macro and micro-level phenomena and illustrates the relationship between macro-level structures and micro-level actions. Treating the diagram in the figure as a wheel that rolls clockwise across time (i.e. as a temporal process of reciprocal casual exchanges between micro and macro-level variables) accurately defines the working framework for analyzing party dominance. For the body of this work, the temporal span starts roughly at the time of independence in Tanzania, where the macro-structural environment inherited from colonial rule fundamentally shaped the coalition that led to the rise of TANU and the latitude for the Party to pursue a relatively ambitious transformative agenda. Throughout the era of single-

25 party rule, this transformative agenda implicated particular social and economic policies. These policies, which are described in more detail in the following paragraphs, cultivated, undermined, and reproduced material and ideational features at the macro-structural level that offered few realistic venues for the micro-level efforts by political entrepreneurs to mobilize resources.

Ideational and Material Resources and Macro-Structural Divergence

While the micro and macro discussion thus far largely implicates the historical framework in which politics will be assessed in subsequent chapters, ideas and capabilities speak more directly to the causal mechanisms which give rise to varying degrees of domination and opposition. Let’s start with describing ideas. By ideas, I am referring to collectively held beliefs and values that constitute what are commonly referred to as identities (Friese, 2002; Straub,

2002), including religious, ethnic, language, and class-based collectivities. Of primary importance is not the inward nature of identities per se (i.e. the meaning of membership to individual members), but the intensity of self-other discourses that constitute the boundaries between identities throughout society and that entice group members into social action (Kreckel,

1994, Straub, 2002). I use the term polarization to describe the intensity of the self-other discourses which constitutes the basis for differentiating one group from another.

The crux of the concept of polarization is found in the prefix “polar”: a continuum whereby binary distinctions are made in order to define the community and issue boundaries. As identified by social constructivists and post-modernists, these macro-structural distinctions are generally constituted over time through micro-level ‘self-other’ discourses (van Dijk, 2004).

These distinctions, which in Africa are so often encouraged by ambitious political entrepreneurs, can be ranked according to the intensity of differentiation. Terms like “good” and “evil”

26 generally represent a “lexical polarization” of sorts and are routinely used by incumbents and opposition entrepreneurs of the past and present as a way for framing the nature of political conflict in order to solidify a political base (van Dijk, 2008). By contrast, terms like “greater” and “fewer” or “more” and “less” tend to be not as polarizing since they denote a sliding scale rather than a more threatening world defined by unchanging polar opposites. At the same time, less polarizing terms require more rigorous articulations, contain more visible grey zones, and are generally less efficient at solidifying a sizable and energetic base.

As illustrated below in figure 2.02, the micro-level consequences of polarization can range from harmonious or highly conciliatory actions to acts of hostility and violence. Where high levels of polarization ensue, groups tend to shore up collective boundaries through rigorous and abrasive self-other discourses. In the former Yugoslavia, Albanians regarded Macedonians as one out of many Slavic groups that have historically tried to dominate the Albanian population

(Safran, 2008, 446), creating a sense of urgency and insecurity which led to state collapse and the outbreak of genocide. In Burundi, the racialization of Hutu and Tutsi distinctions during colonial and post-colonial rule, along with the associated efforts to demonize the other, led to a situation where violence became a normal practice of resolving conflict (Lemarchand, 1994).

On the other hand, a polity might be characterized by low degrees of polarization. In such a homogenous setting, self-other discourse may be substituted by a corporate entity where individuals think as “we” bound together by particular beliefs, practices, and norms (Wendt,

1994, p. 390).

The midway condition in figure 2.02 is most commonly appreciated among advocates of political liberalism. Here, distinctions are defined as basic policy differences rather than radically different perspectives on ideology or perceived life and death struggles between

27 sectarian groups. For liberals of all strains, the ideal basis for governance is a process whereby friends with divergent policy interests bargain amid a consensus over the fundamental rules of political competition (Dahl, 1956; Rawls, 1997).

Hence, ideas motivate people into social action and provide a moral framework for defining the nature of political struggles. The theoretical perspective proposed here predicts that, where polarization is strong, the level of popular contention against party dominance will be higher.

Where political entrepreneurs operate within a context whereby tensions between social groupings run high, we can say that ambitious political actors find it much easier to mobilize communities by evoking images of “us” versus “them”, thus strengthening the ideational basis of party development and, in turn, further polarizing the social and political climate.

Figure 2.02: Polarization Conceptualized

At the same time however, the collective ideas that compose identities may be nothing more than constellations of localized rhetoric with little or no sustained causal force in the national

28 competitive political arena. We can imagine conditions whereby social polarization exists, but the possibilities for collective action is far too fragmented to act as a national political force.

Similarly, we can imagine conditions whereby popular grievances exist and are expressed in the form of sporadic popular protests with little or no organizational coherence and sustenance.

Instead, a group’s material capacities give ideas sustained casual force. Therefore, in addition to polarization of ideas, I am also interested in the degree to which identities are able to be paired material resources, including money, expertise, and organizations. This pairing in turn transforms sporadic protests into social movements, and ultimately, into political organizations with the potential for capturing public offices (Zald & McCarthy, 1987; Canel, 1997). In the electoral arena, where collecting votes matters, this framework predicts that the pairing ideas with financial resources gives rise to a social force (Migdal, 2001) with some discernable casual effect on the competitive political arena. In this piece, I refer to the particular combination identity polarization and the availability of material resources as the level of macro-structural divergence.

The level of structural divergence will have several obvious effects on patterns of politics and regime longevity. First, where the structural environment presents political entrepreneurs with lucrative combinations of ideational and material resources, governing coalitions tend to be more tenuous as political entrepreneurs can more readily exercise the exit option (Sartori, 1976;

Lane & Ersson, 1999). One need not look beyond the African Continent to understand the tenuous position of state leaders when faced local power brokers having sufficient control over valuable commodities and the ideational capital for constructing opposition to the regime through appeals to ethnicity.

At the same time, where divergence is low, pursuing more inclusionary coalitions as an effort to build broader party support poses fewer costs and risks of intense intra-regime fighting.

In short, the “dilemma of scale,” which holds that, as party size grows, intra-party competition becomes increasingly stiff and party unity more difficult to achieve, is less acute of a problem in less structurally divergent settings (Rasmussen, 1969; Chikulo, 1988). Similarly, “clientelistic control” (Foweraker & Landman, 2000, p. 15) is more easily centralized and disciplined by the regime which holds the reigns of state power. Figure 2.03 graphically depicts this dilemma of scale hypothesis as applied to clientele relations.

Figure 2.03: Macro-Structural Divergence and Clientelistic Control

Secondly, as implicated above, where structural divergence is high, regimes will tend to govern more exclusively. Distinct subnational identities, when paired with entrepreneurial political actors able to tap into vast financial resources, can suffice as “a powerful ideological

30 instrument for institutionalizing political participation or mobilization” in the competition for political power between elites (Salih, 2001, p. 27). The knee-jerk reaction by many regimes throughout Africa has been to ramp-up the coercive arm of the state as a method for excluding the most threatening societal demands. At the same time, excluded interests might result in a

‘crisis of mass loyalty’ in the long-run, whereby new social actors with popular bases emerge to challenge the status quo (Cohen, 1982; Offe, 1985; Canel, 1997). Likewise, where structural divergence is high, sustenance under multiparty rules will require more frequent resorts to the blatant manipulation and coercion, resorts which have become less acceptable among Western lenders and donors and thus more costly of an exercise for incumbents.

Stepping back from these details, the theoretical statements made in this section can be most appropriately summarized as arguing that greater levels of divergence at the macro-structural level equate to greater levels of ideational and material capital available to political entrepreneurs for organizing contentious action outside researches of the party in power. In

Kenya for example, a relatively well developed business environment, paired with highly polarized ethnic tensions, provided political entrepreneurs in the lead-up to multiparty reform with a lucrative material and ideational basis for organizing political contention into sustained political parties. These macro-structural conditions however, did not just fall from the sky, but were produced and reproduced by the micro-level politics carried out during earlier periods of authoritarian rule. Let us now turn to a brief discussion about micro-level actions.

Micro-level Actions in the Formation of Party Dominance

So far, I have defined material and ideational causal mechanisms as macro-structural phenomena, and only alluded to the micro-level implications of material and ideational

31 distributions. In this section, I deal more directly with micro-level causes of macro-structural phenomenon and how these phenomena in turn, constitute the framework in which future micro- level actions will take place. In other words, this section gives the wheel in figure 2.01 a spin by previewing the variables presented in the empirical chapters.

As a recap, micro-level analysis can be characterized as the level of action. In the first micro-level iteration, I am referring to the actual exercising of power during the single-party rule period. Exercises of power can include the use of state institutions for subduing rivalries in the short-term or the passage of economic and social development policies which deliberately or inadvertently undermine the growth of opposition in the long-term. Borrowing from Chatharine

Newbury’s (1988) “diachronic approach” to the study of ethnicity in Rwanda, I maintain that different policy priorities in the exercise of power by the state (e.g.: micro-level action) will shape macro-structures in ways that minimize the possibilities for organized social contention against the regime (p. 14).

Indeed, the role of the state in reproducing party dominance is widely reflected in much of the literature on one-party dominance (Giliomee & Simkins, 1999). While T. J. Pempel’s 1990 work on “Uncommon Democracies” concerned party dominance in industrial settings, the following passage is relevant here:

The weightiest political consequence of long-term dominance lies in the ability of the dominant party to shape, over time, the nation’s nexus of public policies, its rules of political conflict, and the benefits and burdens imposed on different socioeconomic sectors of the society (p. 334).

At the micro-level, policy priorities can be loosely classified according to varying degrees of exclusion versus inclusion on one hand, and the varying breadth of the regime’s economic scope on the other hand. The exclusion and inclusion continuum here, which relates most directly to the ideational component of macro-structural divergence, is defined as the degree to

32 which a regime seeks to include all members and sections of society in the political and economic affairs of the state. On one hand, regimes might try manage potential rivalries by repressing and excluding political leaders and masses from the formulation of policy, political participation, social services, access to jobs, economic development, and the cultural composition of the nation. As stated earlier, greater levels of exclusion are considered detrimental to party dominance in a context whereby international and domestic actors are pushing for multiparty reform. Simply stated, exclusion might suffice as a strategy whereby authoritarian rule goes unchallenged by international actors and those domestic groups that feel emboldened by international pressures for democratic governance. However, exclusion generally polarizes the social milieu and sows the seeds for discord that cannot be easily negotiated with once conditions begin to push toward more inclusive governance. Kodmani

(2005) for example, convincingly demonstrates that recent repressions and prohibitions of religious extremism in Egypt have helped polarize the religious climate, thus making it more difficult to deal with religious extremism in the future. Likewise, Douma’s (2006) work speaks more broadly to political and economic exclusion, citing that in Niger and Senegal, the uneven impact of state policies, which disproportionately benefit particular ethnic groups, only escalates intra-state conflict. While polarization may give rise to multiparty competition, it might also bring on far worse. As Safran (2008) notes, “denigration and inferiorization, whether by a country’s elite or by society, is often a forerunner of government policy; it may be the beginning of a process that leads to de-legitimation, scapegoating, exclusion, ethnic cleansing, and finally, genocide” (p. 438).

On the other hand, the policy priorities of a regime might also be distinguished by what I refer to as the economic scope of the state. Whereas inclusion and exclusion speaks directly to

33 the level of polarization, the economic scope of the state casual relates more directly to the distribution of resources manifest at the macro-structural level. In some cases, regimes born out of independence movements attempt to consolidate their gains by extending the state’s dominance over economic production by creating a series of government owned corporations in the industrial, mining, commercial, and agricultural sectors. This was certainly the case for

TANU in Tanzania following a set of 1967 policies which embraced state ownership of the commanding heights of the economy. By contrast, in many other cases, regimes consolidate their gains by erecting a series of regulatory marketing boards and national cooperatives, but, out of pragmatic necessity, also allow private actors to own land and industries. Indeed, this was clearly the approach taken by Kenyatta in Kenya, who encouraged a comparatively higher degree of private entrepreneurship when compared to Nyerere’s policy pursuits in Tanzania.

The central casual argument as it relates to the economic scope of the state is as follows:

States that were able to successfully prioritize state economic ownership were also more likely to generate a class of bureaucratic bourgeois2 elites with material capacities far greater than elites in the commercial or petty bourgeoisie sectors. In turn, at the onset of international multiparty reform pressures, regimes tied to powerful bureaucratic classes were far more economically capable of mobilizing resources in defense of incumbency, versus those political entrepreneurs which had to mobilize resources out from commercial and petty capitalist base which was stifled due to years of state ownership and disincentives.

2 I use the term bureaucratic bourgeoisie interchangeably with other terms like state bureaucrats and high level civil servants. Operationally speaking, the bureaucratic bourgeoisie are a wealth accumulating class occupying positions of stately authority at the intersection between international “neo-colonial” interests and domestic sources of surplus production. Because of their bourgeoisie ambitions, this class is cast in opposition to socialism. However, in the case of Tanzania, state bureaucrats allied with the party leadership stratum in their mutual opposition to the rural petty bourgeoisie for the control over peasant loyalties and agricultural production (Shivji, 1976; Parker, 1979; Okoko, 1987).

By contrast, in states where private sector production was allowed to flourish, the relative capacity of the bureaucratic bourgeoisie was far lower, while commercial and/or petty capitalist capacities were far more significant. In turn, at the onset of international multiparty reform pressures, regimes tied to comparative weak bureaucratic classes were far less economically capable of mobilizing resources in defense of incumbency, versus those political entrepreneurs who were able to mobilize resources for party building out from a lucrative commercial and petty capitalist base.

My empirical sections start with chapter four, where I specifically address the exclusive or repressive characteristics of TANU, and the CCM following the 1977 merger between TANU on the Mainland and the ASP on Zanzibar. The goal of chapter four is to demonstrate two important claims. First, explanations for party dominance which singularly reference exclusion and the exercise of coercive muscle as a principle cause for regime tenure do not go very far in

Tanzania. In comparative terms, the exercise of power during single-party and multiparty eras does not look any more repressive than those cases where incumbents were forced out of office following multiparty elections, namely Zambia, Malawi, and Kenya. Along the same lines, tendencies to repress social forces in Tanzania looks considerably less prevalent when compared to many other cases of party dominance, namely Cameroon, Togo, Gabon, and Zimbabwe.

Secondly, in comparative terms, single-party rule in Tanzania was as competitive as many other cases, and certainly more participatory. The assumption here is that competition and participation are metrics for levels of exclusion or inclusion. In single-party systems, such as

Zambia prior to 1991, Kenya prior to 1991, Malawi prior to 1993, and Tanzania prior to 1992, participation was channeled through an array of political institutions, such as women and youth organizations, football clubs, and regular elections. In each case, competition during legislative

35 elections was relatively intense, albeit bound by the institutional walls of a single party. Some notable academic works (Bratton & van de Walle, 1997; van de Walle, 2003) cite that high degrees of political participation should correlate to high degrees of mass protests once new opportunity structures emerge, including opportunities brought on by international sympathies for good governance. The same works cite that high degrees of political competition should also correlate with more sustained opposition challenges. In Tanzania, on the eve and aftermath of multiparty reform, just the opposite happened.

More broadly defined, the documentation in chapter four will be taken as evidence that historical patterns of governance during single-party Tanzania, which emphasized universal political participation, offers a superior basis for reproducing party dominance after the onset of multiparty reforms when compared to party systems characterized by political exclusion. For starters, where levels of political participation are high and universal (e.g. not discriminating based upon occupation, ethnicity, or religion), the exercise of voice acts as a ‘pressure valve’ that enables citizens to express concerns or dissatisfactions with state policies (Bratton & van de

Walle, 1997, p. 142). While the exercise of voice tends to undermine the growth of a highly divergent maximalist opposition feeding off of broader discontent and alienation, as emerged in some of the most oppressive African cases, it also acts as a feedback channel for the regime to measure public opinion, identify possible vulnerabilities, and make policy adjustments accordingly (Kirchheimer, 1966; Pempel, 1990; Giliomee & Simkins, 1999).

Furthermore, having the chance to vote in party elections, to speak to elected officials, to possess a prized party auxiliary membership card, or to dance, sing or score a goal at an entertainment event organized by the party, imparts something more communitarian in nature.

While participation does not necessarily have to lead to policy influence, it can act as a ritual

36 symbolizing a sense of membership to a broader political community. Participation evolves into something one does because it represents membership to an auxiliary organization, political party, or nation, rather than simply because participation is a means to affect the things that public officials decide.

In sharp contrast, where authoritarianism is characterized by stifled political activity, the regime tends to symbolize something alien and resentful in the collective consciousness of its subjects. This alienation and resentment, as derived from periodic efforts to stamp out opponents by quelling rallies, detaining leaders, and manipulating institutions, may constitute a resource to be mobilized by maximalist political entrepreneurs (Francisco, 1995; Rothchild,

1997; Brumberg, 2002). The threat from opposition entrepreneurs is compounded by the fact that once reforms are implemented, contention may emerge rather abruptly as “people learn for the first time that others like themselves have taken to the streets” (Tarrow, 1998, p. 83). One cannot avoid evoking Moi’s heavy fisted rule as a cause for the intensity of the political pressures and party fragmentation that fuelled the growth of the opposition in Kenya. At the most extreme, the increased brutality associated with the state of emergency following the 1996 coup in Niger certainly contributed to an increased militancy between the government and the guerrilla group, the Front Démocratique Révolutionnaire (Sandbrook, 2000, p. 40).

When compared to the long temporal breadth of the discussion of repression in chapter four, the topic of chapter five, which addresses the macro-structural patterns prevailing in the years of independence in the early 1960s, appears to be out of order at first glance. However, since chapter four is largely about dismissing a particular cause, namely the level of repression, as well as locating the comparative nature of governance in Tanzania across time, I decided to place the macro-structural specifics of the independence period in chapter five rather than four.

The causal statements laid down in chapter five are depicted in figure 2.04 below. I argue that the macro-structural environment in Tanzania at the time of independence was comparatively less divergent than the typical African case of the same period. The consequence of this is no small matter for the consolidation and longevity of single-party, and ultimately, party dominance after multiparty rule. As chapter five argues, low degrees of macro-structural divergence, as a product of the peripheral nature of the Tanzanian colonial economy, translated into weak class and ethnic structures, and consequently a weak material and ideational basis for constructing viable opposition challenges to the consolidation of power by TANU party ideologues.

Figure 2.04: Policy Implications of Macro-Structural Divergence

This macro-structural environment had two important impacts on subsequent micro-level politics. First, while TANU did subdue some fringe opponents through the coercive use of force and actively dismantled the local power structures that were cultivated under colonial rule, the party was also able to expand its organizational presence and permit limited degrees of participation and competition without fear that major societal divisions would translate into

38 intra-party rifts. In short, the dilemma of scale was less of an acute problem in the Tanzanian case. Secondly, with minimal centrifugal tendencies within the party and society, TANU leaders had a relatively free hand in maintaining centralized control over clientele networks and implementing a series of policy measures that are often referred to under the rubric of Ujamaa3 or African Socialism.

Moving onto chapter six, I proceed to more systematically nail down patterns of governance under single-party rule, primarily addressing those policy products of the path-breaking African

Socialist proclamation, the 1967 Arusha Declaration. In this chapter, I argue that specific policies implemented as part of Ujamaa produced and reproduce a diminished level of macro- structural divergence to a degree not found the typical African case.

The policies pursued, as well as their long-term consequences are listed below in figure

2.05. Specifically, one of the foremost concerns in the lead-up to the Arusha was the potential growth of commercial and petty bourgeois classes, ethnic sentiments, and wealth inequalities, each of which was understood not only as antithetical to the aspirations of Ujamaa, but as a potential dialectic to the long-term survival of the party ideologues and the rising bureaucratic elites. Therefore, the economic and social policies of the Arusha Declaration specifically sought to stem the growth of petty and commercial bourgeois classes, whose wealth outside the spatial reach of state and party elites could potentially give rise to a resource base by which ambitious political entrepreneurs could mobilize and organize sustained opposition movements.

Similarly, institutions seen as a potential bastion of localized ethnic sentiment were completely dismantled shortly after independence. This included ‘tribal’ societies and the

3 Ujamaa, which is a Kiswahili word for familyhood, is generally equated with Nyerere’s African Socialist formulations and policies. These include democratic single-party rule, economic and political equality, national ownership of the means of production, and national unity. Throughout the remainder of my work, I use Ujamaa and African Socialism interchangeably, but prefer the term Ujamaa since it more strongly denotes the concept of a national community which ultimately carries significant casual weight in this study.

39 institutions of ‘customary’ authority, most notably the institutions of chiefly rule. At the same time, TANU party branches and state organs expanded rapidly to fill what would otherwise have been an institutional void.

Figure 2.05: Multiparty Consequences of Policies during Single-Party Rule Red = material components, Black = ideational components

As depicted in the table above, in chapter six, I deal specifically with the eradication of extra-party material bases when discussing nationalization and rural development policies. The long-term political consequences of state ownership of the sources of surplus generating capital, along with the deconstruction of local power structures are clearly visible when moving forward to the era of multiparty reform. Opposition leaders such Kenneth Matiba in Kenya, Bakili

Muluzi in Malawi, and Frederick Chiluba in Zambia each had a respectable access to personal wealth, class-based organizations, and varying degrees of consolidated ethnic loyalties which opposition leaders at the time of multiparty reform in Tanzania could only dream of. In short, the net effect of undermining economic bases and ethnic loyalties ensured that, over time, blatant coercive action against class-based or ethnic impulses would be less and less necessary and the costs imposed by international reform pressures would be comparatively minimal.

The leadership code, which attempted to curtail the material wealth of the bureaucratic bourgeoisie, also had implications for the ability to mobilize resources at the time of multiparty reform. What the code managed to do was to constrain multi-nodal concentrations of material

40 wealth within the party-state nexus, nodes which could have otherwise been used by defecting entrepreneurial politicians to fund their own party-building ambitions. At the same time, as indicated in table 2.05, the leadership code had broader implications for the level of popular discontent since state and party leaders could not openly flaunt their personal wealth in the face of a largely poor population.

But, Arusha was more than simple a struggle against the rise and reproduction of economic and institutional bases lucrative enough to give material and ideational life to opposition elements. While these policies might be able to account for the opposition’s lack of organizational vigor following political reform, it cannot explain the lack of spontaneous political protests or the fact that so many Tanzanian’s expressed support for single-party rule during the early 1990s. The Arusha Declaration was also about the solidification of a hegemonic superstructure which would provide a sense of familyhood as an ideological footing for economic development, and in the long-run, the survival of the CCM.

I suggest that this superstructure of sorts, which consists of a loosely constructed concept of nationhood and the CCM’s role as its defender, is what makes the Tanzanian case so exceptional. As mentioned earlier, the comparative literature often misses the importance of ideational frameworks for explaining the persistence of party dominance. What is most surprising about this neglect is that ideational variables are perhaps the most common basis sustaining many of the more competitive political party systems. This variable is found not in the more overt exercises of power explained above, but in the more covert role that ideas play in producing and reproducing regime legitimacy and thus, sustained regime tenure. In a nutshell, regimes exercise power over citizens not only by getting them to do what they otherwise would not do by coercing or co-opting them, but also by mobilizing them to do what they believe is the

41 right thing to do for the good of themselves, their community, or nation. As any stable capitalist democracy illustrates, mobilizing a population to defend their revered political order by appealing to the beliefs in liberty, freedom, and the efficacy of the market is perhaps the most economical basis for power reproduction visible today, one that would have made Machiavelli quite proud.

More specifically to the point of chapter six, I am concerned with the extent to which the dominant regime historically engaged in the process of ‘nation-building’ at an ideational stage.

Three general policy arenas will be assessed as the potential constitutive and causative bases for national identity. First, since collective identities are “communicative constructs” (Straub, 2002, p. 72) that emerge out of discourse about the self and the surrounding world, the presence of a national language is understood here as vital for the economy of identity construction. Of course, most countries have ‘official’ national languages. However, the national languages in

Africa commonly suffer from a lack of diffusion into the daily lives of the peasants, and in some cases, from affiliations with particular sections of the population. In both cases, the national language may come to symbolize something remote and alienating among certain peoples. On the other hand, where there is a broader diffusion of a non-sectarian language, nation-building may come to symbolize something more universal and less threatening, to the point that the language itself becomes a valued symbol of that collective identity (Askew, 2002). In short, regimes that push for the collective usage of a less alienating and more universally spoken language, such as Kiswahili in Tanzania, may ultimately stitch together a stronger degree of convergence rather than divergence. This convergence constitutes a shared macro-structural collective identity that is more resistant to micro-level attempts by political entrepreneurs to mobilize subnational identities.

A second policy arena more directly addresses ideological constructs and will be dealt with mostly within the sub-section on “education for self-reliance.” The crucial point here holds that regimes that construct and propagate relatively “compelling ideologies” are more likely to

“secure long-term incumbency” compared to regimes having little or no guiding ideals or principles (Du Toit, 1999, p. 215). The causal relationship between ideologies and tenure is due to the operation of ideologies as powerful forces for mobilizing a population in support of state policies or in defense of the regime against threats from opponents, which are often portrayed as subversive elements conspiring with the nation’s enemies (Fainsod, 1963; Linz, 2000). As a consequence of the ability to mobilize populations based upon seemingly instinctive commitments to collective ideals, dominant parties can defend their tenure without the frequent resort to repression, which may go to sow the seeds of discord and undermine relations with international donors. In Taiwan for example, the KMT was highly successful in “developing national self-consciousness” and engaging in a “radical” social transformation that diluted

“ethnic tensions”, leaving “little room for an opposition, which can easily be branded as destructive of the efforts at transformation” (Giliomee & Simkins, 1999, p. 26). Likewise, the four-decade long tenure of the Social Democrats in Sweden was reproduced by “socializing subsequent generations of Swedish voters so that much of the once controversial agenda gradually took the air of common sense and the Swedish middle way” (Pempel, 1990, p. 18).

While at times, the rhetoric radiating from the CCM’s more youthful members and radical socialists may remind one of the Chinese Cultural Revolution, ideology here refers more specifically to the somewhat nebulous concepts of patriotism, nationalism, social justice, security and order. Conceptual vagueness itself is often a pragmatic reflection of the presence of a diverse array of powerful interests that would otherwise feel threatened by ideological rigidity

(Linz, 2000). Given this sense of pragmatism, loose constructions may have a greater likelihood of catching and mobilizing a broader section of the population and thus minimizing highly coercive actions that might alienate certain social forces.

The third policy arena relevant to the development of a popular national consciousness, to be discussed in more detail under the sections on “rural development” and “the leadership code” in chapter six, has more to do with the material bases for grievances which raise the probability of finding enough heated discontent for mobilizing political action. Here, I am specifically referring to reduction of discontent through policies that weaken of developmental asymmetries, especially those which correspond to religious or ethnic identity.

As for the leadership code, the inability of the bureaucratic bourgeoisie to amass visible levels of wealth also meant that popular resentment against the regime was somewhat blunted.

Indeed, Tanzanian officials, most notably Nyerere, had considerably more modest lifestyles than their counterparts in neighboring countries. Those officials that did circumvent the leadership code typically did not flaunt their wealth, since it could have denied them reelection or reappointment.

In relation to rural development, I have already mentioned its impacts on the growth of material capacity. As figure 2.05 indicates, the distributive aspects of rural development policies will also have implications for the level of popular contention. The distribution of development resources includes broad development projects, such as schools, health clinics, power generation, and roads. I argue that the symmetries in distributive policies impact macro- structural divergence directly at the level of mass society. Where resource distributions are more universal or inclusive, lower degrees of macro-structural divergence along wealth lines will ensue. For example, the cradle-to-grave welfare state model throughout the Scandinavian states

44 tended to blunt the radicalization of opposition movements and helped secure the prolonged dominance of social democratic parties (Esping-Andersen, 1985; Pempel, 1990). Narrowly confined resource distributions on the other hand, which reflects the tendency to “expel” rather than “accommodate” potential divergent interests, are more likely to sow the seeds of regime demise by creating a empirically verifiable references for mobilizing popular, sometimes uncontrollable contention against the regime (Bratton & van de Walle, 1997; Taylor, 1999;

Salih, 2001). In Kenya, Moi rewarded his ethnic allies in the Rift Valley and Western, and

Coastal Provinces, at the expense of the Kikuyu in the Central Province, which were the prime beneficiaries of patronage from the former President Jomo Kenyatta (Throup & Hornsby, 1998, p. 27). These rewards reinforced patterns of highly divisive ethnic/regional political and economic competition in Kenya and provided rich soil for mobilizing popular discontent against the Moi regime (Rothchild, 1997).

The chapters that follow chapter six will empirically assess the macro-structural consequences of the aforementioned policies, as well as the degree to which macro-structural legacies shape the micro-level politics in the context of multiparty reform. Chapter seven will describe how the comparatively low degrees of macro-structural divergence in Tanzania during the closing of the 1980s, positively interfaced with rising international demands for multiparty reform. The nature of this interface gave the CCM the latitude to reform, without risking a serious opposition challenge in subsequent multiparty elections.

Chapter eight will delve specifically into the empirical question of the macro-structural legacies which undermined the ability of political entrepreneurs to mobilize material resources in their quests to contest against the CCM in multiparty elections. Policies of nationalization and rural development ensured that the party-state nexus remained, by far the most economically

45 capable node of economic and political activity, thus offering clients a material justification for maintaining political loyalty and ensuring disloyal elements found no viable alternative material base.

Chapters nine through eleven speak more clearly, although not solely, to the ideational consequences of delegitimating ethnic loyalties, accommodating religious ones, and ensuring a relatively egalitarian distribution of development resources. Finally, chapter twelve will embark on an empirical assessment of the macro-structural and micro-structural consequences of the

CCM’s ambitious attempt to create a sense of national identity and placing the party as a centerpiece of that identity. But, before moving straight into the empirical sections, a few pages of ink are needed for clarifying with precision some critical methodological issues. As the following chapter shows, connecting theoretical knowledge about the causes of party dominance with empirical facts about actions and structures is most appropriately done through a comparative historical method.

CHAPTER THREE RESEARCH METHODOLOGY

“The time-honored empirical criterion for a satisfactory theory was agreement with the observed facts. Our empirical criterion for a series of theories is that it should produce new facts. The idea of growth and the concept of empirical character are soldered into one.”

Imre Lakatos, 1970

The linking of theory and fact, and the extent to which theory and fact provides intellectual insight, are shaped by one’s perspective on the philosophy of science. More pointedly, research methods are specified by epistemological definitions on the nature and scope of existential, correlational, and explanatory knowledge, as well as ontological definitions about the scope of the explanation accounted for in theoretical constructs (Dessler, 1991). As this chapter pursues a methodological sketch of my work, I will first start by briefly defining the philosophy of science that undergirds my research methods, prior to engaging in a more technical discussion of qualitative research techniques, comparative historical methods, and the selection of comparative cases.

The epistemological and ontological perspective offered here is that of a scientific realist. As a philosophy of science, scientific realism can be defined according to three fundamental characteristics. First, while sharing the view of the empiricist philosophy of science by maintaining that a world exists external to observation, scientific realism contrasts strongly to the post-positivist rejection of the distinction between observation and observer. For scientific realists, the separation between subject and object translates into a project of scientific inquiry which explains events that are external to cognitive representations of the world or distinct from the language structures which shape how observations are interpreted. In comparative terms, whereas empiricism treats observed phenomena as “clusters of sense data” (Hellman, 1983, p.

233) and the post-modern epistemology conceptualizes reality as structured by discourses of sameness and difference (Wendt, 1999), scientific realists generally maintain that the existence of real world phenomena does not logically nor causally incite the necessity of an observer

(Hellman, 1983, p. 233). In short, atoms, kitchen tables, or class and ethnic consciousness exist as products of real causal relations or constitutive properties that exist whether or not we observe their existence as researcher.

Secondly, scientific realists define theoretical constructions of unobserved processes and

“hidden structures” as central to developing an explanatory framework for a particular phenomenon (Hellman, 1983; Warren, 1989). This position is distinct to the philosophy obtained by empiricism, which concludes that unobservable phenomena are simply unknowable and therefore, the subject of nothing more than baseless speculation. Guided by the belief that true knowledge can only come from what we observe, empiricists deduce explanations from the logical relationships between general laws and observations and incorporate unobservables as descriptions of associations between variables. This so called deductive-nomological (D-N) model has proven to be a powerful tool for translating observed existential knowledge into knowledge on correlations between variables (Dessler, 1991).

For the scientific realist however, the task of science is to move into a terrain where explanations for correlations are established, something the D-N model and correlational knowledge alone cannot provide (Dessler, 1989). Establishing explanatory knowledge begins when the scientific realist “baptizes an unobservable phenomenon by proposing a description of its properties and some hypotheses about how these relate to observable effects” (Wendt, 1999, p. 63). Unobservables such as states, political regimes, and social identities are definable

48 theoretical constructs with causal and constitutive consequences. While the properties of constructs and their consequences are shaped by the behavior of individuals as typically described by empiricists, they are not reducible to micro-level actions since constructs provide a field in which individual action takes place, a position consistent with the theory discussion in the previous chapter.

Thirdly, scientific realists generally embrace the form of “sophisticated falsificationism” argued by Lakatos. In contrast to Popper’s view on falsification, where disconfirming empirical observations overturn scientific laws, Lakatos contended that falsification often takes place when existing theories are made more comprehensive by the matching of new observations with new conceptualizations (Cook & Campbell, 1979). A progressive research program would thus be defined as a theory being able to account for the phenomenon explained by an existing framework, while additionally accounting for a few previously unexplained phenomena

(Dessler, 1989). For example, while Newton’s inverse square law of gravitation was falsified by

Einstein’s theory on general relativity because the latter could better account for the gravitational bending of space and time, Newton’s conceptualization was merely incomplete rather than wrong. Similarly, citing coercion as a modality for the reproduction of tenure is not necessarily wrong when applied to the Tanzanian case, but merely incomplete.

Data Analysis: From ‘Discrete Causality’ to Causal Process

Measurements made during a solar eclipse in 1919 on the African Island of Príncipe by the

English astrophysicist Arthur Eddington, empirically established that the mass of the Sun altered the trajectory of light. Indeed, this experiment was an achievement of monumental significance.

Prior to this, Einstein’s theory of General Relativity predicted that gravity wells produced by

49 massive objects should alter the course of light that passes through its contours. Without

Eddington’s measurements however, or ones similar to it, Einstein’s theory would have been little more than a series of abstract field equations without any substance for supplanting the classical mechanics of the Newtonian world.

Whether one theorizes about the possible consequences of gravity for the path traversed by light, antiretroviral drugs on HIV, or previous economic and social policies on the length of regime tenure, theoretical claims must be linked to data of some qualitative or quantitative value in order to function as a mode of scientific enquiry. In the realm of political studies, the most frequent method appearing in graduate courses and academic journals is quantitative methods rather than qualitative ones (Bennett et. al., 2003). This favoritism in part reflects the fact that quantitative data has some clear advantages over qualitative data, such as easily comparable and orderable values and a greater potential for research transparency. Additionally, when paired with sophisticated statistical software, quantitative techniques are sometimes able to establish robust correlations between variables at high levels of confidence, results of which are almost never attainable through qualitative methods alone.

The presentation of tables in the empirical chapters, along with the collection of extensive election data in the appendices, testifies to my predilection to use quantitative techniques when numerical data is available and quantification is of value for understanding casual relations. To be sure, the dependent variable (level of party dominance) is measured through data from election results and public opinion surveys. At the same time, for macro-structural divergence, I rely on a host of economic indicators and election and survey data for explicating measurements on access to material resources, polarization intensities, the saliency of sub-national loyalties

50 versus national ones, and the expressed reasons for popular support for the CCM over the opposition.

At the same time, when bundled together from the first to final pages, the dissertation is largely a qualitative one. The description of the process from dominance at independence to dominance following reform is almost exclusively qualitative. There are no composite indicators for material capacity or polarization. Therefore, while the dependent variable can be compared cross-sectionally based on quantitative data on election results, independent variables will be compared through nouns and adjectives rather than numbers.

Qualitative work is, with some truth, often associated with small n case study methods.

Case studies, or the study of specific “phenomenon of scientific interest” with the aim of generalizing causal statements to other similar cases (Mahoney & Rueschemeyer, 2003, p. 17), have some distinct advantages over large n statistical studies. First, because larger global studies try to maximize the number of cases, there is often a significant degree of conceptual overstretching, at the expense of conceptual refinement. Hence, smaller case studies have the advantage of more robust concept validity.

Secondly, qualitative case studies are useful for exploring questions of “complex causality”, especially phenomenon such as feedback loops, path dependencies, and, most notably,

“sequential interactions between individual agents and social structures” (Mahoney &

Rueschemeyer, 2003, p. 13). As Marshall (1984) points out, qualitative case studies are much more apt for exploring “meaning, to delve in-depth into complexities and process” and to navigate research questions characterized by unidentified “relevant variables and interconnections” (Mahoney & Rueschemeyer, 2003, p. 26). As the theoretical sketch outlined in the previous chapter might suggest, the presence of interacting or “mutually reciprocating”

51 variables (Cook & Campbell, 1979, p. 17) and “combinatorial explanations” (Ragin, 1987, p.

27), renders qualitative techniques as the most efficacious route to explaining the case of continued party dominance in Tanzania after the onset of political reforms.

Finally, an equally important justification for qualitative case studies is linked to the particular class of knowledge my work attempts to establish. Without a doubt, quantitative techniques are extremely useful for translating existential data into correlational knowledge.

Regressing numbers on unemployment and inflation leaves economists highly confident that there is a significant inverse relationship between these two variables. Similar techniques have permitted political scientists to establish a clear correlation between economic performance and political survival in office. Correlational studies are crucial for developing scientific insights to the extent that our knowledge of the world would be far less than what it is today without quantification and statisticians.

However, when correlational statements are equated to casual ones, the cause and effect is merely assumed while the interconnectedness between variables is never established. This reliance on “discrete causality” by quantification and correlation leads to a stunted and possibly delusional understanding of how the world works in reality. Uncovering genuine causal connections are not manageable through quantification alone. Instead, explanations for causal relations come from qualitative descriptions of the process of getting from the independent variable to the dependent one. Due to catalogued observations, physicists know that gravity distorts space and time and hence, alters the path traveled by light. But, the covert and somewhat allusive reasons behind this distortion have since been explored at the theoretic and microscopic levels through quantum and string theories. In other words, quantitative measurements of the ‘gravitational lensing’ of light and the presence of massive celestial objects

52 says nothing about underlying casual relations between gravity and the path traveled by electromagnetic energy. In this piece, in order to explicate underlying causal bases of party dominance, we will need to dig deeper than commitments to quantitative analysis allows us to.

One widely cited complication in qualitative work arises when researchers try to assign precise values to variables. Perhaps the most promising remedy to the problem of value assignment in qualitative analysis is the application of the comparative method. With the comparative method, causes and outcomes are compared with the intent to “develop, test, and refine causal, explanatory hypotheses about events and structures integral to macro-units…”

(Skocpol, 1979, p. ). Loosely speaking, the comparative method employs the Mills Method of similarity and difference to study “how different conditions or causes fit together in one setting and contrast that with how they fit together in another setting” (Ragin, 1987, p. 13; Peters,

1998). This method is especially useful for eliminating hypothesized causes which are shared among cases with divergent outcomes (Mahoney, 2000). Hence, because rules governing elections and divisions of legislative and executive authority are similar across cases like Kenya,

Zambia, Malawi, and Tanzania, institutional design arguments are not truly sufficient for explaining the persistence in incumbency. Likewise, while Malawi, Zambia, Kenya, and

Tanzania have similar records on respect for civil liberties and political rights, it makes little sense to couch an explanation for the CCM’s survival in Tanzania by referencing the frequency of repression and electoral manipulation per se.

A more problematic proposition is eliminating hypothesized causes when they are not shared among cases with similar outcomes. As a consequence, under a strict adherence to the determinism of the Mills Method (Sobel, 1995; Mahoney, 2000), we would be forced to conclude that because egalitarian distributive policies in Tanzania were not present in Botswana,

53 egalitarian distributive policies are therefore not necessary for generating long-standing incumbency. Yet, they may have been necessary in the Tanzanian case. In statistical terms, we run the risk of erroneously rejecting the null hypothesis that distributive policies are important for regime tenure.

The reality of social science is that phenomena do not often lend well to the positivist demands for necessary and sufficient conditions as a basis for concluding the existence of causal relationships (Cook & Campbell, 1979). More commonly, social scientists are faced with combinatorial explanations, where two independent variables X1 and X2 or X3 and X4 combine to yield a particular outcome Y. None of the independent variables alone are sufficient to cause the outcome Y, nor does the outcome necessitate the presence of any one particular independent variable (Ragin, 1987). Instead, concomitant variation, or probabilistic statements might offer a more realistic understanding, given the absence of laboratories and limited sample sizes often faced by social scientists when investigating the complexities of the social world.

An added complication is that a particular causal variable might only have a hypothesized effect in conjunction with an antecedent variable, constituting a temporal series of causally linked variables that can only be captured through a historical analysis of combinations of variables (Mahoney, 2000; Pierson, 2004). For example, to describe the basis for someone’s vote by citing their level of exposure to campaign advertisements misses the truly meaty part of the causal processes. Campaign imageries only have their specific effects in conjunction with values and beliefs that were socialized in the distant past. Similarly, explaining a jury’s verdict in a particular case can only be understood in reference to the arguments made by the attorneys, which conjure up a rationale based on previously constructed shared beliefs and values along with prior court decisions.

In conjunction with the generative exchanges between micro and macro-level variables discussed in the previous chapter, the implication of temporality seems to call for a comparative historical approach. To be sure, evoking temporal explanations is not just an academic exercise, but a sort of common sense approach to the way we piece together the world around us. As

Philip Abrams (1982) so plainly stated when introducing the first chapter of his work on

“historical sociology”:

Try asking serious questions about the contemporary world and see if you can do without historical answers. Whether it is a matter of conflict in the Middle East or in Northern Ireland, or racism in urban ghettoes, of poverty and social problems on the Clyde or the Tyne, or the fall of governments in Italy or Chile, we tend to assume that an adequate answer, one that satisfactorily explains whatever it is that puzzles us, will be one that is couched in historical terms (p. 1).

Much of what we know today about democracy and authoritarianism, such as basic concepts and typologies and causal variables giving rising to governance and regime sustenance, is built out from the comparative historical method (Mahoney, 2000, p. 164). While the importance of history has been eclipsed by the elevated respect for rational choice approaches, scholars from a broad range of social science fields are calling for a “return to history”

(Mahoney & Rueschemeyer, 2003, p. 8). The method adopted here heeds to this calling.

The crux of the adopted comparative method maintains that actions in the past, most notably macro-structural configurations and subsequent micro-level politics, provide structural frameworks for understanding current political phenomenon. Similarly, present actions have

“downstream” consequences for future phenomenon (Pierson, 2004).

The dependence of the present on the past is captured in concept of path dependence and is specifically implicated in processes of positive feedback or self-reinforcement. Path dependence emerges when the initial choices and actions of political actors become “locked in” with the

55 passage of time due to actors’ perceptions of costs and uncertainties associated with course changes, the exercise of power by interests that were vested by earlier choices, and the development of norms and expectation that are positively linked to the legitimacy of the status quo (Krasner, 1989; Alexander, 2001; Pierson, 2004). For example, the initial causes behind regime consolidation, such as the rapid rise of opposition to colonial rule, may be absent in the decades following independence. Yet, the most durable regimes seem to survive as if the initial causes were present. What explains this continuity? I argue that the comparatively underdeveloped nature of Tanganyika at the time of independence allowed the consolidation of power by ideologues and civil servants committed to policies which had the consequences of undermining the material base for independent political action and weakened the saliency of sub-national identities, which in turn, undermined the mobilizational capacities of political entrepreneurs at the time of multiparty reform. None of the three contrasting cases discussed below exhibit a process similar to the Tanzanian case. In each case, either out of necessity or desire, post-independence leaders embraced a development path that would accentuate ethnic or regional divisions and cultivate the growth of class structures which served as the ideational and material bases for mobilizing opposition to single-party rule.

In light of this comparative method, the flow of analysis throughout this piece will compare and contrast historical political developments among a handful of cases, some African and a few non-African. Three of these cases will be introduced in the empirical chapters alongside the

Tanzanian case, while the remaining cases will be reserved for the “comparative perspective” chapter. Let us now look more specifically at the cases selected for comparative analysis and the rationale for selecting them.

Case Selection: Finding Robust Comparisons

In order to establish reasonable casual inferences, the selection of cases in any small n study should be attentive to “maximizing experimental variance, minimizing error variance, and controlling extraneous variance” (Peters, 1998, p. 30), while enveloping these selections within a sound theoretical framework. As a recap from chapter two, I am interested in explaining why some regimes are able to dominate their respective polities for prolonged periods of time under multiparty elections, while maintaining relatively quality records on civil liberties and political rights. Therefore, in terms of the dependent variable, I am concerned with three facts: 1) the regime has maintained a solid control over executive and legislative authority for more than a decade, 2) the past three elections have been conducted under multiparty elections, and 3) the regime has a relatively good comparative standing in terms of civil liberties and political rights.

To this end, my primary case study is Tanzania. The Chama cha Mapinduzi has governed for well over forty years while the last three elections have been multiparty ones. Furthermore, in comparison to many other African cases, the CCM scores relatively well on civil liberties and political rights, a point taken up in the subsequent chapter.

Moving forward with Tanzania as the primary case and informed by the framework adopted in chapter two, the task now is to restate the main independent variables as they relate to

Tanzania in order to find a few viable comparative cases. At the macro-structural level, a cursory overview suggests that divergence has been quite low historically. Due to the

Tanganyika’s incorporation as a peripheral component of British colonial interests, the growth of the division of labor, the consolidation of land, and vertical differentiation along ethnic lines were quite minimal when compared to most other African cases. At the micro-level, another cursory overview suggests that single-party rule was characterized by an economically

57 expansive public sector, and policies which attempted to avoid excluding major sections of society from the political and economic affairs of the state. Development efforts, as much as they failed to generate sustained macro-economic growth, were never directed toward advancing the interests of any particular section of the population. Similarly, political participation was encouraged throughout the entire Mainland polity. Only on Zanzibar do we find a history of repression and exclusion similar to that found elsewhere on the Continent.

Is there a secondary case where similar patterns of divergence and politics prevailed at the time of independence? To some extent, Botswana shares a few similarities with the Tanzanian case. While the textures of politics at the time of independence, and the policies pursued afterwards significantly departed from the Tanzanian model, two fundamental conditions were similar. First, the Botswana Democratic Party (BDP) came to power amid a macro-structural environment easily classified as lacking significant levels of divergence. Secondly, while post- independence development policies were fundamentally inegalitarian and political participation less vibrant than in Tanzania, a strong degree of elite cohesion was achieved because no particular ethnic section of the country was excluded from the policy-making arena and the distribution of development benefits. As will be forcefully argued in the comparative perspective chapter, these two common characteristics are crucial casual bases for party dominance in both countries.

In an attempt to maximize experimental variance, I select three cases where patterns of polarization and exclusion have been far more normal: Kenya under the Kenya African National

Union (KANU), Zambia under the United National Independence Party (UNIP), and Malawi under the Malawi Congress Party (MCP). In each case, regimes came to power at the time of independence amid moderate to strong degrees of divergence, especially along ethnic lines,

58 making independence environment far more competitive than that found in Tanganyika during the same era. Furthermore, this level of divergence enticed the post-independence leaders in each case to shore up hegemony through policies that were fundamentally more exclusive than the Tanzanian case. In all three cases, attention to equality at the elite and mass levels was far less prevalent, while sections of society were permitted to accumulate lucrative levels of economic wealth. Moreover, political exclusion in elections and the definition of the nation clearly cut across regional and ethnic lines. As a consequence, the divergence of the independence era carried over into the era of multiparty reform, giving rise to an environment whereby entrepreneurial political actors were materially equipped to channel sub-national identities into sustained political parties. More to the point, Malawi, Zambia, and Kenya each in their own way, highlight the long-term contradictions in trying to reconcile ethnified political and class developments, with international and domestic pressures for reform and the prospects of continued dominance following reform. Table 3.01 provides a schematic view of my case selection.

There are two additional notes with respect to case selections. First, what is the universe of cases in which these arguments apply? The answer to this question will be explicated more specifically through a comparative discussion in part three. Suffice it to say that the answer to this question is a complicated one, simply because explanations for incumbency in other countries implicate a different context. Nevertheless, there are some generalizations worth pursuing. Specifically, to what extent is the degree of inclusionary governance tied to low levels of polarization and, thus party dominance? Malaysia under UNMO, Mexico under PRI, Taiwan under the KMT, and Singapore under PAP offer up interesting comparisons. The conclusion clearly suggests that minimizing divergence is essential to party dominance, and the best route

59 to achieving this is through politics which seeks to include divergent social forces in political and economic life and development. Policies which unify large sections of the population around national or revolutionary symbols and share power and wealth among all prominent elites are more likely generate a macro-structural context conducive to prolonged party dominance after the onset of reform pressures.

Table 3.01: Case Comparison Schematic Dep Variable Independent Variable Party Dominance under Macro‐Level Divergence Micro‐Level Politics multiparty elections Expansive public Exclusion sector Tanzania yes weak High Low Botswana yes weak low low/moderate Malawi no moderate high Moderate Kenya no strong moderate High Zambia no moderate moderate Moderate

Secondly, the distinction between governance on Mainland Tanzania and on Zanzibar is frequently highlighted throughout my work. More specifically, nearly everything that is said about Tanzania as a whole (i.e. weak divergence and policies which seek to undermine sub- national loyalties) is not reflected in the history of politics on the Isles. Therefore, I will, in the concluding chapter, separate the two by analyzing divergence and governance in relation to the radically different political scenarios in both parts of the United Republic today.

Before moving in to the depths of the empirical discussion, let me clarify what is meant by

“relatively favorable human rights standings,” specifically in relation to civil liberties and political rights. Indeed, given the periodic use of security forces against CCM opponents on

Zanzibar, I expect this assertion to receive some criticism. One of my goals for the following chapter is to provide some clarity to this assertion and deflect some of the criticism.

CHAPTER FOUR REJECTING DISSENT - REPRESSION IN TANZANIA

“The seed of revolution is repression.”

Woodrow Wilson, unknown

"We are informed that if repression has indeed been the fundamental link between power, knowledge, and sexuality since the classical age, it stands to reason that we will not be able to free ourselves from it except at a considerable cost: nothing less than a transgression of laws, a lifting of prohibitions, an irruption of speech, a reinstating of pleasure within reality, and a whole new economy in the mechanisms of power will be required."

Michel Foucault, 1977

Suppression of some form is a characteristic common to every institutional order in every society throughout the globe. Governments, corporations, interest groups, and schools all engage in practices that not only limit the expression of interests and ideas threatening to the status quo, but shape our basic understandings of appropriate behavior, who our enemies and friends are, and the very context in which social action takes place. Reflecting over this, another quotation of Michel Foucault comes to mind: “In its function, the power to punish is not essentially different from that of curing or educating” (1995, p. 300).

The conclusions found throughout much of the scholarly work in comparative political studies today causally link the sustenance of regime tenure under authoritarian rule to the persistence of illiberal forms of suppression, which I think is more appropriately captured in the term repression. Repression in this context includes the most naked use of security forces to harass, detain, or even murder opposition leaders, or to beat, teargas, or shoot protesters.

Repression also includes less overt tactics, such as the manipulation of elections or the revocation of business licenses to supporters of opposition parties. Repressive modes of domination are widely practiced among Africa’s range of political regimes, even after the emergence of the so called “third wave.” As noted by Mkandawire

(1999), authoritarianism in the past “established a certain style of policymaking that has left a legacy which complicated democratic policymaking in Africa,” including centralized executive authority, a “cult of secrecy” surrounding the policymaking process, and the disproportionately strong role played by international actors in shaping domestic politics (p. 125).

The goal of this chapter is to refine the earlier statements claiming that in Tanzania, the level of repression has historically not been so fundamentally different than levels found in most other

African cases. During more recent periods, the level of repression has arguably been less prevalent. While dissent has often times been punished, especially under single-party rule, and while election manipulation and harassment do indeed take place, general trends suggest that there is nothing special about the Tanzanian experience to citing repression as a more significant role in political domination when compared to most other African cases.

The basic roadmap for this chapter starts with a brief overview of the arguments linking repression with dominance, before moving on to examine the Tanzanian case both specifically and comparatively. The data clearly demonstrates that, while repression as described above was certainly a foundational part of the Tanzanian political experience, past and present repression trends in Tanzania look no more outstanding than most other African countries. In fact, the data suggests that, while the percentage of votes cast for CCM’s has increased over the past three elections, the government’s record on human rights has improved at a rate which outpaces many of the records of neighboring regimes.

Repression and Dominant Parties

Repression, like all forms of suppression, implicates an attempt by incumbents to reduce the uncertainties of political conflict by reducing competition and participation (Park, 1991; Bratton

& van de Walle, 1997; Linz, 2000). For repression under semi-authoritarian rule, incumbents hope to placate international and domestic demands by organizing periodic elections, yet continue to hedge against the risks of uncertain election outcomes by suppressing civil liberties and political rights, intimidating and disenfranchising voters, buying votes, harassing opposition supporters and leaders, and suppressing the emergence of dissent within the media (Karl, 1995;

Diamond, 2002; Schelder, 2002).

While the crux of this chapter suggests that, since there are no clear historical and contemporary differences in the level of repression among the African cases selected for this piece, it is still worth pointing to some of the possible effects that repression has on political contestation. By far, the most widely suggested consequence is a stunted opposition. In the short-term, flexing the coercive muscle of the state simply raises the costs of expressing voice against a regime which holds the reigns, bridle, and bit of the military and police services

(Booth & Richard, 1995). The murders and kidnappings that typified military regimes in Latin

America during the 1970s and 80s (González, 1995), the “liquidation” of the private traders, old intelligentsia, and kulaks during the Great Purge in the former Soviet Union (Fainsod, 1963), not only all attempted to physically eradicate existing dissent, but also demonstrated to other would-be political opponents that dissent could be a prohibitively costly and unwise course of action.

In the long-term, repression is widely cited as leading to an atmosphere of pessimism, where politics is seen throughout the population as a “stale, corrupt, elite-dominated domain” having

63 nothing of value to offer its citizens (Carothers, 2002, p. 10). By contrast, authoritarian regimes which permit some limited degree of pluralism and participation will foster norms of political participation and nodes of competition by which anti-regime sentiments can be organized and sustained as viable competitors in a multiparty environment (Bratton & van De Walle, 1997;

Geddes, 1999).

However, other research suggests that, under some conditions, repression might intensify overt conflict, and by extension, render more forceful challenges to incumbency once the international climate begins to favor multiparty elections. The “predator-prey” model of social conflict, which suggests that increases in repression cause increases in protests (Tsebelis &

Sprague, 1989; Francisco, 1995), has been widely supported by empirical data from Europe and the Third World (Lee, Maline, and Moore, 2000). Chehabi & Linz’s 1998 publication on

Sultanistic regimes clearly illustrates that high degrees of repression and widespread political and economic exclusion give rise to stronger maximalist oppositions. The Rational Actor model for domestic conflict, proposed by Mark Lichbach (1987), illustrates that increases in nonviolent government repression reduces nonviolent dissent, while at the same time, increases violent dissent. In fact, Lichbach concludes that the relationship between repression and dissent depends a great deal on the government’s simultaneous efforts at “accommodating” dissent.

My position is that repression and exclusion, when combined as a mode of domination, only polarizes social forces to the point where popular protests against incumbency surface once international pressures offer up a new array of opportunity structures. This position is in stark contrast to the thesis that lower degrees of repression (i.e. higher degrees of participation and competition) under single-party regimes offer the most viable route for the rise of protests and political competition in the lead-up to, and aftermath of multiparty reform. As this chapter

64 shows, evidence demonstrates quite clearly that repression in Tanzania has been on par with other African cases. The central argument is, while histories of repression are characteristic of all cases, in Tanzania, where repression was accompanied by an expansive public sector paired with policies which emphasized political and economic equality and a universal national identity, the macro-structural environment at time of reform was not conducive to rise of opposition parties. Thus, where opposition is weak, the CCM can dominate while achieving respectable human rights record today.

Batons, Guns, and Prisons: Repression under Single-Party Rule

In Africa, the state was developed as the institutional apparatus for coercively enforcing colonial authority. Given the development of a centralized and repressive constellation of bureaucratic institutions under colonial rule, the post-colonial state was to continue to assume highly repressive forms of domination over societies that, for one reason or another, tended to resist domination from state institutions (Alavi, 1972; Saul, 1974; Migdal, 1988).

The historical patterns of governance in Tanzania are replete with instances of the most naked forms of repression. As the brutal response to the Bushiri uprising of the late 1880s and

Maji Maji rebellion in the first decade of the 1900s illustrate, the highly militant German colonial rule of Deutsch Ostafrika had no tolerance for organized political dissent. The real prospects for organized political activity under British rule were hardly an improvement from that of the Germans.

Since independence in 1961, there has been a marked willingness to deploy the Tanzania

People’s Defense Force (TPDF), the Party’s paramilitary force (Jeshi la Mgambo), the Tanzania

Police Force (TPF), along with an assortment of other militias and vigilante groups, in order to

65 stifle opposition voices (Shivji, 1991; Brennan, 2006; Commonwealth Human Rights Initiative,

2006). Organized labor was subdued by jailing and exiling labor leaders or by exerting direct party control through an affiliate organization, the National Union of Tanganyika Workers

(NUTA). The right to strike was effectively curtailed by the 1962 Trade Disputes Act, while labor conduct was managed through the creation of a Permanent Labor Tribunal (PLT), and

Workers’ Committees (MacDonald, 1966; McHenry Jr., 1994; Chambua, 2006). Likewise, the eradication of the Native Authority System, as the hallmark institution of British colonial rule in rural society, was characterized by detentions and deportations, prior to the replacement of the entire system with local TANU commissioners under the supervision of the Minister of Local

Government (Maguire, 1969; Hopkins, 1971).

The ensuing conflict, which will be explicated in more detail in the following chapter, between the Party and existing or emerging social forces, such as labor, chiefs, ethnic associations, and cooperative societies, played out most intensely during the brief period of multiparty politics prior to the passing of the interim single-party constitution of 1965. Again, parties opposed to TANU were subdued through a series of coercive moves on the part of

TANU and state security organizations, most notably the Party’s youth league (TYL).

Christopher Kasanga Tumbo, leader of the Peoples’ Democratic Party (PDP) and head of the

Africanization1 faction within the TFL, was jailed and deported for his outspoken leadership

1 One of the most pressing debates at the time of independence was over the speed of which expatriate civil servants would be replaced by African civil servants. The TFL, along with the ANC and PCP, spearheaded the call for rapidly replacing the European expatriates and Asian commercial elite. From this perspective, the most important issue was the creation of jobs and experience for the African workforce, who for years was treated as subjects under colonial rule. Bienen (1970) cites data showing that by 1960, there were no African judges or magistrates, no African permanent secretaries, and no principle assistant secretaries. Data for 1961 reported some 1,170 Africans holding middle and senior civil service posts, out of a total of 4,452 posts (p. 124). Data for industrial and commercial employment shows that, at around the time of independence, less than 5 percent of the workforce was employed in executive or professional capacities (p. 271). Buchert (1994) cites that, at the time of independence, some 87 percent of the highest grade professionals, administrators, and commercial and industrial managers were either Asians or Europeans (p. 53). For the principle party leaders however, an equally important 66 against TANU. Government officials sympathetic or supportive of the opposition Africa

National Congress (ANC) were pressured to give up their district council seats. Chief Francis

Masanja, an outspoken TANU MP from Kwimba, was pressured to resign from TANU and subsequently deported after he was suspected of secretly supporting the ANC (Maguire 1969,

Bienen, 1970). In short, the consolidation of TANU’s position at the time of independence was buttressed through highly repressive tactics. Yet, as demonstrated below, the adoption of repression during the formative years was a common response among Africa’s newly independent states.

Throughout the single-party era, the coercive capacity of the Tanzanian state was buttressed by a collection of laws and ordinances passed during colonial and post-colonial times. As head of the executive, the President himself has had an enormous array of powers at his disposal, enough for President Nyerere to acknowledge shortly after independence that he had “sufficient powers under the constitution to be a dictator” (as cited in Hopkins, 1971, p. 27). The President appoints the Chief Justice, judges of the High Court and Court of Appeals, members of commissions established by the Constitution, and persons to ministerial posts (United Republic of Tanzania, 1992, p. 155). She can declare a state of emergency, detain nearly anyone without a warrant for an undetermined period of time, and exercise influence over the media and party auxiliary organizations. After the passing of the 1965 Interim Constitution, which essentially validated an already obvious trend toward single-party rule (Baregu, 1997), executive power has been legally defined as a prerogative of the TANU on the Mainland and the ASP on Zanzibar.

concern was the necessity of expertise in steering positions during the crucial transition to self-government. Again, citing Bienen (1970), “Neither the civil service nor the educational system had been geared to an early independence” (p. 124). Therefore, by the time independence day approached, there were only a handful of Africans with the professional qualifications sufficient enough to work as high and middle-level managers and executives.

A number of other laws assist in enhancing the powers of the President. The Collective

Punishment Ordinance of 1921, which was originally passed to hold villages accountable for land burning practices, essentially empowers the President to hold an entire village or tribe accountable for crimes committed within village boundaries. The Regions and Regional

Commissioners Act of 1962 and the Area Commissioners Act of 1962 grants regional and district commissioners the powers to arrest and detain any person for a 48 hour period when there is cause to believe this person is likely to disturb public peace or tranquility (United

Republic of Tanzania, 1992, p. 168). Under the Deportation Ordinance of 1921, the President can deport anyone “where it is shown by evidence on oath to the satisfaction of the president that such person is conducting himself so as to be dangerous to peace and good order…or is endeavoring to excite enmity between the people of Tanzania and the Government.” The most notorious of these measures is The Preventive Detention Act. Passed by the Kawawa government in 1962, this Act allows the President to detain without a specified duration2

(Baregu, 1997) “persons who, in their opinions, are conducting themselves in such a manner as to endanger peace and good order of the country” (United Republic of Tanzania, 1992, p. 166).

With some recent notable exceptions3, the courts have failed to serve as institutions for restraining arbitrary government action.. Professor Shivji notes that the courts in Tanzania reflect a “timidity and mediocrity” and a “loyalty born out of pressures and expectations of favor from the executive” (as cited in Peter, 1997, p. 124). Separately, Shivji defines the

2 This Act was amended in 1985, stipulating the detainee must be released after a period of fifteen days, or be notified about the reasons for arrest.

3 The passage of the Fifth Constitutional Amendment Act of 1984, which enacted the Bill of Rights, established a legal focal point for the Court of Appeals and High Court in making claims against state authorities, especially in relation to the seemingly arbitrary power to arrest, detain, and penalize (see for example Chamchua Marwa v. Officer i/c of Musoma Prison & and Attorney General, 1988; D.P.P. v. Daudi Pete, 1989; Happy George Washington Maeda v. Regional Prisons Officer, Arusha). Still, verdicts are typically ignored by Inspector General of Police, the Minister of Home Affairs, and the lower courts (Mwaikusa, 1996; Peter, 1997).

“authoritarian direction” of law in post-independence Tanzania as derived from a type of ‘crude positivist’ approach to law, which focused on applying rules as defined by the other two branches of government without ever grounding law on conceptions of rights, necessitating strong burdens of proof on the part of the state, or interpreting the role of the court as a guard against injustices done by other branches of government (Shivji, 1991, p. 117).

Yet, none of these acts and institutions of coercion makes the Tanzania case exceptional.

After striking the names of people, parties, and places, the institutional and legal environment above could describe nearly any African country formerly subject to British colonial rule. Laws like the Preservation of Public Security Acts in Kenya and Zambia for example, were simply different names for the nearly identical set of legal previsions and similarly styled jargon and justification found in the Prevention Detention Act in Tanzania.

Likewise, the flexing of the state’s coercive muscle during the single-party era in Tanzania was comparable to the practices of other single-party regimes. Perhaps the most notable exceptions were during the villagization4 measures of the 1970s, and the frequent resort to violence on the part of the Revolutionary Council5 on Zanzibar. Aside from these exceptions, the breakup of the ANC and the PDP during the early 1960s was nowhere near as turbulent as the “considerable official harassment” experienced by the Kenya African Democratic Union

(KADU) and the Kenya People’s Union (KPU) during the 1960s in Kenya (National Election

Monitoring Unit, 1993; Throup, 1993).

4 Villagization refers to the forced relocation of peasants between 1973 and 1976, in response to the failures of voluntary resettlement plans.

5 Following the 1964 Revolution, the government was supervised by a 30 member interim Revolutionary Council. Despite the fact that a Constituent Assembly was supposed to be called within one year following the Council’s formation, Zanzibar was effectively governed by Council decree until the promulgation of the 1979 Constitution.

Repression toward the KPU was especially harsh. Led by the prominent Lou6 politician

Oginga Odinga, the party’s ambitions were more akin to Nyerere’s socialist stance, and was thus seen as a direct challenge to the growing conservatism and pro-capitalist orientation of KANU under Kenyatta. Additionally, the KPU’s populism was threatening to expand its base beyond the Lou community into KANU’s base among the Kikuyu7. The threat of conversion was especially acute among the poorest of the Kikuyu, who grew tired of KANU’s empty promises of development. Thus, the KPU presented a serious threat to KANU and, as a consequence, was dealt with through some of the harshest forms of repression visited during the Kenyatta years.

During the late 1960s, KPU officials were constantly harassed, party branch offices were refused registration, and party rallies blocked by local government offices. Crucially, the KPU was denied the ability to hold Harambee8 (self-help) rallies, thus cutting off the KPU’s potential to build support through patronage. Finally, the 1969 assassination of Tom Mboya, widely considered to be Lou’s representative in KANU, was seen by many as orchestrated by the

Kikuyu elite who feared Mboya was on the road to succeeding Kenyatta as president. The angry crowds demonstrating against his assassination met the full force of the state security apparatus, in one case leaving 100 dead (Throup & Hornsby, 1998).

Another high profile murder in Kenya took place in 1975, when Josiah Mwangi Kariuki, a

Kikuyu dissident which again attempted to mobilize the Kikuyu poor against the KANU elite, was beaten and shot dead. Government reports implicated senior government officials in the

6 The Lou constitutes the third largest ethnic group in Kenya and are mostly concentrated alongside Lake Victoria in the western portions of the country.

7 The Kikuyu constitutes Kenya’s largest ethnic group, mostly found on the fertile grounds of the central highlands.

8 Harambee in Kiswahili means “all pull together” and was the name given to Kenya’s local self-help development efforts. Harambee rallies would be held to occasion financial donations by wealthy entrepreneurial politicians or by state grants attracted by well connected members of parliament.

70 crime. The murder coincided with a new era of repression, and Kenyatta made it clear that dissenting backbencher MPs would be crushed “like a hawk among the chickens” (as cited in

Throup & Hornsby, 1998, p. 20).

In Zambia, the period between single-party rule in 1972 and independence in 1964 was highly competitive, but also relatively violent. Unified by the concordal enthusiasm of independence, by the late 1960s, the ethnic coalition among the leadership circle within the

United National Independence Party (UNIP) began to unravel. Against the backdrop of greater centralization that was often interpreted through an ethnic lens as Bemba9 encroachment

(Sichone & Simutanyi, 1996), the Lozi10 faction within UNIP grew tired of their declining position within UNIP and subsequently defected to the African National Congress (ANC) just before the 1968 election. Prior to this, the ANC was largely confined to the Tonga in the

Southern and Central Provinces. UNIP’s position was even more tenuous following the fragmentation of the Bemba faction, when Simon Kapwepwe, a Bemba leader from the

Northern Province, defected to form the United Progressive Party (UPP) in 1971 (Tordoff,

1988; 1993).

The competitiveness during this period was subdued by a series of repressive and co-opting measures that typifies regime consolidation throughout Africa during this period. The destruction of the property owned by those in support of opposition parties, the identification and harassment of opposition sympathizers within Zambian Railways, and the 1968 opposition and UNIP clashes in Zambia’s Copperbelt that resulted in a ban against the United Party (UP),

9 The Bemba is mainly concentrated in the Northern, Luapula, and Copperbelt provinces. Chibemba, a Bantu language, is the most widely spoken tongue in Zambia.

10 The Lozi is concentrated in western Zambia, in a region known as Barotseland. 71 appear quite violent in comparison to TANU’s success in subduing the opposition in Tanzania

(Rasmussen, 1969; Tordoff, 1988).

Malawi undoubtedly illustrates the most repressive example of the four cases presented thus far. The aftermath of the 1964 Cabinet Crisis, where six cabinet ministers attacked President

Hastings Banda’s attempt to introduce school fees and his failure to take a stand against apartheid (Dzimbir, 2005), and where wider discontent was directed toward concerns that northerners were disproportionately represented in politics and disproportionately benefiting from scholarships (McCracken, 2002), resulted in ministers and unionists being exiled and detained without trial on one hand, and ministers struggling to demonstrate their loyalties to

Banda and the MCP on the other hand.

Following the formal adoption of a one-party state in 1966, Banda boldly proclaimed himself to be “president for life” and used the MCP to shore up personal loyalties (Phiri & Ross,

1998). Most notably, the Malawi Young Pioneers (MYP), an elite youth segment of the MPC

Youth League, became Banda’s “eyes and ears” (Dzimbir, 2005, p. 63) to circumvent the more professionally oriented military. As described by Phiri (2000), the Young Pioneers, along with the MPC Youth League and Police, “were a major instrument for the operationalization of Dr.

Banda’s one-party state dictatorship and terrorism” and the four cornerstones of the nation:

“unity, loyalty, obedience, and discipline” (p. 3). Kanyama Chiume, one the Banda’s opponents during the cabinet crisis, provides a more colorful depiction of Banda’s rule, where “youths were ordered to spy on members of their own families…became infused with terrible delight in beating and killing people after the cabinet crisis” (as cited in Dzimbir, 2005, p. 62).

Quantified data on human rights practices supports the conclusion that the level of repression in Tanzania during the single-party years was unexceptional. With the risk of

72 oversimplifying concepts such as civil liberties (CL) and political rights (PR), the data from

Freedom House illustrated in figures 4.01 and 4.02 clearly shows that the level of repression in

Tanzania was generally comparable to other African cases.

The first figure establishes comparative scores, including the sub-Saharan average, of political rights, which includes measures of the electoral process, political competition and participation, and government functioning. On a scale from 1 to 7, where lower values equal greater respect for political rights, the cases selected here range from 5 to 7. Tanzania consistently scored a 6 in every year between 1973 and 1989, not far above the average African score of 5.7. The most repressive regimes included Benin and Somalia, while, not surprisingly, the least repressive were Botswana and Mauritius. As for the comparative cases described in the previous paragraphs, Tanzania’s scores are lower than Malawi’s, with an average of 6.4, but higher than Kenya and Zambia, with averages of 5.4 and 5.2 respectively.

Figure 4.01: Political Rights

Source: Freedom House, 2005

The CL trends on figure 4.02 are somewhat similar to the PR scores, although with more variation within each case and between all cases. Measurements here capture variables such as freedom of expression, rule of law, individual rights, and freedom of association. Again, out of a

73 range of 1 to 7, Tanzania consistently scores 6 in every year, slightly higher than the sub-

Saharan average of 5.5. Out of the comparative cases, Malawi scored the most repressive average, at 6.6, while Kenya and Zambia scored the lowest, at 4.8 and 5.1 respectively.

There is plenty of research which points to a causal link between the production of a climate of fear during the most repressive times of single-party rule and a popular unwillingness to go against the incumbent even after voters and would-be oppositionists have greater opportunities to do so (Chehabi & Linz, 1998, Posusney, 2004). However, this casual link is clearly not born out of the data presented above. Given the similar civil liberties and political rights scores across these cases, and the fact that the introduction of multiparty rules resulted in the rise of oppositionists with enough support to unseat incumbents in some case (Malawi, Zambia, Kenya) but not others (Cameroon, Gabon, Tanzania), explaining the persistence of hegemony by singular references to legacies of fear is simply not robust enough.

Figure 4.02: Civil Liberties

Source: Freedom House, 2005

Perhaps the failures to correlate repression levels with regime outcomes following reform has to do with the fact that Freedom House indicators are too narrow to capture the rather broad spectrum of repression. What happens if we give treatment to the level of political competition

74 and political participation generally? To some extent, these two expanded slices are captured in the Freedom House data reported in figures 4.01 and 4.02, specifically the four questions related to political pluralism and participation under political rights. Still, it is worth investing a few pages to explicate the arguments in more detail. As will be shown below, because Tanzania’s single-party rule can be described as relatively competitive and participatory, some prevalent theoretical claims within the literature on African politics would predict that the push for multiparty change in Tanzania should be accompanied by both widespread popular protests and a somewhat durable array of opposition parties, thus posing a substantial challenge to post- reform CCM rule. Yet, as the early 1990s unfolded, just the opposite was true.

According to some works, historical patterns of authoritarian rule that severely curtail political competition and participation do not bode well as a pathway for the rise of sizable and sustainable contention once new political opportunity structures are realized (Gabriel, 1999;

Erdmann, 2002; Aidoo, 2008). This point has been rigorously defended by Bratton and van De

Walle’s (1997) work on democratic transitions from neo-patrimonial rule in Africa. In this seminal work, the authors make a number of credible claims, two of which are important here.

First, because neo-patrimonial rule is sustained by the mutual exchange of political support in return for material reward, exchanges which span from local level administrators on up to the highest office in the executive (Bayart, 1993; Bratton & van de Walle, 1997; Taylor, 1999), the push for political change often comes from those that find themselves outside the entitlements of political access (Bratton & van de Walle, 1997; van de Walle, 2003). Since all of our comparative cases include variations of neo-patrimonial rule, this claim by itself, cannot serve as a robust explanation for the Tanzanian case. Secondly, while democratic transitions from neo-patrimonial rule to democratic rule are rare, transition prospects are highest when transiting

75 from competitive single-party regimes (Bratton & van de Walle, 1997; van de Walle, 2002).

While this piece is not about democratic transitions per se, the reasons for the comparatively promising prospects of democratic transitions from competitive single-party regimes speaks directly those factors which influence the degree of party dominance after multiparty change.

The two principle characteristics of competitive single-party neo-patrimonial rule include relatively energetic levels of political participation and competition. Where high levels of participation ensue, citizen involvement in politics emerges as an act normally used when political questions arise. At the time of political reform, it is more probable that citizens will see popular protest as an appropriate channel for expressing disapproval against authoritarianism and growing economic mismanagement for example. Where authoritarian periods are characterized by high levels of competition, political entrepreneurs developed a space for organizing clientele networks as a basis for contesting elections. These networks in turn, constituted the capital for mobilizing political supporters into relatively sustained array of opposition parties in the aftermath of reform (Bratton & van De Walle, 1997).

Scholarly works on Tanzania authored during the early years of independence have been challenged for their over idealization of the degree of competition and participation within a restricted single-party framework (Martin, 1975; Baregu, 1997). Nevertheless, in comparative terms, evidence does support the claim that Tanzania, from the time of independence until the onset of multiparty reform, evolved with a single-party electoral system characterized by intra- party competition and a strong participatory spirit (Nellis, 1972).

Most of this democratic spirit was projected by Nyerere himself. For the independence leader and subsequent President, elections were seen as a necessary instrument for exerting the

“popular will” peacefully, allowing masses to “achieve changes in the laws…[and] personnel”

76 which governed their lives (Nyerere, 1968, p. 5). The process of apportioning officials with law- making capacity was to reflect this democratic spirit. In depth research done by Hopkins (1971) indicated that, at the time of the rise of socialism in Tanzania, “democracy and democratic institutions [were] highly valued by most members of the elite, and the rights of criticism and freedom of thought [were] widely accepted” (p. 236). More recently, Pratt (2002) cites that the emphasis on democratic engagement of all citizens was one of Nyerere’s most important legacies (p. 40).

From the 1962 presidential election, where Nyerere soundly defeated Mtemvu with 99.2 percent of the votes cast, to Nyerere’s decision not to contest the 1985 election, bids for the presidential office were however, hardly competitive. At the conclusion of the CCM annual conferences, or the joint TANU-ASP conferences prior to the election of 1980, Nyerere emerged each time as the default choice. In the run-up to the 1970 election, Zanzibar’s President

Abeid Karume stated on record that because of the obvious consensus among TANU delegates was in favor of Nyerere, there was no need to hold elections and therefore Nyerere should be made president for life (Mutahaba & Okema, 1990). Unlike Banda’s status as Malawi’s

President for life, Karume’s words never made it into the force of law in Tanzania. However, during the general elections, the pattern that emerged following the 1965 single-party “yes or no” presidential ballot was an overwhelming endorsement of Nyerere’s tenure.

The low tolerance for contestation in presidential elections in Tanzania was hardly unique among the population of single-party regimes found across the African continent. In Zambia, in the single-party elections between 1972 and 1991, Kaunda was selected on a similarly styled yes, no ticket. In Malawi’s single-party election between 1964 and 1994, where all citizens were required to be members of the MCP, Banda was unanimously selected to be president by the

National Assembly, until being declared president for life in 1971. A similar arrangement obtained in Kenya, where Jomo Kenyatta, who presided under defacto single-party rule, and

Moi, who presided over a constitutional single-party state, were each elected unopposed out of the National Assembly between 1969 and 1988.

Among single-party regimes, elections for parliamentary seats were more or less open to contestation. Tanzania is no exception. Under the single-party system, contesting a parliamentary seat started with the Party’s screening process11. Prior to the 1970 election, nearly any Tanzanian citizen of sound mind over the age of 21 with a TANU membership card and without a recent criminal record could run for office (United Republic of Tanzania, 1998).

Hence, there was a reasonable amount of grassroots influence in the type of person to be selected in the nomination process (Okema, 1990). It is worth noting that, by the 1970 election, new guidelines permitted only those committed to principles of socialism, to stand for election under a TANU, and later, CCM ticket.

After candidates filed nomination papers with the district returning officer, the screening process began with the preferential votes cast by party delegates at the Annual District

Conference (ADC). Typically, delegates to the conferences included Party district officers, resident regional officers, constituent MPs, two branch delegates along with branch chairpersons, and delegates from affiliated institutions such as trade, women’s, and cooperative organizations (Harris, 1967; Bienen, 1970). After voting at the ADCs, the names of the largest vote-getters were then referred to their respective Party executive committees, starting with the

District Party Committee (DPC), then the Regional Party Committee (RPC), the Central

Committee (CC), and finally the National Executive Committee (NEC). This process, which

11The description of the nomination process under single-party rule also applies generally to the vetting process within the CCM under multiparty rule. 78 was guided by internal Party norms and interests rather than formal legislation (Mpangala,

1994), allowed Party committees to gauge candidate leadership qualities, commitments, and ethics, as well as ensured candidates were easily disciplined by the Party command and control hierarchy (Van Donge & Liviga, 1989; Mvungi & Mhina, 1990)12.

The official opening of the campaign period followed the NEC screening. All campaign meetings throughout each constituency were attended by each of the TANU/CCM parliamentary contenders13, typically consisting of two contenders. Outside the growth of illegally funded private campaigning during the 1980s, which included lobbying members of the ADCs, offering gifts in exchange for support, and the use of institutional machinery as venues for holding campaign events (Mpangala, 1994), campaigns under single-party rule were exclusively financed by the state.

Parliamentary turnover was quite common in single-party Tanzania, a trend that has persisted after multiparty reform. The most telling was the watershed election of 1965, which demonstrated that some degree of election choice could be attained within a single-party framework. At the very least, the election served as a relief valve for those wanting to throw out lavishly living government officials who broke ties with their respective localities after taking

12 Overturning an ADC decision was justified on a somewhat nebulous array of qualities, including “dismissal from leadership, making trouble, sexual promiscuity, racketeering, individualistic tendencies, lying, lack of leadership, being indiscreet or untrustworthy, misappropriation of Party funds, and many ‘sins’, including drunkenness, and participation in polluting the political air” (Bavu, 1990a, p. 26), More often than not, nomination changes reflected considerations of patronage rather than ideological or behavioral concerns. However, overturning an ADC verdict was a risky affair, and often resulted in local protests, burning registration cards, voter abstinence, and ultimately a rejection of the Party’s chosen. The number of NEC changes is as follows: 14 in 1965, 30 in 1970, 15 in 1980, and 43 in 1985 (Okema, 1990, p. 54). Out of all of the candidate changes between 1960 and 1990, the electorate rejected around 49 percent of the candidates preferred by the NEC, suggesting that the party leaders did not always have the final say (Baregu & Mushi, 1994, p. 113).

13 Venues and timetables are arranged by a campaign committee, composed of the District Party Commissioner (DPC), Election Supervisors, as well as representatives of the Returning Officer, the candidates, and the District Police (Liviga, 1990). The candidates were not permitted to travel to and from campaign venues without the escort of election supervisors (Harris, 1967).

79 their seat in the nation’s capital. Thus, elections functioned as a way of legitimating the CCM’s tenure and the single-party framework. While the turnover for the most important ministerial posts was much lower, the overall turnover of the 1965 election was impressively high. Less than a quarter of the incumbent members of parliament were returned (Morgenthau, 1965, p.

15), and two ministers and six junior ministers defeated (Maguire, 1969, p. 373). Three out of five incumbent area commissioners were defeated, along with eighteen out of thirty-four local government officials (Bienen, 1970, p. 401). In the Sukumaland14, none of the seven incumbent members of parliament were re-elected, including the powerful Minister of Finance Paul

Bomani, and the former national vice-president of TANU John Rupia (Maguire, 1969, p. 374).

These turnover rates were carried out in subsequent elections. In most cases, incumbent failures to secure a popular vote had less to do with a lack of ideological conviction or being implicated in a scandal and more to do with their failures to bring home development projects and other services. At the same time, the National Executive Committee typically rejected incumbents based on a demonstrated lack of ideological conviction or being implicated in a scandal. Given an environment where developmental resources were scarce and the fact that politicians typically viewed political office as a means to an income, it is little wonder that turnover was relatively high where expectations include ideological commitment, cleanliness, and ability to bring benefits to constituents. By the 1985 election, Van Donge and Liviga (1989) concluded that “Few Tanzanian MPs serve more than one term and, if they seek re-election, they often fail, which means that parliament is not a closed shop but a relatively open institution” (p. 51).

14Sukumaland is a term used to describe the territory inhabited by the ethnic Sukuma. The Sukuma is largely found to the southeast of Lake Victoria. 80

The expression of dissent and dissatisfaction during the parliamentary campaigns is also a strong indicator of the degree to which political competition was tolerated. Data on parliamentary political campaigns during single-party rule indicates that open dissent appeared to be a normal part of the political campaigns. While open dissent may, in part, be little more than political theater, it also meant that incumbent failures would be brought into the open and candidates would be forced to openly navigate the tensions between the material demands of the constituents, and the ideological demands of the higher Party echelons.

During single-party elections, codes of conduct and election laws allowed the audience to ask three questions to each candidate, although the chairperson reserves the right to reject questions deemed irrelevant or not specific enough. At times, chairpersons displayed a marked tendency toward restricting challenging questions. In one case in Dodoma during the 1985 elections, all six questions were rejected as bad questions. Some of the rejected questions asked why campaign promises never resulted in actual policy, or why elected officials only come around during election times (Mvungi & Mhina, 1990, p. 118).

In most cases however, questions were accepted and were quite scathing at times. Table 4.01 details about the volume of questions during the 1985 election in Morogoro Urban. Attendees consistently used about three-quarters of the space allocated for questions, indicating a respectable level of interest among the attendees in probing the political aspirants. It is possible that this number would have been higher if more time would have been allotted to allowing the audience to formulate questions. In one case in Dodoma Urban, the chairman started the question session for each candidate by counting to three. If no questions were asked during this brief time frame, all three chances were forfeited (Mvungi & Mhina, 1990, p. 118).

The data on table 4.01 also shows that, in the case of Morogoro Urban, questions were rejected only three percent of the time. While this is a small slice of a very large election process, out of the five case studies presented in a 1990 volume entitled “Tanzania: Democracy in Transition,” only one case (Dodoma Urban) was reported to have seriously high rates of question rejection (Mvungi & Mhina, 1990).

Table 4.01: Questioning During Campaign – Morogoro Urban, 1985 Candidate Expected Questions Rejected Percentage Percentage Questions Asked Questions Asked Rejected

Malinza 120 92 2 77% 2% Khan 120 88 3 73% 3% Total 240 180 5 75% 3% Source: Liviga, 1990, p. 128

The nature of the questions asked indicates a general openness to speak one’s mind. One incumbent was asked “why did you decide to contest for this seat when you know you are not capable?” (as cited in Mwakyembe, 1990, p. 140). In a separate case, another incumbent was questioned over his concern over the quality of beer produced by Tanzania Breweries: “You as our member of Parliament fully know the water problem, but instead of bringing us water you decide to take ‘dirty’ beer to the Parliament as if that is our problem – is this what we sent you to the Parliament for?” (as cited in Munishi & Mtengeti-Migiro, 1990, p. 183). Frustration was voiced in Dodoma with the frequently heard question: “if we elect you, will you disappear into the city like your predecessors and only come back after five years to ask us for votes again?”

(as cited in Mvungi & Mhina, 1990, p. 112). In a campaign meeting in Mbozi during the 1985 elections, the audience responded to the call for questions by stating “we have no questions! Off you go!” (as cited in Mwakyembe, 1990, p. 141).

No doubt, the competitive political environment described above exhibited some clear constraints, a few of which eroded over time in the lead-up to the multiparty reforms. For

82 starters, like the other comparative cases discussed at length so far, the single-party framework in Tanzania effectively ensured that competition would have to come from within the sole political party. Secondly, the Party control over nominations of parliamentary contenders, which meant that district, regional, and national executive committees could veto candidates nominated at the Annual District Conferences, effectively translated into competition for the favor of the Party leaders rather than a singular focus on cultivating a base among the electorate.

During the first single-party elections, sixteen candidates were overturned by higher party organs. In 1990, during the last single-party election, twenty-three candidates were rejected in such a manner (Harris, 1967; Mpangala, 1994).

However, there was a competitive framework that would, and did, entice political entrepreneurs to build powerful personal political networks between themselves, the executive leaders, and grassroots constituents. Some of these officials, most notably Augustine Mrema, emerged as popular opposition figures in the lead-up to multiparty elections. Perhaps the most competitive aspect emerged following the economic liberalizations measures negotiated in

1984, which led to the rise of a more prominent position of business interests and non- governmental organizations (Mmuya & Chaligha, 1992). While private campaigns took place in earlier elections, the move toward socialism in the late 1960s instituted more discipline in election campaigns and effectively barred the nascent business community from participating in them (Kiondo, 1994). However, following economic liberalization private campaigns were conducted openly in nearly every constituency and introduced a new sense of unfairness in the election process. These included privately lobbying party ward and branch leaders, the use of private funds in financing local projects and tip giving, using schools and hospitals as campaign

83 resources, enlisting support from local religious, ethnic, and business notables, and campaigning by tapping religious or ethnic sentiments (Mpangala, 1994).

A comparative assessment of electoral competition in Tanzania would certainly be strengthened by comparisons of incumbent turnover levels between all four cases.

Unfortunately, such data is hard to come by. The only confident figures are the 49 percent turnover rates in Kenya’s 1969 and 1974 election (Throup, 1993, p. 377), compared to the two reported figures for pre-form Tanzania, 75 percent in 1965 and 26 percent in 1990 (Luanda,

1994, p. 259). Other spotty indicators suggest that, in terms of the number of competitors in the nomination processes, Tanzania’s elections had significantly more contestants than found in single-party Zambia15.

As far as Malawi is concerned, what can be confidently stated is that Banda exerted an almost exclusive personal control over the lists of candidates presented to the electorate

(Kalpeni, 1992; Ihonvberere, 2003). In fact, during the elections of 1971 and 1976, Banda simply nominated one candidate for each constituency who was then elected unopposed

(Meinhardt & Patel, 2003). To a considerably greater extent than Zambia, Kenya, and Tanzania, candidates in Malawi were almost singularly concerned with demonstrating their personal loyalties to the President and the MCP than they were about cultivating support among grassroots constituents.

At the same time, qualitative accounts indicate that in Kenya and Zambia there was greater space for constituency choice and generally more room for political entrepreneurs to cultivate personal ties to the electorate when compared to what was the case in Tanzania. In Kenya in particular, election competition was almost exclusively carried out as a private, non-party affair.

15In Tanzania, the ratio of contestants to contestable seats was 7.5 in 1965, 10.6 in 1970, 9.6 in 1985, and 8.5 in 1990 (Van Donge & Livinga, 1989, p. 47; Luanda, 1994, p. 258). 84

There were no ideological screening processes, although Moi did introduce measures to ensure more personal accountability to him. Members of parliament competed on the basis of their personal contributions to local development projects, along with their abilities to attract development assistance from the executive. It was furthermore, quite common for politicians to use ethnicity as a basis for mobilizing local support, a process heavily shunned in Tanzania. As a consequence, members of parliament in Kenya and Zambia developed very powerful personalized clientele networks (Barkan & Okumu, 1978; Barkan, 1994) that indeed acted as a resource for party building and constituency cultivating following multiparty reform. While

Banda in Malawi exercised more personal control over clientele networks to grassroots supporters, this control was at times challenged by powerful local notables through the language of regional and ethnic loyalties.

The political space for competition under single-party rule in Tanzania may have been somewhat more constrained than what was found in Kenya under Kenyatta and Zambia under

Kaunda. However, rates of political participation in Tanzania have been far more outstanding

(Chazan, 1982). As defined by Bratton and van de Walle’s (1997) description of democratic transitions from neo-patrimonial single-party rule, high levels of political participation raise the likelihood of popular protests as an impetus for multiparty reform. Let us briefly look at the data on political participation.

The Annual District Conferences (ADC), where local party members selected the list of parliamentary candidates to be forwarded to higher party organs, marked the most significant event for bringing together local party activists, and was thus well attended on average. In a small sample given by Mpangala (1994), attendance rates for the 1990 preferential voting

85 ranged from 60 percent in Kibaha to 90 percent Babati. Other turnout rates included 73 percent in Mbeya, 65 percent in Lindi, and 67 percent in Songea Rural (p. 40).

Attendance rates at campaign meetings or rallies were mixed, appearing to vary based upon the appeal of the parliamentary contenders, the legitimacy of the Party’s screening process, and the presence of national political figures at the rally. Campaign venues during the single-party era included CCM offices, stadiums, primary schools, markets, assembly halls at the University, factories, and army barracks (Liviga, 1990). In some cases, campaign meetings would bring the community together under a shade tree, occasioned by songs and dances. Harris (1967) cites that in a number of the more remote parts of the country, where populations were highly dispersed, campaign meetings would be the “only community gathering in months” (p. 37).

The scantly available data suggests that, while campaigns for parliamentary contenders could indeed be lively, many were simply dull and uninteresting. A rough estimate by one observer in the 1985 elections placed the typical campaign attendance between 40 to 70 percent of the registered voters (Liviga, 1990). Another study conducted in the months following the

1970 election cites that around 47 percent of those sampled stated that they managed to attend at least one campaign rally, while nearly two-thirds of those who knew about the campaign actually showed up (Hall & Lucas, 1974, p. 179). However, in cases where there was widespread popular opposition to the Party’s candidate, attendance was quite low. In one case an observer noted about a 14 percent attendance rate at campaign meetings. In other cases, attendance might be high during the start of a campaign, typically reserved for the presidential contest, but then evaporate once a local candidate took the platform (Mwakyembe, 1990).

According to most of the data from the 1985 election, low turnout to campaign rallies reflected the fact that the majority of the voters decided who they were voting for prior to the start of the

86 campaign period16 (Bavu, 1990a; 1990b), or reflected a clear dissatisfaction in the candidates running for office.

Comparative data on participation is largely restricted to contrasting the more systematic data on rates of voter registration and turnout, rather than on the scant and questionable data on nomination or campaign rally attendance. Furthermore, registration and turnout rates are useful indicators of the rate of participation among the general population, most of which cannot be considered party cadres or activists. Where voter participation is found to be historically high, one might expect to find the development of norms of political action which could overcome collective action problems in other venues of participation, namely political protests (Knack,

1992; Bratton, 1994).

Throughout the election process from the time of independence until the first post-reform election, the CCM has in practice been the predominant agent in the selection of candidates, along with the registration and mobilization of voters. At the same time, the government, largely through the National Election Commission was - and still is - predominant over the technical aspects of registration and voting, such as providing registration lists, voter identification cards, and ballot materials (Mpangala, 1994, p. 51).

Poor communication, the lack of experience on the part of Party, government officials, and the population generally, paired with sparse populations were acute obstacles to raising participation rates in early elections throughout all of Africa. Tanzania was certainly no exception. During the course of the 1965 election, the local TANU organization exhibited strong efforts to register as many people as possible, while in some areas, efforts were far less commendable. All-in-all, during the first single-party election, constituency registration ranged

16 In Ilala during the 1985 election for example, nearly 45 percent of the voters had decided on their choice before campaign day (Bavu, 1990a, p. 33). 87 from slightly more than 30 percent to over 90 percent (Harris, 1967, p. 28). Nevertheless, the election turnout was a great improvement from the elections of 1960 and 1962, highlighting the increased capacities of the Party to mobilize voters. For Bienen (1970), the fact the TANU was able to mobilize half of the adult population in 1965 to cast a ballot on election-day represented

“a real achievement in a free, honest election and registration” (p. 404).

Turnout in later elections was relatively consistent with the 1965 election. Table 4.02 is a sample showing some turnout variations from the last single-party parliamentary election. As the sample below suggests, the rate of turnout and registration remained relatively high in nearly every constituency, with the exception of the notable lower rates in Mtwara Urban.

Table 4.02: Sample Turnouts for the 1990 Election Constituency Eligible Registered Number of Registration Voting Rate Voters Voters Voters Rate Among Eligible Voters

Singida Urban 30,000 25,919 21,473 86% 72% Mtwara Urban 27,301 19,135 14,482 70% 53% Mbeya Rural 65,416 58,850 47,657 90% 73% Dodoma Urban 93,430 79,900 61,750 86% 66% Rungwe 101,148 88,302 63,010 87% 62% Lindi Urban 13,100 12,262 8,254 94% 63% Mbozi 145,669 104,444 83,324 72% 57% Mbeya Urban 61,730 50,054 35,815 81% 58% Source: Mpangala, 1994, p. 47

The data presented in appendix B, albeit a highly incomplete set, places turnout rates in

Tanzania in comparative perspective. At no point in post-independence Tanzania, does the ratio of votes cast to registered voters fall below 70 percent. In Kenya, while pre-independence elections bore wide participation largely due to the high level of competition between KADU and KANU, voter turnout after independence remained well below the 70 percent mark. In

Zambia, while the first post-independence elections of 1968 were a remarkable achievement in terms of participation levels, the elections in 1973 and 1991 both stand out as remarkably low.

Even in Malawi, where the first two-post reform elections were quite impressive, the most recent multiparty election witnessed a notable decline in participation. In sum, the data in appendix C shows that, while participation in Tanzania has varied considerably, overall participation in elections in Tanzania after independence has been remarkabley higher than

Kenya, Zambia, and Malawi.

High rates of participation, according to Bratton and van de Walle (1997) should correspond with relatively high rates of political protests in the lead-up to multiparty reform. But, just the opposite took place. The size of the domestic protests to single-party rule generally, and the

CCM specifically, was significantly smaller than protests against authoritarian incumbents in nearly any other African country. Perhaps the relationship between political participation and the rise of political protest is more complicated than the argument suggests by Bratton and van de Walle. Perhaps single-party rule by the CCM was commonly seen as a better alternative to multiparty elections and rule by an opposition party. Before addressing these questions, a brief review of repression after multiparty reform is in order. Indeed, what will be demonstrated in the following paragraphs is that post-reform repression cannot be inserted as a robust explanation for party dominance in Tanzania largely because, in comparative terms, the CCM does not appear all that repressive.

Repression after Multiparty Reform

The rise of opposition parties and political protests in Tanzania were quite modest in comparative perspective. At the same time, what can be said about the patterns of repression since multiparty reform? As clarified from the outset, the CCM’s dominance vis-à-vis opposition parties appear to have become more entrenched. Perhaps the knee-jerk reaction is to

89 suggest that increased harassment and intimidation against the opposition is a sufficient explanation. Indeed, many of the opposition leaders I interviewed suggested this to be the case.

However, the data does not support such a conclusion.

Since the surge of multiparty reforms in Africa, the level of overt repression in Tanzania does not appear abnormally high or low when compared to many other cases of party dominance. As numerous cases suggest, the police and the TPDF do indeed periodically arrest and detain opposition party leaders17, deploy excessive force for breaking-up opposition rallies18, and commit seemingly random acts of violence against citizens for unclear reasons19.

At the same time, a number of surveys show that, while repression might be a characteristic of daily political life, Tanzania does not appear to be a comparatively repressive polity. A survey conducted by Research and Education for Democracy in Tanzania (REDET) during the middle of the 1990s shows that coercion is the least reported factor behind the decision to register to vote, as shown in table 4.03. In the same survey, while few people avoid discussing politics with public officials, nearly 34 percent cited that they do not avoid discussing public

17 For example, in October, 1999, TLP leader Augustine Mrema was arrested for making “derogatory” comments about the wife of President Mkapa (U.S. Department of State, 2000g). In the spring of 2001, CUF leader Seif Shariff Hamad was arrested for holding an unlawful meeting (U.S. State Department, 2002g). In November, 1999, October, 2001, and August, 2002, DP leader Christopher Mtikila was arrested on charges of sedition (U.S. State Department, 2000g; 2001g; 2003g).

18 For example, in 1994, police fired teargas on an opposition march on the National Assembly (State Department, 1995g). In 1999, police field force units used clubs to break-up a peaceful demonstration held by the NCCR- Maguezi (State Department, 2000g). Tear gas was routinely used against CUF demonstrators in 2000 and 2001 (State Department, 2001g; 2002g),

19 For example, in December 2003, some 100 TPDF soldiers intervened in a “fight over a woman” in which the soldiers attempted to seduce. The intervention, which was described as a “rampage” resulted in the injury of four civilians and the destruction of seven vehicles (Tarimo, 2003, December 13). In Arusha during April 2004, following the death of a soldier, TPDF personnel went on a “rampage” by indiscriminately “beating civilians”. Some 20 civilians were injured as a result of the attacks (U.S. State Department, 2005g, 1C).

90 affairs with anybody in particular (Mushi, Mukandala & Baregu, 2001, p. 102)20. Another

REDET survey attempted, through a series of questions evaluating the 2005 election, to assess the amount of trust that Tanzanians have in the political system. The results, which are listed in table 4.04, clearly show that, while some 11.5 percent did not believe the elections would be free and fair, the overwhelming majority believed they would be peaceful and free and fair

(Research and Education for Democracy in Tanzania, 2004, p. 16).

Table 4.03: Reason For Registering to Vote Feelings When Going to Register Frequency Percentage I feel satisfaction 3022 61.8 I feel a sense of duty 850 17.4 I feel coerced 60 1.2 I get no feeling 156 3.2 Other 96 2.0 Don’t know/no answer 706 14.4 Source: Mushi, Mukandala & Baregu, 2001, p. 162

Table 4.04: Evaluate the Upcoming 2005 elections Responses 2005 elections will be peaceful 2005 elections will be free and fair Yes 78.2% 76.8% To some extent 8.5% 8.6% No 10.9% 11.5% Don’t know 2.4% 3.1% Source: Research and Education for Democracy in Tanzania, 2004, p. 16

Macro-level data reflect these trends. Again, with the risk of oversimplifying concepts such as civil liberties (CL) and political rights (PR), I constructed two figures illustrating Freedom

House scores for countries where regime change has not taken place as a consequence of

20 The possible interpretation of the result from this particular question (“if you wanted to discuss political and government affairs, mention people with whom you fell it is better not to discuss such topics”) are especially wide. First, some 38.7 percent of the respondents either did not know or refused to answer. In fact, such a response was quite common throughout this REDET survey. In one question (reasons for not discussing political affairs with some people), over 75 percent either did not know or would not answer. It might be argued that there is a general apprehension toward participating in such surveys simply because of fear that doing so might result in reprisal from the authorities. Secondly, members of the opposition were the most frequently cited specific group that is avoided during political discussion (6.8 percent). Again, there is no clear indication as to why this is the case. On one hand, it could be that members of the opposition are simply not trusted, or seen as politically irrelevant. On the other hand, an equally plausible interpretation is that being seen in public while discussing affairs with members of the opposition, could be perceived by some as personally risky. Nevertheless, the fact that 33.7 percent cited that they avoid nobody is a clear indication that, at least in some quarters of Tanzanian life, it is possible to openly discuss public affairs (Mushi, Mukandala & Baregu, 2001). 91 reform. Since KANU did survive the first two multiparty elections, I felt safe to include Kenya in this sample.

The trends in figures 4.03 and 4.04 suggest one important observation: Assuming civil liberties and political rights are robust measures of levels of repression, Tanzania has traversed from greater to lesser levels of repression at a more progressive pace than the African average.

Most importantly, the level of repression in Tanzania is considerably lower than most other cases of prolonged regime tenure. Since Tanzania’s first multiparty election in 1995, PR and CL ratings have averaged 4.3 and 3.7 respectively, versus the African averages of 4.5 for PR and

4.3 for CL. By contrast, in Togo, where the Rassemblement du Peuple Togolais (RPT) has succeeded in routing the opposition since first multiparty election in 1993, PR and CL scores have averaged 5.8 and 5.0 respectively. In Gabon, the Parti Démocratique Gabonais (PDG) continues to dominate with post-reform ratings of 5.1 and 3.9 for PR and CL respectively. The ratings are most extreme in Cameroon. Here, despite holding the first multiparty elections in

1992, the Rassemblement démocratique du Peuple Camerounais (RDPC) continues in office with Freedom House scores of 6.4 for political rights and 5.6 for civil liberties.

Figure 4.03: Political Rights Following Reform

Source: Freedom House, 2005

If we compare Tanzania’s PR and CL scores with those cases where multiparty reform resulted in regime change, it is clear that political environment in Tanzania has indeed grown comparatively less repressive. In Malawi, between the years immediately prior to reform and

1994, when the MCP lost control of the legislative and executive branches, PR and CL ratings were 6.5 and 6.0 respectively. In Kenya, ratings for the years spanning from 1990 to the 2002 election, which witnessed the death knell of KANU’s uninterrupted tenure, were 5.8 for PR and

5.5 for CL. By contrast, Tanzania’s ratings from 1990 up through 2008 were 4.7 and 4.1 for PR and CL respectively.

Figure 4.04: Civil Liberties Following Reform

Source: Freedom House, 2005

What conclusions can be drawn from the graphical trends in figures 4.03 and 4.04, and the ratings discussed above? First, repression is neither a necessary nor a sufficient condition for regime survival. Regimes with poor records for respecting civil liberties and political rights have survived multiparty elections, but have also been brought to their knees in the years following multiparty reform. Secondly, the CCM government has indeed become measurably less repressive since the first multiparty election in 1995, but the incumbent has also managed to widen its margin of victory in the two elections since 1995.

Data on human rights abuses confirms the patterns in the Freedom House scores. Human rights abuses tell us something about the regime’s willingness to engage in coercive acts in order to defend dominance in the short-term. Popular demands for freedom of association and freedom of speech for example, clearly pose an immediate threat to the incumbent’s ability to manage the sphere of political ideas and mobilization. Consequently, repressive regimes are generally inclined to exploit the coercive power of the state as a method for removing the threat.

Conversely, respect for human rights suggests that a regime is willing to accept democratic practices because of established norms of the rule of law, traditions of reciprocity, or simply because the regime wants to minimize the risks that current dissent might foment into more concerted and hardened challenges in the future (Scarritt & McMillan, 2000).

The U.S. State Department provides concise statements on country specific instances of human rights abuses from the early 1990s up to today. With the goal of confirming and refining earlier statements on repression, I compiled quantitative measurements of human rights abuses, including politically related disappearances, arrests and detentions, destruction of property, harassment, and interference with the media, from the largely qualitative data available in these reports. I selected seven cases of prolonged party dominance. These include Tanzania, Kenya,

Gabon, Togo, Botswana, Kazakhstan, and Malaysia. Again, while KANU was ousted in 2002, the incumbent did manage to survive the first two multiparty elections. Therefore, I felt it would be of value to include it in this data set. I then counted instances21 of human rights violation in each case between 1994 and 2006. The specific details of the methods of analysis and the aggregate data are presented in appendix A.

21While the details of data analysis are reserved to the appendix, it is important to note that figures represent instances of human rights violations. While it might be interesting to know just how many people were affected in each instance, State Department reports do not consistently report effected numbers. 94

Tanzania’s commonplaceness is again observed. As figure 4.05 illustrates, with the exception of the peak period in 2000, where Tanzania made up of more than 40 percent of the human rights abuses, the level of politically related human rights violations in Tanzania are on par with the six other case selections. Only Botswana and Gabon stand out as exceptionally low cases, although the ratings for Gabon would be far worse if election rigging was accounted for in this data set. Kenya, Kazakhstan, and Togo each stand out as the most probable violators of human rights. Kenya contributed to some 20 percent, and Kazakhstan and Togo contributed 21 percent each to the total volume of human rights abuses. Violations in are especially notable along the lines of politically related deaths and killings (Togo), media interference (Malaysia), and action against taken against political assembly (Kenya).

Figure 4.05: Politically Related Human Rights Violations

Source: U.S. State Department Country Reports on Human Rights Practices, 1995-2007

As for Tanzania, the persistence of the Registrar of Political Parties in denying the registration of Reverend Christopher Mtikila’s Democratic Party (DP) since the onset of multiparty reform22 (State Department, 2003g) leaves the country as one of the most likely to deny registration to an opposition party. Nearly 30 percent of the denial of registration cases took place in Tanzania. The inclination for detaining political opponents is also quite high,

22 The DP was finally registered in 2002.

95 largely reflecting the 1997 arrest of 18 CUF members on treason charges, which were dismissed by the Zanzibar High Court in November 2000. The high rates of harassment and intimidation partially reflect reprisals against protests over these detainees, culminating in the January, 2000 confrontation between CUF protestors and the police (State Department, 2001g)23, and thus the spike in human rights violations for that year. Finally, the police in Tanzania show a marked tendency toward denying the ability of opposition parties from holding rallies and meetings, especially during election years (State Department, 1996g; 2001g; 2006g).

However, after spending countless hours reading through dozens of State Department reports, a few additional comparisons can be made. First, peaks in human rights abuses in

Tanzania clearly coincide with election years, especially during 2000 and 2005 elections, where abuse levels were higher than adjacent years and were significantly higher than the average yearly contribution (at +27.2 and +8.2 respectively). Together, all three elections account for nearly one-third of the human rights abuses in Tanzania reported between 1994 and 2006.

Similar spikes are observed in other cases. In Gabon during the 2005 Presidential election,

Bongo’s countrywide ban on all opposition marches, the killing by security forces of a Union du

Peuple Gabonais (UPG) affiliate, and the arrest and detention of 14 UPG members, along with 9 others, for protesting the re-election of Bongo, account for the Country’s overwhelmingly high level of human rights violations for 2005. Likewise, in Togo, following the February 2005 death of President Gnassingbé Eyadéma, the controversial succession by his son Faure Gnassingbé, and the simultaneous blocking of Parliamentary speaker Fambare Natchaba Quattara’s claim over the Presidency, ignited a firestorm of protests which resulted in some 400 to 500 deaths from post-election demonstrations and raids by government security forces and RPT militia. In

23 Approximately 70 people were injured and several arrested when hundreds of CUF supporters attempted to assemble to observe the treason trial (State Department, 2001g). 96

Kazakhstan, the 2004 Mazhilis election and the 2005 Presidential election were the occasion for widespread attacks against journalists, and the use of the ill-defined Extremism Law to deny opposition parties the opportunities to rally. Therefore, while it can be cautiously stated that spikes in politically related human rights abuses may be more pronounced in Tanzania during election periods, most of the other cases show a similar tendency to reign in the state security measures during times of contestation.

Secondly, state officials on the Tanzanian Mainland appear much less inclined to publicly implicate the role of the state in repression, or to make unveiled threats against otherwise legitimate government critics. It is rare to find Tanzania officials making statements similar to the Malaysian Deputy Information Minister’s blatant implication of UMNO’s control over the media when stating that the opposition would be published in government media when they have something “specific or good to say” (State Department, 2001f) or the assistant minister to

Moi in Kenya, Fred Gumo, when threatening journalists with beatings for critically reporting on the Luhya ethnic group (State Department, 2000e). In a comparative sense, officials in Tanzania seem far more attentive toward the art of public relations.

Thirdly, all of the politically related killings in Tanzania are due to clashes between political parties or between demonstrators and police. Unlike the murders of Zamanbek Nurkadilov24 and

Altynbek Sarsenbaiuly25 in Kazakhstan, or the Crispin Odhiambo Mbai in Kenya26, there were

24 Zamanbek Nurkadilov, a former regime insider who claimed to have documents implicating President Nazarbayev in a corruption scheme known as “Kazakhgate”, was shot three times in his home in November 2005, just prior to the presidential elections (“Suspicions raised”, 2006, February 14; U.S. State Department, 2006d).

25 Altynbek Sarsenbaiuly, another former regime insider before co-chairing a leading opposition party called True Ak Zhol, was shot in February 2006 along with his driver and bodyguard, allegedly by officers of the National Security Committee (“Suspicions raised”, 2006, February 14; U.S. State Department, 2006d).

26 University don Crispin Odhiambo Mbai was killed on September 14th, 2003 when three assailants broke into his home and shot him several times. Mbai was a personal friend of chief opposition leader to the Kibaki 97 no reports of Tanzanian security officials deliberately targeting high profile opposition figures for assassination. Furthermore, the death tolls from confrontations were typically far lower than

Kenya and Togo. Given the four to five-hundred deaths following the 2005 elections in Togo, or the countless deaths following the politically incited ethnic clashes in Kenya since the onset of multiparty competition, killings as tactical weapons of intimidation appear far less prevalent in

Tanzania.

The final and most important observation concerns the geography of repression in Tanzania.

Given the two-tiered governing structure in Tanzania, authorities on the lightly populated archipelago of Zanzibar have sufficient autonomy in tailoring the tactics of repression to local conditions, without being penalized by cooler headed elements on the Mainland. These tactical differences are born out in the data presented in figure 4.06 and appendix A. When controlling for human rights abuses on Zanzibar, the volume of abuses for Tanzania declines dramatically

(although election year spikes remain). Indeed, all cases of long-term political detainees took place on Zanzibar, along with 73 percent of the cases of harassment, intimidation and violence,

83 percent of the cases of property destruction, and 61 percent of the arrests, detentions, or exiles. All-in-all, 61 percent of the cases of politically related human rights violations between

1994 and 2006 were restricted to the lightly populated Isles. While the denial of registration and surveillance still stand out as exceptionally high for the country as a whole (both of which impact Zanzibar), the remaining categories are all sufficiently below the mean of the seven country sample. When controlling for Zanzibar, out of the 1,817 politically related human rights

government, Raila Odinga, and a key player in efforts to trim executive power though constitutional reform. These efforts have been sharply opposed by what has repeatedly been called “the Mount Kenya Mafia” – Kikuyu “nationalists” and business elite who helped put Kibaki into the state house. Other allies to Odinga have been harassed and threatened with job dismissals and death, reportedly by theses Kikuyu nationalists (Murder most foul, 2003). 98 violations tabulated in this sample, Tanzania’s contribution is approximately 7.2 percent, about the same as Gabon.

Figure 4.06: Politically Related Human Rights Violations (Excluding Zanzibar)

Source: U.S. State Department Country Reports on Human Rights Practices, 1995-2007

This statement on the geography of repression in Tanzania has two specific functions for the remainder of this analysis. First, the statement acts as a disclaimer in an attempt to deflect some obvious lines of criticism against earlier claims that comparatively speaking, repression in

Tanzania has historically been unexceptional and, more recently, has actually been comparatively less prevalent. Obviously governance on Zanzibar is an exception to this claim.

More importantly, locating the geography of repression allows the treatment of Zanzibar as more than just an exception to everything said about Tanzania generally. It opens up the possibility for treating Zanzibar as a separate comparative case, pointing to the causal consequences that macro-structural divergence and attempts to mitigate divergence through exclusionary and repressive tactics has on regime survival after the onset of multiparty rules.

Considering the level of state violence against CUF on the Isles, paired with razor thin election victories, naked repression constitutes a highly valuable tactic for the reproduction of incumbency, albeit with the enormous costs of escalating episodes of arrests, detentions, reprisals, and deaths during election years, and the dwindling opportunities for peacefully resolving political conflict.

Conclusions

The volumes of data presented in this chapter points to three conclusions about the centrality of repression’s past and present in sustaining the CCM’s tenure, despite the introduction of multiparty elections. First, given the fact that the data in figures 4.01 and 4.02 indicate that single-party rule in Tanzania was no more or less repressive than the typical African case, and that many regimes with similar histories of repression found themselves out of office after the introduction of multiparty constitutions, there is little empirical basis for asserting that legacies of fear among the population of opposing the CCM, by itself, is a significant factor behind the

Party’s post-reform tenure. Given the paranoia and brutal attacks against opposing forces during

Moi’s tenure in Kenya or Banda’s rule in Malawi, both cases are among the most repressive instances of single-party rule in sub-Saharan Africa. Yet traditions of repression failed to generate a common reluctance toward expressing dissent during the lead-up and aftermath of multiparty reform. Why should we expect Tanzania to operate any differently? At the very least, we must conclude that any casual link between past repression and dominance in the present is sustained only through other interrelated variables.

Secondly, judging from the data presented in figures 4.03 to 4.06, the post-reform level of overt repression does not appear as a significant explanation for continued CCM rule. More specifically, explanations that intend to link restrictions on civil liberties and political rights, or patterns of human rights abuses as raising the costs of opposing the incumbent, does not seem to go very far in the Tanzanian case. In that case, we would expect to find higher levels of constraints and abuses.

Finally, according to Bratton and van De Walle (1997), single-party regimes with traditions of political participation and competition show greater promise in transitioning to polities with

100 more vibrant levels of protests and viable opposition parties. That was indeed the case of Kenya and Zambia. The applicability of this argument to Malawi and Tanzania is unclear at best. In

Malawi, the restricted political space under Banda’s personalistic rule should not have resulted in the UDF’s victory over the MCP and Banda in the first multiparty election in 1994. As for

Tanzania, given the high levels of political participation, we should at least have seen multiparty change accompanied by sizable, albeit disorganized protests.

In the chapter to come, I begin my assessment of the historical causes behind party dominance in Tanzania. In chapter five, I argue that the peripheral status of the Tanzanian colonial economy is the principle initial cause behind the scale of the CCM’s dominance today.

Being peripherally incorporated into the regional and global economy meant that social dislocations from economic development and colonial authoritarianism were comparatively minimal. Hence, at the time of independence, the constellation of social forces in Tanganyika lacked the material base or the mutual hostilities necessary for coalescing into a political force with the wherewithal for challenging the consolidation of power by a mostly urban array of political ideologues and civil servants. These political elites in turn had a relatively free hand in implementing sets of policies which had the long-run consequences of reproducing low degrees of macro-structural divergence. I argue that this macro-structural condition at the time of multiparty reform is the primary reason why opposition parties and protests appeared to be so insignificant. Let us turn to the beginning point of this argument.

101

CHAPTER FIVE STRUCTURES AND THE RISE OF SINGLE-PARTY RULE

“In Tanzania, it was more than one hundred tribal nits which lost their freedom; it was one nation that regained it.”

Julius Kambarage Nyerere, 1969

“We are fast learning that independence is not enough to rid us of the consequences of colonial rule.”

Kwame Nkrumah, 1963

I owe an intellectual debt to Thorvald Gran, a distinguished professor in Public

Administration and Organizational Theory at the University of Bergen, who, when presented with my research agenda, asked me “haven’t CCM elements always dominated Tanzanian politics?” Professor Gran, who spent considerable time in Tanzania during the 1970s, was undeniably correct. The Tanzania African National Union (TANU), which is the Mainland antecedent to the Chama Cha Mapinduzi, emerged from the struggle for independence only to compete in early multiparty elections from a seemingly uncontestable position. This statement on contestability could easily apply to the CCM in today’s multiparty elections. Indeed, when looking at the distribution of votes from Tanzania’s first multiparty elections, which took place between independence in 1961 and the formal proscription of opposition parties in 1965, one is struck by their similarities to the current election outcomes. This similarity speaks directly to the country’s remarkable degree of political stability, continuity, and the ability of the CCM to survive in office.

Professor Gran also asked the obvious: “what was the essence of this hegemony, and how does this hegemony relate to the current status of the CCM?” This chapter addresses the first part of this question. Why did TANU emerge in such a dominant position in the first place? The

102 answer in the shortest possible form is that the peripheral status of the Tanganyika (present-day

Tanzania Mainland) colonial economy, where land consolidation and the growth of a landless class was relatively mild, led to the development of an array of social forces that were fundamentally less polarized than what was found in many neighboring countries. Specifically, the lack of land consolidation meant that the causal force of a rural petty bourgeoisie was in a comparatively weak position for consolidating ethnic sentiments as a modality for challenging the consolidation of political power by the emerging urban elite. In short, the macro-structural environment permitted the convergence of social forces around not only the need for independence, but also on the need to embrace TANU as the primary agent for bringing about national unity and social development.

The bulk of this chapter is dedicated to fleshing out this argument in more detail. I start first by elaborating some theoretical concepts that form the crux of the analysis in this chapter, most notability the term collective identity and how identities relate to political parties. Then, I shift gears into an empirical discussion of the Tanzanian case, beginning with a comparative assessment of the degree of TANU dominance at the time of independence. Afterwards, I flesh out some colonial background to the Tanzanian case, beginning with a brief discussion of colonial rule and African political movements generally. Afterwards, I turn the attention more specifically to on the rise of TANU, the nature of the independence movement and political party, and finally move on to assess the overall structural context in which TANU operated.

Since the history of Zanzibar is radically different from that of the Mainland, before concluding the chapter, I will separately discuss the social, economic, and political background to the rise of the ASP and Revolution on the Isles.

103

The Material and Ideational Substance of Political Contestation

As a recap from the second chapter, identities1 are potentially important resources for political actors seeking to advance political agendas, even if those agendas reflect something other than the interests of those groups they claim to represent. Sidney Tarrow (1998) makes this point clear when claiming that political activists build social and political networks by tapping “collective action frames that justify, dignify, and animate collective action”. Social movement literature captures this process when using terms like “cognitive frames”,

“ideological packages”, and “cultural discourses” which are central components of identities and can be used by movement leaders to “inspire people to collective action” (p. 21).

Similar processes are often implicated in cultivating dominance. Brown (1982) argues that throughout post-colonial Africa, regimes have frequently deployed communal terminologies and references as a way of maximizing support and minimizing opposition, most notably in places like Nigeria (p. 39). In other cases, dominance may be cultivated by references to broader symbols of nationalism, unity, and the defense against subversive threats (Eriksen, 1991;

Bačová, 1998; Chachage, 1998).

The construction of political parties can be conceptualized in a similar manner. Przeworski and Spague (1986), and later Herbert Kitschelt (1995), in articulating a rational-choice institutional approach to Lipset and Rokkan’s seminal 1967 work on the origins of political party systems, focused on the incentives faced by political elites to strengthen or dilute social cleavage structures. When faced with competitive pressures, elites tend to solidify political constituencies by strategically emphasizing certain salient collective symbols, while de-

1 Collective identities are understood here as “the collective aspects of subjectivity” emerging out from an individual’s membership to a particular group and the “alleged” symbols, rituals, beliefs, and values of the group (Friese, 2002, p. 2). The dynamics of collective meaning are shaped by self-other discourses which define the boundaries between community insiders and outsiders (Straub, 2002; van Dijk, 2008). 104 emphasizing others (Norris, 2003). A variety of other authors have noted similar processes in political party consolidation in Africa, where elites employ collective symbols as a route for unifying a national whole or as a means for challenging incumbents (Brown, 1982; Lemarchand,

1994; Chachage; 1998).

However, just because political elites attempt to strategically mobilize or de-mobilize collective identities does not mean they while succeed in doing so. There are broad array of intervening variables between elite action and political outcome. Election rules and degree of power separation between the executive and legislative functions for example, can have profound impacts on the ability of political entrepreneurs to transform collectivities into instruments with casual political force (Sartori, 1976; Koford & Heckert, 1987).

Yet, as identified earlier, institutional rules governing political conflict in Tanzania, Zambia,

Kenya, and Malawi, have not shown considerable variation since the time of independence. All four cases have experienced prolonged periods of de jure single party rule. In the aftermath of multiparty reform, each case emerged as powerful presidential system with similar electoral systems, namely the first-past-the-post designs. Instead, what have varied significantly are 1) the degree to which identities are polarized, and 2) the ability of political entrepreneurs to translate identities into social forces with consequential impacts on the nature of multiparty competition.

As a recap of the discussion in chapter two, let’s look at each in turn.

The concept of polarization is well accounted for throughout the various social science fields. For Giovanni Sartori (1976), the degree of polarization between political parties is conceptualized in terms of “ideological distance”, where distributions of political power across parties favor those at the ideological extreme. Where polarized pluralism prevailed, political compromises are unattainable, governing coalitions are unworkable and political instability

105 becomes highly probable. In refining some of Sartori’s claims, Evans (2002) talks about polarization as the ideological distance between those clusters of preferences which constitute political parties.

The heart of the concept of polarization is found in the prefix polar: a continuum whereby binary distinctions are made in order to define the community and issue boundaries, distinctions often based on ‘self-other’ discourses as identified in social constructivist and post-modern research (van Dijk, 2004). Terms like “good” and “evil” generally represent a “lexical polarization” that has been used by rulers of the past and present as a way for framing the nature of political conflict (van Dijk, 2008). By contrast, terms like “greater” and “fewer” or “more” and “less” tend to be less polarizing since they denote a sliding scale rather than a world defined by rigid polar opposites.

Unfortunately, due to scope and resource limitations, the measurement of polarization in this chapter will not delve into these self-other discourses. Instead, I rely on historical accounts of polarization between social identities. My working definition of polarization addresses an array of institutional and economic factors which have been widely cited as conditional to the emergence of sectarian polarization throughout Africa. These include the penetration of cash crop production, labor differentiation, and colonial tactics of divide and rule. Polarization ensues out of these conditions primarily because various communities begin to see one another as a threat to their respective corporal interests. These structural formations, which remained relatively cohesive when opposition to colonialism was the dominant issue, were also sources of divisions that mapped onto political party formations once contesting elections for legislative assemblies and executive offices became part of the equation.

106

By itself, polarization between identities cannot capture the actual translation of identities to political parties. The single biggest shortcoming in discourse analysis is that it fails to theoretically account for the causal impact of lexical polarization, as previously defined, on political outcomes. I argue that the causal impact of collectivities will depend on the ability of political elites to mobilize material resources to advance personal and/or group interests. In short, while discourse might tell us about the possible constitution of actors and issues, under the logic of political competition, material resources speak to the causal power that actors have in establishing issue dominance and agenda setting. Where polarization prevails, and actors are able to muster resources for organizational building, mobilizing a population into a sustained political movement will be far easier than in an environment where polarization is weak and/or resources are scarce.

In reference to the Tanzanian case, my claim is that the peripheral status of the colonial economy in Tanganyika produced two interrelated macro-structural patterns having profound effects on the ability of TANU to consolidate power and dominate post-independence Tanzania.

First, comparatively speaking, the minimal amounts of land consolidation and capital accumulation throughout rural Tanganyika meant that elements from the petty bourgeoisie were simply unable to muster together the material resources for building sustained party organizations in opposition to TANU’s rise (Shivji, 1976). Secondly, the relatively minimal resort to divide and rule tactics by the British, which, as practiced in neighboring countries, elevated the economic and political status of one ethnic group at the expense of another, meant that ethnic consolidation and polarization among Tanganyika’s 120 language groups was minimal. Taken together, the ideational and material structures at the time of independence ensured a reduced likelihood that the most powerful and ambitious rural elites would be able to

107 emerge as rural spokespersons for popular ethnic movements with the capacity to challenge the urban intellectuals for control of the state (Hopkins, 1971; Hyden, 1994; Maxon, 1994)2.

TANU Dominance in Comparative Perspective

While the push for independence may have been a major incentive for formulating nationalist political parties throughout much of Africa, in many cases the rise of political parties had as much to do with efforts on the part of political elites to stake out sub-national claims and personal ambitions in colonial legislative assemblies. As the smell of decolonization was in the air, political elites understood that those actors which dominated the legislative assemblies would have enormous institutional clout to dictate the shape of independence. While competition over legislative seats was intense in Malawi, Kenya, and Zambia, from the time of its founding in 1954, TANU in Tanganyika was rivaled by no other organization outside the colonial administration itself.

During the terminal years of the 1950s, Nyerere wanted a sweeping victory for TANU in

Tanganyika’s first competitive popularity contest for the British concocted Legislative Council

(Legco)3. Whereas before, Council members were appointed by the colonial governor, now a

2 A criticism that emerged during the early stages of formulating this work, one that I take very seriously, concerned the Eurocentricity of concepts such as polarization and social cleavages, and the inapplicability of much of the political party literature to the specificities of the African context. Mahmood Mamdani, a leading critic of Eurocentric Africianist scholarship, could not have said it better: My work is suffering from too much Eurocentric weight. While applying concepts that sound an awful lot like social cleavage analysis to political party formation in Africa may indeed seem far too Eurocentric, it must be remembered that the entire project of multiparty competition, as a theoretical construct and historical fact, is also based on Eurocentric histories and processes. While underlying political networks in Africa may differ remarkably from those found in the histories or Europe, political actors quickly adapt to the prevailing political rules. Social divergence and competition between self-interested actors over the spoils of the state’s connections to international capital translates into persistent attempts to tap, define, and redefine social cleavages as a support base for landing a seat in public office and launching credible threats against political opponents.

3 Created in 1926 under the demands of colonial governor Sir Donald Cameron, the Legislative Council was originally composed as an organization for the representation of European interests in Tanganyika. Throughout the course of several decades, the Legco was transformed into a fully elected body (MacDonald, 1966) based on 108 total of ten constituencies would select one African, one Asian, and one European to sit on the controversial multiracial Council. Nyerere, who was under the threat of a prison sentence due to a libel conviction for authoring a dissenting publication printed in the Voice of TANU, managed to escape confinement in order to fight for a sweeping TANU victory in the 1958 election.

Nyerere’s work was not in vein. In a contest between three political parties, TANU won a stunning 29 out of 30 seats. Of the 29 TANU victors, 5 were appointed to sit on the 12 member cabinet, dominated by the British colonial governor (Maguire, 1969, p. 339).

The contest for the all important 1960 election, which were contests for an majority elected and expanded Legco rather than a majority appointed one, would apportion the participants in the negotiations with Britain over the Marlborough House independence Constitution (Baregu,

1997, p. 52). This landmark election was an even more impressive victory for TANU. The party secured 70 out of the 71 seats, with 82.8% of the votes cast (Bienen, 1970, 56)4. The victors of the election would also eventually sit on the first National Assembly after independence. Table

5.01 illustrates TANU’s overwhelming dominance in the Tanganyika’s final election under colonial rule.

Table 5.01: Results of the 1960 Legislative Council Elections in Tanganyika Votes Percent Tanganyikan African National Union (TANU) 100,851 82.9% African National Congress (ANC) 337 0.3% Independent Candidates 20,527 16.9% Total 121,715 100% Source: Bienen, 1970, p. 56

TANU’s dominant position grew even more evident in the years immediately following independence. Maguire (1969) boldly, but accurately claims that, during the 1960s, “TANU

a ‘multiracial’ tripartite voting system that guaranteed representation to Asian, European, and eventually Africans (Bienen, 1970). 4 The sole TANU outsider, Mr. Sarwatt, was a former TANU candidate in Mbulu, but was rejected by the party’s national headquarters. Soon afterwards he was elected to the Legco on an independent ticket, before he finally rejoined TANU (Bienen, 1970).

109 could boast of having one of the most efficient, effective and united popularly based, nationalist party organizations on the African continent” (p. 338). To be sure, in the 1962 multiparty presidential elections, Julius Nyerere soundly defeated his closet rival, Zuberi Mtemvu, with

99.2% of the votes cast5. In comparative perspective, the relative vote gathering capacity of

TANU in early elections is most clearly evident when comparing the results of early elections across a larger sample of African countries. Figure 5.01, which depicts the percentage of the votes cast for the winning parties, mirrors Maguire’s sentiments. Out of all the countries that gained independence in the 1950s and 60s, TANU hegemony is clearly matched only in three other cases: the Union Progressiste Sénégalaise (UPS) in Senegal, the Mouvement d'Évolution

Sociale de l'Afrique Noire (MESAN) in the Central African Republic, and the Parti

Démocratique de Guinée (PDG) in Guinea.

Figure 5.01: Percentage of Seats Won by Largest Party in Early Elections

Source: African Elections Database, 2007

5 A note of caution should be directed toward overestimating the vote getting capacity of TANU in these early elections. In absolute terms, TANU’s ability for mobilizing votes throughout rural Tanzania at the time of independence was quite limited. These limitations were clear in the 1962 presidential elections, where less than a quarter of the potential electorate actually voted.

110

The life span and cycle of opposition parties in Tanganyika was undeniably short and unimpressive. In the final years of colonial rule, TANU’s emerging hegemony was challenged in the Council elections by a collection of political notables and interests with no clear popular base. The United Tanganyika Party (UPT), a staunch advocate of the multiracial power sharing design of the Legco, was formed in 1957 by European and Asian business persons and chiefs, and with the assistance of colonial sponsorship from Governor Edward Twining6. By 1959, the

UPT was nowhere to be found. The African National Congress (ANC), a staunch advocate for rapid Africanization of state institutions and major business, was formed in 1958 by a former

TANU assistant secretary general Zuberi Mtemvu7, who broke away from TANU due to the party’s lack of stern opposition to Asian and European representation (Mmuya & Chaligha,

1994; Okema, 1996). By 1963, Mtemvu had rejoined TANU and the ANC vanished from the political scene. Other parties, such as the People’s Democratic Party (PDP), led by an ex-TANU leader and railway union head named Kasanga Tumbo, the People’s Convention Party (PCP), founded by J. D. Chipaka8 among others, and the Nationalist Enterprise Party (NEP), headed by

Sheikh Yahya Hussein9, rose and fell just as quickly as the UPT and ANC. The plight for all of these parties was their inability to cultivate ties to well-funded patrons and sub-national groups

6 The UTP was also composed of a bloc representing settler interests. However, since the UTP backed a multiracial power sharing plan, the staunchest racists within the white settler community found the party too accommodating toward the interests of Africans (Baregu, 1997).

7 Zuberi Mtemvu, who was a TANU provincial secretary prior to the creation of the ANC, disagreed with Nyerere’s more open and conciliatory approach to racial politics. For Mtemvu, who preached radical African nationalism, governance in Tanganyika should have been left totally to Africans. He rejected any talk of opening TANU membership to Asian and Europeans. The agenda of the ANC reflected these sentiments by embracing a rapid Africanization of the civil service and racialized criteria for citizenship (Bienen, 1970).

8 Chipaka was a founding member of the ANC prior to forming the PCP.

9 After the failure of the NEP, in 1963 Sheikh Hussein went attempted to form a broad opposition front called the African Independence Movement. The AIM was an ambitious, but failed attempt to bring together a diverse array of some of the most notable opposition leaders, including Chipaka, Tumbo, and Fundikira, under a common political platform (Beinen, 1970). 111 with strong senses of corporate identity, a point elaborated in the subsequent sections of this chapter. Most arose out of interests in preserving the institutions of chieftaincy, Africanization, or loose religious distinctions and typically reflected a “kaleidoscope of interpersonal attachments and alliances of various spokesmen” (Bienen 1970, p. 59).

The relative weakness of these opposition parties is more fully appreciated in a comparative perspective. In Kenya, the Kenya African National Union (KANU) clearly emerged as a decisive victor in pre and post-colonial elections. However, inter-party competition at the time was far more intense than what was found in Tanganyika. Although the Party’s position would later change, KANU initially appeared with a more left-leaning agenda, one staunchly opposed to colonial and settler interests. By contrast, the Kenya African Democratic Union (KADU) emerged as a more conservative adversary, sometimes aligning with the settler interests of

Michael Blumdell’s New Kenya Party (Barkan, 1994; Hyden, 1994; Throup & Hornsby, 1998).

Table 5.02: Results of the 1961 Legislative Council Elections in Kenya Votes Percent Kenya African National Union (KANU) 557,415 63% Kenya African Democratic Union (KADU) 141,566 16% Others 185,805 21% Total 884,786 100% Source: Adar, 1999

Table 5.02 depicted above shows that, while KANU clearly dominated the last election to be held under colonial authority, KADU gave the victor a far stiffer challenge than the more radical

ANC did to TANU. Whereas TANU managed to scoop all but one of the Council seats in the

1960 election, KANU secured 19, while KADU seated 11 candidates. In the May 1963 elections for the House of Representatives, just eight months prior to independence, KANU secured 83 out of 124 seats. KANU’s victory margins might be impressive when compared to the victory margins of Francisco Macías Nguema’s 36 percent in the first round of the 1968 presidential

112 elections in Equatorial Guinea or the 1962 victory of the Sierra Leone People’s Party (SLPP) based on a mere 35 percent of the votes. Still, KANU’s relative strength pales in comparison to

TANU.

In present day Zambia, formerly the colony of Northern Rhodesia, the United National

Independence Party (UNIP) emerged from the pre-independence election competition and the struggle for independence in a marginally better position than KANU. While UNIP and the

African National Congress (ANC) formed a legislative coalition out of the highly competitive

1962 Legislative Council election, UNIP managed to score 94 percent of the votes in the 1964 election, securing 55 out of 65 Council seats. However, in post-independence Zambia, UNIP was faced with persistent challenges from the ANC, the United Party (UP), and the United

Progressive Party (UPP) until the formal proscription of opposition parties in 1972. While UNIP won slightly more than 73 percent of the vote in the 1968 elections (see table 5.03), the African

National Congress won 23 seats, compared to UNIP’s 81 seats.

Table 5.03: Results of the 1968 National Assembly Election in Zambia Votes Percent United National Independence Party (UNIP) 657,764 73% African National Congress (ANC) 228,277 25% Independents 12,619 1% Source: African Election Database, 2007w

At the same time, political contention grew more turbulent while the state grew more repressive. The government banned the UP after a series of clashes ensued between UP and

UNIP supporters in the Copperbelt (Tordoff, 1988; Burnell, 2001). After the UPP threatened to make inroads into UNIP vital support bases in the Copperbelt, and the escalation of violence during the 1971 by-elections (Gertzel & Szeftel, 1984), a single-party state was finally promulgated in December, 1972.

113

While national unity may have been measurably higher in the lead-up to independence in

Malawi, a combination of political manipulation and ethnic and regional saliency rendered the months following independence highly turbulent. Despite Banda’s calls for greater levels of centralization, the Nyasaland African Congress (NAC), the MCP’s forerunner, was an ineffective federation of sorts to the extent that it was “stricken” with those ethnic divisions it

“purportedly sought to obviate” (Tangri, 1968, p. 40). While the MCP dominated pre- independence elections, against the few notable contestants, such as the United Federal Party and the Nyasaland Constitutional Party (NCP), the calls for unity in the lead-up to independence gave way to political fragmentation along ethnic lines, a post-independence leadership crisis, and the ascendancy of Banda’s personalistic rule as the modality for generating cohesion (Ross,

1967).

The analysis in the following paragraphs will more specifically demonstrate that the fundamental differences between these cases at the time of independence lay in their respective differences in macro-structural divergence. These differences presented political entrepreneurs at the time of independence with varying capacities for mobilizing constituents and building political parties. Let us now comparatively explicate this macro-structural divergence in

Tanzania and relate it to party formation.

The Colonial Background

Administered as part of Deutsch Ostafrika (German East Africa) until 1919, the territory officially named Tanganyika by the British in 1920 was originally annexed by Germany in

1885, following the infamous Berlin Conference. In 1890, Britain and Germany signed an agreement recognizing Zanzibar as a British Protectorate and while Germany was recognized by

114 the British as the colonial authority in German East Africa. In 1891, the German government in

Berlin officially took over the direct administration of the Country’s East African colony10.

Colonial capital, paired with forced labor and years of military campaigns designed to insight “frightfulness” (Schrecklichkeit) spawned the growth of profitable coffee, tea, and cotton plantations throughout parts of Tanganyika (MacDonald, 1966). Local economies were more tightly linked to regional and global market demands, as agricultural products were carried along the newly laid rail lines stretching from Tanga to Arusha and Dar es Salaam to Kigoma. In the process, massive swaths of land was alienated from local control, African males were forced into labor with little or no compensation, while German investors and administrators profited enormously from the ability of the colonial state to discipline local production.

Colonial conquest and subjugation was however, periodically contested by popular protests.

During 1888, Arabs along the coastal areas in northern Tanganyika, led by Bushiri bin Salem el

Harthi, revolted against encroaching Christian dominance and German attempts at annexing the last remaining enclaves under local control (MacDonald, 1966; Mbogoni, 2004). The Hehe in

Iringa, the Chagga in Kilimanjaro, and the Gogo and Tao in the south each at various times launched intense armed resistance campaigns to forced labor and the confiscation of land.

When, in 1902, the colonial government forced villages to establish communal farms and grow cotton, Tanganyikans from various localities revolted in what is famously known as the Maji

Maji Rebellion (1905-1907)11.

Inter-imperial rivalry would again shape the contours of colonial rule, when German East

Africa was divided up and administered by two different European powers following Germany’s

10 Initially, German control over its East African possessions was administered by the Gesellschaft fur Deutsche Kolonisation (Society for German Colonization). 11 The Maji Maji Rebellion has since, occupied an important place in Tanzanian history, being portrayed by many political leaders at the time of independence in the early 1960s as a symbol of the first stirrings of Tanzanian nationalism. 115 defeat in World War One. Ruanda-Urundi, which was separated and renamed Rwanda and

Burundi after independence, was administered by the Belgians. To the east, Tanganyika was placed under British control, as a League of Nations Mandate territory. After World War II,

Tanganyika was converted to a United Nations Trust Territory, but continued to be administered under the authority of Britain. It was during the period of British colonial rule, which at times encouraged the development of local self-help societies, where TANU’s forerunner, a self-help organization called Tanganyika African Association (TAA), was born.

Throughout most of colonial Africa, the initial opposition to colonial rule was limited to the emerging class of urban professionals and civil servants, most of which were educated in

European institutions. Thanks to the ability to form self-help organizations, by the 1920s and

1930s this growing bourgeoisie pressed for more adequate social services, and greater employment opportunities, while increasingly serving as advocates against cases of land confiscation.

At the same time, the expanding colonial legislative assemblies, when paired with an increased determination among self-help organizations to wrestle state power away from

European powers, acted as the institutional and interest mixture necessary for expanding the scope of conflict from urban society into rural society by networking with local cooperative societies and ethnic associations. Given the necessity of competing for legislative seats, and the fact that so many elites were relatively well versed with political institutions in Western societies, political parties served as the logical organizational means for mobilizing support for combating colonial rule. These urban to rural self-help networks surfaced as a readymade basis for building capable political parties.

116

While struggles for legislative seats provided the competitive arena for expanding political party activities, two additional factors would transform this competitive arena into one favoring the emergence of party dominance. First, the most salient concern shared throughout African societies was the prospects for terminating colonial rule and establishing majority rule. Given these shared concerns, unity within the framework of a dominant party was not only relatively easy, but seemed utterly necessary in the struggle for independence. Secondly, academics wedded to systems theory, as well as Africa’s emerging urban elite, conceptualized political parties as having distinct nation-building functions rather than functioning specifically as a collection of organizations representing the competing interests of sub-national groups. These functions included articulating a nationalist challenge to colonial authorities and assuming control of the state apparatus as a protector and integrator of a national whole which was otherwise highly divided between mutually hostile sectarian interests (Kilson, 1963; Colman &

Rosberg, 1970; Salih, 2001).

This ideal of an all inclusive ‘national whole’ again represented the ‘modernist’ world view within the Western educated urban elite, one that equated the existence of a state with the existence of the nation (Zolberg, 1966). The fact that the reality on the ground was far different from the aspirations of this model was quite clear in the eruption of inter-ethnic tensions following independence. In what Joel Migdal (1988) characterizes as “strong societies, weak states”, African leaders quickly understood the limits of the state’s capacity to enforce the national integration agenda, amid powerful sub-national constituents who understood the state as at best an institution for advancing personal ambitions or group interests, and at worse, as a

“threat to their income, their autonomy, even their lives” (p. 14). Stressing the Eurocentricism of the nation-state concept, Shivji notes (1976) that “the slogans of nationalism and freedom,

117 equality, etc. that the petty bourgeoisie shouted on the eve of independence were merely echoes of the ideology of the metropolitan bourgeoisie without their social or economic content” (p.

20). While some African leaders such as Kwame Nkrumah in Ghana or Léopold Sédar Senghor in Senegal, genuinely believed that their respective political parties were organizations with the capabilities to shape a popular national sentiment, emerging single-party rule more clearly functioned to suppress rivalry from competing petty bourgeoisie throughout the countryside

(Collier, 1982). As state leaders grew increasingly insecure about their capacities to carry out nation-building functions or ensure their own tenure, challengers to state encroachment were increasingly harassed, arrested, exiled, or killed. At the ideational level, oppositionists were incorporated into nationalist discourses as a mechanism for shoring up incumbent support when labeled as “illegitimate” and “traitorous” elements aligned with foreign powers wishing to bring down self-governance (Kilson, 1963, p. 266).

The history of TANU generally followed the development path outlined above. Similar to

KANU in Kenya12 or UNIP in Zambia13, TANU in Tanganyika was an outgrowth of a welfare association which expanded its networks with other more rural associations. Established in Dar es Salaam in 1929 by Sir Donald Cameroon, a more progressive British governor by the standards of the time, the Tanganyika African Association (TAA) provided an assortment of social services for blacks, and later acted as a forum for political discussion (Bienen, 1970). At the same time, ‘tribal’ unions consisting of African traders, farmers, and government employees, sprang up throughout the countryside. Similar to claims found in most other colonial territories

12 The origins of KANU can be directly traced to the formation of the Kikuyu Association, Kikuyu Central Association, the North Kavirondo Central Association, the Ukamba Members Association, and eventually the Kenya African Association (Spencer, 1985, p. 2).

13 The origins of UNIP can be indirectly traced to 1946 formation of the Federation of African Welfare Societies.

118 in Africa, these rural political and economic organizations in Tanganyika arose in opposition to the Native Authority System14 and colonial intervention in agricultural and land tenure

(Agonafer, 1994).

After the Second World War, the TAA rapidly made inroads into the more remote quarters of Tanganyika, capitalizing on peasant rejection of colonial agricultural policies and the British

Native Authority system, which empowered seemingly unaccountable chiefs, subordinate chiefs and headmen as agents of rural control. Additionally, the TAA began to push for restricting

Asian access to rural markets, hoping instead to empower Africans as the primary village traders

(Shivji, 1976). With expanded political activities, alliances with local ethnic associations and cooperative societies were formed, which, by the 1950s, often functioned as local TAA branch offices (Hopkins, 1971).

Data on TAA membership illustrates the rapid expansion of the TAA. In 1940, British authorities reported that the TAA had only 100 members. The only TAA office outside of Dar es Salaam was located in Dodoma, in central Tanganyika. By 1948 however, a United Nations visiting mission was informed by TAA officials that the organization boasted some 39 branches and 1,780 members. When a second mission visited Tanganyika in 1951, the TAA reported to have some 5,000 members and branch offices in nearly every town (as cited in Bienen, 1970, p.

28).

While the TAA networks were expanding, more politically astute and active Africans, such as Julius Kambarage Nyerere and Oscar Kambona, started to shape the agenda of the TAA.

With more active members and leaders, the organization’s normal routine increasingly tackled

14 Native authority ordinance of 1923 and Native Courts ordinance of 1929 established in Tanganyika what came to be known as “indirect rule”. The Native Authorities included chiefs and petty chiefs or village headmen, who were charged with executing the rules made by the colonial administration (Bienen, 1970).

119 issues of discrimination, land alienation, and more frequently engaged in political and legal action against colonial politics (Hatch, 1976). The Meru Land Case15 and the apparent intensions of the British colonial authorities to link Tanganyika with the East African

Federation16 surfaced as monumental rallying issues for the TAA. By July 7th, 1954, a day known as Saba Saba (seven seven), the TAA’s ascendance to the status of political party was finalized when its name was changed to the Tanganyika African National Union (TANU).

Despite energetic efforts at party-building, TANU’s capacity to successfully contest elections and shape politics throughout the countryside was still quite limited. Except through networks of personal friendship and kinship, urban Party leaders had no clear means of exerting effective control over rural party leaders throughout much of rural Tanzania (Hyden, 1980).

With the state still being administered by the British, and a commercial sector dominated by the

Asian community who were less than sympathetic to the aspirations of TANU, the urban elite had no real economic base to effectively stitch together a rural/urban party network.

Throughout Africa, similar challenges of effective control were especially troublesome for rulers with the air of western education and the goal of nation-building and economic development. Whether one talks about a disjunction in the language of modern versus traditional values (Lerner, 1958), the mobilized versus the non-mobilized (Deutsch, 1961), or the modern

15 The Meru Land Case grew out of an effort by the British to confiscate land from some 3,000 African farmers in order to make way for European settlements. Due to the Trusteeship status of Tanganyika, and the determination of activists within the TAA, the Meru Land dispute made its way to the United Nations. A Meru Citizens Union leader named Kirilo Japhet Ayo became the first Tanganyika African to address the UN (Bienen, 1970; Agonafer, 1994). When Japhet returned to Tanganyika, he was given a warm welcome by the TAA, who understood land alienation and Japhet’s trip to New York as important opportunities for uniting Africans around a common political cause. With the assistance of TAA leaders, the Meru leader toured throughout the countryside to give publicity to the Meru Land dispute (MacDonald, 1966).

16 In 1947, a Central Legislative Assembly for East Africa was proposed, which was correctly seen by TAA leaders as a step toward linking Tanganyika to the East African Federation. The fear for many Tanganyikans was that, as a member of this Federation, the political and economic agenda of the colony would be dominated by the sizable white settler population in Kenya, which could have slowed the pace toward independent majority rule (Bienen, 1970). 120 versus residual (Zolberg, 1966), a critical urban-rural “bifurcation” emerged in nearly every

African society, each section having its own sets of institutions and practices (Mamdani, 1996).

The challenge for new states was to somehow reconcile this bifurcation.

From a comparative perspective however, this challenge was somewhat unique in Tanzania.

Unlike many of Tanzania’s neighbors, where rural petty bourgeois elements found themselves in well established positions for constituting political movements, the rural petty bourgeoisie in

Tanganyika simply lacked the material resources or ideational basis for constituting serious political movements. Hence, the expansion of the TAA was far less likely to face the centrifugal pulls associated with the dilemma of scale, nor was it likely to face stiff external rivalries from political entrepreneurs mobilizing peasants through the language of ethnic identity. Only among the cooperative unions and ethnic associations in Sukumaland (Maguire, 1969) and Kilimanjaro

(Hopkins, 1971) were there notable challenges to TANU encroachment. At the same time, because the agenda of the urban intellectual elite prevailed, TANU never developed close associations with any particular ethnic group to the extent found among the new parties in the three other cases. The most significant challenge that TANU faced was successfully spearheading the drive to independence and wrestling control of the commercial sector away from the Asian communities, who settled along the coast generations ago and functioned as traders, intermediaries and creditors in relation to the production and consumption of African peasants (Shivji, 1976). Let us now turn to a more detailed account of what exactly TANU was and an analysis of the social structures which enabled TANU to capture state power without a serious fight against domestic opposition.

121

Political Factions and the Rising Party Leadership Stratum

As mentioned previously, independence movements throughout Africa were defined by relatively intense rifts between sections of rural society or between urban and rural society. The formative tensions within TANU were characteristically different, largely because the power of social forces to viably exercise the exit option was constrained by a social milieu less conducive to mobilizing material and ideational resources. This did not however, translate into a party without internal divisions. Instead, intra-party rifts, some of which gave rise to opposition parties like the ANC or the PDP, were quite intense and centered around internal divisions within the urban petty bourgeoisie and between sections of the urban petty bourgeoisie and organized labor. In the previous chapter, I suggested that these rifts were patched up coercively.

In the next chapter, I will show in more detail the degree to which factional leaders were accommodated within the single-party framework. What is important in this chapter is the provision of a brief overview of the issues underlying these divisions, and the manner in which the macro-structural legacies of colonial rule, along with a little bit of luck, shaped their respective capacities for action.

Taking a cue from Okoko (1987), I divide the petty bourgeoisie in Tanganyika at the time of independence into three factions: 1) those hailing from the cash crop regions which were interested in a conservative approach to economic development following independence, 2) those which constituted an urban-based professional class of senior civil servants and traders who were interested in dominating the state and commercial sector, and 3) those leftist ideologues Okoko appropriately dubs as the “leadership stratum” of TANU, which were committed to an assortment socialist principles. While these categories are highly fluid, especially the latter two, together they constituted the bourgeoisie component of TANU at the

122 time of independence. During the later years of colonialism, this bourgeois coalition had two common interests: to end colonial rule and to assume leadership over the commercial sector.

A second factional tension emerged between the leadership stratum and organized labor, which increasingly pushed for commitments to Africanize the economy and the workforce in general, and most importantly the civil service positions within the state. At the same time, for some senior and middle-level party officials, ending colonialism also meant ending Asian domination of the commercial economy. As Shivji (1976) points out, for shopkeepers and commercial farmers in particular, this “called for a struggle against the colonial state” that granted the privileged position of this commercial class (p. 49). In short, there was a convergence of interests between non-leftist elements of the petty bourgeoisie and labor over the need to Africanize the economy.

As it turned out, the leadership stratum would emerge as the real shakers of TANU’s agenda, versus those with bourgeois impulses that would eventually emerge as a bureaucratic bourgeoisie with the state as its material base (Okoko, 1987). The dominant position of the leadership stratum stemmed from several sources. First, one cannot ignore the level of charisma commanded by Nyerere himself, an allure which translated into a colossal asset when competing for peasant votes. Secondly, the biggest advantage for the ideologues laid with the disadvantages faced by labor and the remaining elements of the petty bourgeoisie, namely their respective inabilities to mobilize resources compared to those party leaders who commanded popular support and increased backing from the British in the drive toward independence. It is this second advantage which makes Tanzania unique. In neighboring countries, rural and urban bourgeois political entrepreneurs were powerful enough to assert themselves as contenders against the ideological ambitions of leftist bourgeois intellecturals. In Tanganyika, the

123 peripheral nature of the colonial economy denied these social forces the material and ideational bases for mobilizing sizable popular followings and building sustainable party organizations as an obstacle to the ideological ambitions of the leadership stratum.

The Structural Underpinning of the Rising Leadership Stratum

While the expansion of regional trade in Africa was taking place for centuries prior to the dawn of colonial rule, colonial conquest and the colonial state emerged as a major stimulus for social transformations in order to link productivity in Africa to the consumption demands of the metropole. Whether the colonial authorities were guided by class interests or racial and cultural senses of moral superiority, the net effect of land confiscation and consolidation, land and head taxes, and education for modernization to name a few, widened asymmetries in resource distributions and swelled the ranks of migrant plantation and mine wage workers. In southern

Uganda, the British encouraged the consolidation of the freehold system into larger, indigenously owned estates for the production of cotton and coffee (Karugire, 1996). In Zambia, the production of copper in the Northern Province after the 1920s contributed to the emergence of a sizable class of wage earners working in the mines, while land alienation for commercial farming in the Southern Province produced a well of agricultural laborers employed on medium to large scale farms (Gertzel, Baylies, & Szeftel, 1984). Similarly, in Kenya, the confiscation of land for the development of European owned cash crop plantations in the “White Highlands” contributed to the growth of a large landless African labor reserve (Agonafer, 1994).

Colonies such as the Bechuanaland Protectorate (Botswana) and Tanganyika on the other hand, were established and maintained for balancing against colonial aggression from European rivals and acting as peripheral supply regions for more lucrative production in neighboring

124 colonies (Agonafer, 1994). Administered with the ultimate goal of transferring authority to

South Africa, “absurdly low levels of revenue and lack of consistent policy of internal development” characterized British rule in Bechuanaland (Picard, 1987, p. 97; Du Toit, 1999, p.

196). In Tanganyika, the growth of export oriented cash crop farming was much more limited when compared to Southern Uganda and Kenya, and often involved free-holding peasants rather than tenant farmers and large scale plantations dominated by white settlers17 (Hyden, 1980).

Only in the north-western highlands, the western plateau, in the foothills of Kilimanjaro, and along the central rail line, was the production of cash crops, such as coffee, cotton, and sisal, actively promoted prior to independence (Buchert, 1994)18. Therefore, Tanganyikan politics at the time of independence was, comparatively speaking, devoid of an “African yeoman farmer class” as an outgrowth of land consolidation and commercial farming (Bienen, 1970, p. 270).

Whereas the relatively strong economic base derived from agricultural production in Kenya translated into a post-colonial state dominated by an urban bourgeoisie allied with a particularly powerful ethnic section of rural petty bourgeoisie, in Tanzania, the economic base of the rural petty bourgeoisie was far too weak and fragmented to challenge the rising tide of the party

17 When compared to some of its neighbors, such as Kenya, Zambia, Mozambique, and Rhodesia (Zimbabwe), the settler community in Tanganyika was quite weak. First, the relative size of the settler community in Tanganyika was small in number. The Germans never actively encouraged permanent settlements in German East Africa, and the ability of the British to launch massive land alienation campaigns that encouraged the growth of white settlements was constrained by its obligations to the Mandate System (Agonafer, 1994). Secondly, the European population in Tanzania was a heterogeneous bunch, including Greeks, Swiss and Italians that moved into the territory to purchase German farms that were auctioned off after the World War I (Hopkins, 1971; Hyden, 1980). Finally, Tanganyika existed as a Trust territory under the U.N. Trusteeship Council. In comparative perspective, this status constrained the ability of the British to conduct widespread land alienation campaigns and generally strengthen international attention toward national self-determination in Tanganyika (Bienen, 1970; Baregu, 1997).

18 As Hyden (1980) points out, the production of cash crops in Tanganyika was, as promoted by the Germans, a mixture of free-hold peasant production and plantation production. In Kenya by contrast, cash crop production was often taken up by the white settler community on large plantations, as the owners of land, and by Africans as wage workers. In fact, in parts of Kenya, peasant production of cash crops was outright illegal (p. 41). 125 ideologues or serve as a viable partner among conservative elements of the urban petty- bourgeoisie (Shivji, 1976).

The lack of land consolidation also meant that the dislocation of peasants for employment on cash crop estates was comparatively low. As Freund (1981) notes, “even in intensely capitalized and crowded Kilimanjaro, the rural poor have tended to become very junior clients rather than landless workers” (p. 492). Economic relations were based upon exchange between producers, and therefore “peasant consciousness is in part that of the petty commodity producer” (p. 492), rather than as a subordinate wage worker. Therefore, when compared to the well of workers in the copper mines of Zambia or the plantations of Uganda and Kenya, the basis for the rise of a powerful proletariat in Tanganyika with the ability to define TANU’s policy ambitions was comparatively weak.

These class developments had noteworthy implications for the role that regional and ethnic loyalties would play in Tanzania’s post-independence polity. In the typical fashion throughout

Africa, the penetration of the colonial state, the introduction of new market forces, and the increased division of labor, accentuated political competition and reconstituted the boundaries of those local identities which consisted of overlapping loyalties based on kinship, tradition, and practice (Rasmussen, 1969). Colonial administrative techniques seeking to institute territorially- based ‘tribal’ societies, paired with increased competition over access to markets and the colonial state, gave rise to more territorially bound, and ideologically rigid partitions between self and other, often described in the language of ethnicity, tribe, or race (Migdal, 1988). When writing on Burundi for example, Lemarchand (1994) notes that the fluidity within the pre- colonial stratified kinship networks was undermined by the “discovery” of a hardened “Hutu-

126

Tutsi dichotomy” due to colonial differentiation and rapid growth of political and economic competition (p. 14).

At the onset of struggles for independence, these vertical hierarchies of ethnic or racial divisions of labor, bounded only in common by a singular repressive territorial state intervening between local production and the international market, factionalized and polarized the emerging nascent petty bourgeoisie, resulting in violence, repression, and coups. In Ghana (formerly the

Gold Coast), the Akan speaking Ashanti in South Central Ghana accumulated enormous returns from mineral mining and cocoa cultivation during the colonial years. Paired with the reproduction of British dominance through the sowing of hostilities between Ghana’s pre- colonial kingdoms, ethnic divisions of labor were a major source of tension which erupted into the 1966 military coup once the drop in cocoa prices undermined clientelistic glues.

The consequences of economic and ethnic tensions were even more extreme in Uganda, where economic and political privileges accrued by Bantu speaking Kingdom of Buganda in the south solidified sharp social divisions and lasting mistrust between the southern Kingdom and the Bunyoro Kingdom in the western portions. These divisions were even more complex in that both the Bunyoro and Buganda were in uneasy relations with the more fragmented Nilotic speaking groups in the northern parts of the country. The abrasiveness of differentiating discourses rendered post-colonial coalition-making a seemingly impossible undertaking, temporally circumvented by Obote’s exclusionary rule prior to his removal by Army Chief of

Staff Adi Amin in the coup of 1971 (Karugire, 1996; Byarugaba, 1998).

Similar processes played out in Zambia. In a 1971 study, Scarritt shows that political elites in Zambia at the time of independence expressed a strong “wariness” toward the destructive potential of ethnic sentiment when mobilized by opportunistic politicians (p. 36). The uneven

127 penetration of capitalism and the application of indirect rule left Zambian politics a game of competing locally and regionally-based notables. By the late 1960s, the politics of nation- building driven by the United National Independence Party (UNIP) and Kenneth Kaunda, began to disintegrate as intra-party factionalism resulted in greater degrees of inter-party competition.

Leaders such as Harry Nkumbula’s African National Congress (ANC), which was dominant in the Southern Province, blended Tonga identity and discontent among wealthy peasants against the lack of provincial development. While the United National Independence Party (UNIP) boasted a broader base, especially among mine workers, the party maintained almost no support in the Southern and Western Provinces.

The regional limitations of UNIP were most obvious in the four 1968 by-elections in the

Southern Province, where, despite heavy state investments in roads, schools, and health clinics, the ANC clinched all four seats and 72 percent of all the votes cast. In the general elections which followed that same year, the ANC swept, with wide election margins, 13 out of the 14 seats in the Southern Province (as cited in Rasmussen, 1969, p. 411-412). Even within the incumbent’s stronghold in the Copperbelt and among mine workers, UNIP’s dominance was threatened in the late 1960s and early 1970s due to a combination of Lozi and labor grievances and the corresponding formation of the United Party (UP) and the short-lived United

Progressive Party (UPP) (Gertzel, Baylies, & Szeftel, 1984; Sichone & Simutanyi, 1996; Du

Toit, 1999). The intra-UNIP fracturing, brought on by the acuteness of the “dilemma of scale” in the context of strong macro-structural divergence, produced another windfall for the ANC after the banning of the UP just prior to the 1968 election. The results of the election in the Barotse

Province, a former UP stronghold, produced surprising victories for the ANC rather than

128 increased victory margins for UNIP, which captured only 3 out of the Province’s 11 seats (as cited in Rasmussen, 1969, p. 415).

Due to comparatively greater levels of migration and employment in colonial civil service, the class of people that would emerge as Malawi’s independence leadership possessed a sense of national togetherness that was perhaps more coherent than what was found in Zambia and

Kenya (Ross, 1967). However, the 1964 cabinet crisis only accentuated ethnic and regional cleavage structures to the point that all of those forced out of Banda’s government were either from the Northern or Southern Regions. The rump government following the cabinet crisis was almost exclusively dominated by Chewa from the Central Region, all loyal to Banda, who built an ethnic support base by appealing to the unjust treatment of the Chewa and the stereotypes of inferiority which legitimated that treatment (Kaspin, 1995). As Ross (1967) states, “Banda decided that he could exploit Chewa feelings of inferiority and exclusion, to create a system of personal political power” (p. 89).

Developments Kenya parralled those which transpired in elsewhere. The appropriation of land during colonial rule in Kenya in the largely Kikuyu occupied Central Highlands, translated into countless numbers of displaced Kikuyu migrants and wage workers. In fact, the Mau Mau uprising, which took place in colonial Kenya between 1952 and 1955, makes for an illuminating comparison to earlier uprisings in Tanganyika. The uprising was essentially a revolt by Kikuyu peasants and squatters in opposition not only to British colonial rule, but also to their Kikuyu chiefly agents and commercial farmers who enforced and benefited from land confiscation

(Spencer, 1985). This popular revolt could have served as a class-based rallying point for a broader peasant movement not unlike that found in Tanganyika at the time of independence.

However, while Kenyatta was jailed and the Kikuyu based KAU was banned, the Lou,

129

Abaluhya, and Kalenjin “remained on the sidelines” due to their reluctance to come to the defense of ‘outsiders’ and fear of British retaliation (Throup & Hornsby, 1998, p. 7).

The Mau Mau uprising highlights two interesting points. First, ethnic lines of conflict in

Kenya became a distinct line for defining the scope of conflict to a far greater extent than one finds in the Maji Maji rebellion in Tanganyika only fifty-years earlier. These partitions would greatly impact the formation of political parties, including the Kenya African Democratic Union

(KADU) among the less affluent groups and the Kenya People’s Union (KPU) among the Lou in Nyasaland. Secondly, the constitution of the Kenyan African Union (KAU) and the Mau

Mau, appeared to many of the smaller and less economically affluent ethnic groups, such as the

Kelenjin, Turkana, Samburu, and Maasai, as an anti-colonial struggle representing the claims of the more affluent groups, such as the Embu, Meru, Luo, along with Kikuyu. Yet, the Mau Mau uprising was appropriated by KANU as a symbol of independence and nationhood, with the consequences of alienating those ethnic groups that were not involved in the revolt. This is in stark contrast to TANU’s use of Maji Maji imagery, which was similarly used as a symbol for rallying national sentiment without evoking sub-national tensions like the ones found in Kenya.

In Tanganyika, the colonial economy neither translated into pronounced regional differentiation nor vertical integration among ethnic groups in economic and political life

(Barkan 1994; Hyden 1994; Maxon, 1994). Only the Chagga and Haya had modicum of economic clout, due to the production of cash crops and the penetration of Christian missions into the north-western highlands and Kilimanjaro (Buchert; 1994). Paired with the previously discussed impotency of rural elites with powerful petty bourgeois interests, there was no real sense of inequality or exploitation which could have otherwise acted as an impetus for the growth and consolidation of regionally or ethnically defined populism. Hence, there were no

130 significant social divisions for weakening TANU as an independence movement or for competing against TANU for peasant loyalties.

In short, the social structural condition at the time of independence in Tanganyika was favorable for the rapid growth of an independence movement led by a left-wing urban-based petty bourgeoisie. First, the number of peasants dislocated due to land confiscation, the consolidation of cash crop plantations, and mining operations was minimal at best. Therefore, labor leaders lacked a strong popular base for successfully exercising the exit option. Secondly, the consolidation of capital among the highly fragmented rural petty bourgeoisie meant that none could become successful political entrepreneurs in their own right, nor could they offer up much to enhance the bargaining position of the urban conservatives. Finally, because colonial divide and rule tactics were somewhat more constrained in Tanganyika, the growth of ethnic- based hostilities among the peasantry was minimal by comparison and thus denied political entrepreneurs an ideational basis for mobilizing support. In a comparison between political developments in Tanzania and West Africa, Goulbourne (1978) points out that “What little opposition there was [TANU] was able to surmount more easily than if there had been a greater degree of social divergence in society of the kind more prevalent in some former West African colonies” (p. 382). Given the minimal degrees of macro-structural divergence, TANU could more easily accommodate divergent interests by expanding clientelistic control without facing serious intra-party centrifugal pressures.

Still, a cautionary note needs to be inserted here. While grievances colored by ethnic discourses have been comparatively absent in Tanzania, this absence should not be construed as suggesting that Tanzania is devoid of ethnic discourses or national-level social cleavage structures. Indeed, references to ethnicity and locality have historically been quite common on

131

Tanzania Mainland and especially Zanzibar. Yet, for reasons described above, these discourses have failed to translate into national-scale partitions with the ability to underpin political party formations.

The social-structural conditions at the time of independence on Zanzibar merit special attention, as the historical processes leading to present multiparty environment on the Isles offers a valuable comparative case against the processes and outcome on the Mainland. As will be shown by the conclusion of my research, the shear level of violence between the main opposition party, the Civic United Front (CUF), and the CCM today, along with the latter’s razor thin election victories, is a product of the inability of the post-independence Tanzanian government, through exclusion and coercion, to resolve long-standing class and racial animosities on the Isles. Let us now take a look at the basis for these long-standing animosities.

Race, Class, and the Crisis on the Isles

While lines of national-level social cleavages in Tanganyika appeared fragmented and localized at the time of independence, the 1964 Union between Zanzibar and Tanganyika brought with it an entirely new line of conflict to the national political arena, and ultimately into the ruling party itself. Given the radical dissimilarity between the historical processes of these formerly separate political entities, the comparatively widespread use of naked repression on

Zanzibar, and the waves of Zanzibari migrants making their ways to the Mainland, the question over the Isle’s incorporation (some say annexation) into a United Republic has evolved into the single-most divisive line of conflict in post-colonial Tanzanian politics. Later, I argue that the current hostilities, repression, and violence that so often accompanies the Union question, is rapidly transforming into one of the single biggest threats to continued CCM dominance.

132

When compared to Tanganyika, the historical constellations of conflict producing independence in Zanzibar in December 1963, and the revolution in January 1964, reflected a complex web of class and race relations that produced a more tenuous, and consequently, a more overtly repressive form of dominance on the Isles. The basis for the rise of this complex web is the clove and coconut plantation economy, which produced a racialized class structure where the surplus value was generated largely by immigrant slave and wage labor from the

Mainland rather than from local smallholders. Furthermore, whereas the main opponents to majority rule in Tanganyika either lost their raison d’être after the termination of colonial rule, or were too numerically insignificant to play the game of electoral competition, on Zanzibar, the main players in the racialized aristocracy remained on the Isles as a continual challenge to the hegemony of the Afro-Shirazi Party (ASP) following the 1964 revolution. Let us briefly detail the economic and political circumstances leading up to Zanzibar’s volatile situation at independence. For the sake of reference, I included a map of Zanzibar, as presented below in figure 5.02.

Prior to the 1800s, most of the migrants to Zanzibar engaged in the trading of slaves and ivory with other parts of East Africa, Arabia, and the Far East. It was however, under the Omani

Arab rule starting in 1832, when Sultan Said of Oman, relocated his capital to Zanzibar, which spawned the rapid development of a highly lucrative plantation economy (Othman, 2006).

Mostly confined to the smaller northern island of Pemba, labor inputs in clove production were based almost exclusively on the work of low wage and slave laborers, who either voluntarily or forcibly migrated from the African mainland. In the typical fashion of colonially sanctioned enterprise development, the capital formation for initializing plantation upstart and expansion

133 occurred through land confiscation, which displaced countless indigenous peasants, many of which survived through employment on clove plantations.

Figure 5.02: The Zanzibar Islands

Source: CIA, 1977

By the end of the 1800s, when the Islands formally became a British protectorate, the Arab aristocracy emerged as the dominant economic class, with a material base in the fields of the plantation that was legitimized among dominant actors through a racist superstructure. The

Prevailing conceptualization among Arabs as well as Europeans toward the African as racially and culturally inferior provided a moral framework for the plantation economy, characterized by a “master-slave” and later “employer-employee” relationship between Arabs and Africans respectively (Mpangala, 2000, p. 58). While slavery was formally outlawed in 1897, a condition demanded by Britain in exchange for continued protection for the Sultanate amid inter-imperial

134 rivalry (Othman 2006), this distinctly racialized class structure ensued throughout the colonial period.

Occupying a social station somewhere between the wealth Arab landowners and African workers, the middle-tiered Arabs and Asians, along with the descendants of Persian immigrants, often identified themselves as the Shirazi. Hailing largely from Pemba, the economic base of the

Shirazi ranged from larger plantation holdings to smallholder agricultural production19. The

Shirazi themselves are often considered the ‘indigenous peoples’ of the Isles since, as a racial and ethnic category, they have lived on the Isles for many generations. As an identity with a clear sense of their belongingness to the Isles, the Shirazi were mobilized in the years prior to independence in opposition to foreign dominance. This included opposition to Arab ties with the

Sultan, the continuation of British protection, and vast majority of those living on the Islands who migrated from the Mainland (Mpangala, 2000). In short, racial categories reproduced class distinctions, where Arab landlords were at the top, followed by Asian business owners and

Shirazi smallholders. At the bottom were African slaves, squatters, and domestic servants

(Mmuya & Chaligha, 1994).

Amid these differentiated structures, the consolidation of an independence movement on

Zanzibar with a distinctly unified national character as seen in Tanganyika was compromised.

When combined with large doses of political opportunism, the likelihood of any unified national movement to assume dominance over the reigns of political authority without inciting a violent contestation from opponents was greatly reduced. Civil society organizations, which included the Arab Association, the African Association, the Shirazi Association, and the Indian

19 The Shirazi identify themselves as decedents of the wave of eighth century Persian migrants, who fled from the Arab Islamic empire following the fall of the Persian Sassanian Empire and settled alongside the existing populations of the Wahadimu and Watumbatu on Unguja and Pemba.

135

Association, were outgrowths of the violent and oppressive class and racial struggles which characterized Zanzibar society. Political parties in turn, were outgrowths of these ethnic and racial welfare associations (Mpangala, 2000).

By the late 1950s, several parties competed for seats in the Zanzibar Legislative Council and aspired to lead the country to independence. The Zanzibar Nationalist Party (ZNP) emerged out of the Arab Association and was spearheaded by the political aspirations of the Arab aristocracy. Led by Sheikh Ali Muhsin Barwani, who was a staunch nationalist from a leading

Arab landowning family and a close follower and friend of Nasser, the party arose in 1955 with enormous access to wealth and international support. The main competitor of the ZNP was the

Afro-Shirazi Party (ASP), led by an Unguja native and former seaman named Sheikh Abeid

Amani Karume. As the name suggests, the ASP grew out of the Shirazi and African

Associations and represented the African peasants and Shirazi villagers in opposition to the

Arab aristocracy (Clayton, 1981). Together, these were the two major players giving political momentum to independence.

During the later years of the 1950s, each party made some attempts at constructing a broader nationalistic agenda by which a movement toward independence could be anchored. For the

ZNP, an attempt was made to use religion as a common rallying point when it called for the formation of a multiracial Muslim state as a replacement to the now vulnerable British colonial state (Clayton, 1981). Both the ASP and the ZNP formally agreed to piece together a common coordinating body for national independence, although actual cooperation was a rare phenomenon. Just prior to independence, Nyerere reflected in the lack of cooperation on the

Isles when he stated that “politically the (Zanzibar) parties all agreed to one objective but they opposed each other because of race” (as cited in Othman, 2006, p. 38).

136

It was through this economic and political competition, one colored with racial discourses, that sharpened class demarcations on the Isles in the lead-up to independence (Clayton, 1981).

As Ayang (1970) points out, the political parties themselves were vehicles “for disseminating age-long ethnic hatred” and this tension “grew deeper and uglier especially during an election campaign” (p. 88). The combative verbal exchanges between the parties capture much of the racialized class tensions channeled through national-level political institutions. For the ASP, the strong distaste for Arab aristocracy was reflected in the slogan “This Land is Ours, the Trees are

Theirs”, a reference to the crops on the clove and coconut plantations (as cited in Mmuya &

Chaligha, 1994, p. 41). In 1962 Karume bluntly stated at “We therefore now clearly state that no government will be acceptable to the majority of the people unless that government is predominately African in character” (as cited in Mpangala, 2000, p. 60). For the ZNP, the ASP represented the interests of a societal underclass and dominated by Africans in Tanganyika

(Othman, 2006). Prominent ZNP leaders repeatedly uttered unguarded “anti-European and anti-

African” themes as a tactic for shoring up their political base, especially under the competitive pressures of approaching elections (Clayton, 1981, p. 40).

The racial epithets between leaders of the ASP and the ZNP were reflected in the performance of the elections in the late 1950s and early 1960s, as the competition between the two parties was tight. While the ASP managed to secure 5 out of the 6 seats in the 1957 election, conceding the 6th seat to a Muslim League, the “fraudulent elections” of 1961 and 1963 (as cited in Mmuya & Chaligha, 1994, p. 41) showed just how tenuous the position of the ASP was.

The ASP’s fragility stemmed from two sources. The first source lay within the demographic characteristics of the Isles. As reported by Maliyamkono & Kanyongolo (2003), the Shirazi population, if united, would outstrip the combined Arab and African population. Some 148

137 thousand Shirazi inhabited the Isles, compared to 51 thousand Africans and 45 thousand Arabs

(p. 138). Secondly, even more strenuous for the fragility of the ASP, the similar economic base of the Pemba Shirazi and the Arab aristocracy yielded sufficient convergence of interests for making the ZNP a viable coalition partner for some elements of the Shirazi (Mmuya & Chaligha

1994).

In 1957, these two circumstances were brought to bear on the ASP when the landowning

Shirazi on Pemba, led by Mohamed Shamte, Ameri Tajo, and Ali Sherif, withdrew from the party. The basis for the defection reflected the tensions between those Shirazi who considered themselves as indigenous Zanzibaris and the rightful owners of any nationalist discourse, versus

Africans, who were identified by the Shirazi as hailing from the Mainland and thus foreigners on the soil of the Isles (Clayton, 1981). Indeed, the ASP did enjoy close ties with the TAA and

TANU20, which the Shirazi on Pemba outright rejected and used as evidence of an African conspiracy against them. But, rather than defect into the Arab dominated ZNP, the Pemba faction formed their own party, the Zanzibar and Pemba Peoples Party (ZPPP).

What ultimately saved the ASP from total electoral annihilation was the internal split with the Shirazi community, whose political orientations were divided by differing experiences under the racialized aristocracy. Given the “abundance of arable land with less land alienation” and the

“predominance of smallholders rather than large, slave-tilled plantations,” colonial rule on

Pemba was far less “polarizing” than the experience on Unguja (Maliyamkono & Kanyongolo,

2003, p. 136). On the larger island of Unguja, the Shirazi experienced land deprivation, poverty,

20 The Zanzibar African Association (ZAA) operated as a branch of the Tanganyika African Association (TAA). The actual merger between the ZAA and the Shirazi Association was taken under the advice of Nyerere. During the elections in the late 1950s and early 1960s on Zanzibar, TANU members such as Bibi Titi Mohamed and Ali Mwinyi traveled to the Isles and openly campaigned for the ASP (Mpangala, 2000).

138 and exploitation not unlike the hardships experienced by Africans. Thus, in the Unguja context, the two communities shared similar political aspirations (Maliyamkono, 2000).

As a consequence of the class fissure within the Shirazi community, in the 1961 and 1963 elections the share of the votes for the ASP, which stilled touted an enormous following on

Unguja, out stripped the individual vote totals of the ZNP and ZPPP. With a mere plurality however, a coalition between the two spelled trouble for the ASP. The ZNP, which enjoyed a sizable breadth of support on both islands, found the concentrated pockets of support for the

ZPPP on Pemba an attractive way for tapping into some of the most energetic communities on the Isles. As clear from table 5.04, in coalition, the ZNP and the ZPPP managed to eclipse the

ASP in the Legislative Council.

Table 5.04: Pre-Independence Election Results on Zanzibar Election Date ASP ZNP ZPPP Number of % of Votes Number of % of Votes Number of % of Votes Seats Seats Seats 1957 5 35% 0 22% ‐ ‐ 1961 (Jan) 10 43% 9 39% 3 18% 1961 (June) 10 51% 10 35% 3 14% 1963 13 54% 12 30% 6 16% Source: Mmuya & Chaligha, 1994

Given the racial and class inflammations, especially between the Arabs in the ZNP and the

Africans in the ASP, paired with the high stakes leading the negotiations for independence, it is no surprise that the 1961 and 1963 multiparty elections would look an awful lot like the multiparty elections on Zanzibar after 1992. Outright fraud and manipulation was commonplace in both elections. At the conclusion of the 1961 election, riots broke out resulting in some 68 deaths, 400 injuries, and 1000 arrests (Othman 2006). Similar patterns emerged in the 1963 election, which apportioned the main Zanzibari parties to the drafting of the documents at the constitutional conference in London.

139

Political repression continued after the December, 1963 independence, as the ruling aristocracy was quite aware of the tenuousness of its hold on power and moved to ban opposition where it was found. By the early weeks of 1964, a revolution toppled the ZNP/ZPPP government and the monarch and abrogated the 1963 constitution adopted in London. With the assistance of the Umma Party (Peoples Party), which represented a dissenting leftist faction of the ZNP who rejected the racist rhetoric of the Arab elite, the ASP was propelled into power. In what has been defined as “a revolt of the landless peasantry and laboring masses against the landed aristocracy and political oligarchy” (Othman, 2006), the objectives of the Revolution were to remove the Sultanate protected by the 1963 Constitution, as well as to ensure a rapid dismantling of feudalist and capitalist institutions in favor of a socialist development path

(Mpangala, 2000).

Despite this revolutionary victory, the ASP was still exposed to the possibility of a counter revolution or a coup attempt. Given the historical closeness between the Isles and the Mainland, paired with the aforementioned cooperation between the ASP and TANU, a union between the two republics seemed to be the next logical step for entrenching the ASP’s dominance. In April,

1964, the Articles of the Union formally united The People’s Republic of Zanzibar and The

Republic of Tanganyika.

Still, the Union came with a high price tag for Tanzania as a whole, primarily because it widened the friction of conflict beyond the Isles into Tanzania’s national political arena. Not only would the ASP reproduce its dominance through a level of brutality surpassing anything found in Kenya under Moi or Malawi under Banda, but Tanzanian leaders outside the Isles would routinely find themselves stifling political debate over the question of the wisdom and legality of the Union.

140

The rise of multiparty politics in the early 1990s became a catalyst by which objections to the Union would shift from high-level internal party debates to ones taking place more openly during campaigns. While the racial overtones found in Zanzibar’s past can still surface today when encountering justifications for Zanzibar’s autonomy, opposition elites have instead directed their contemporary objections toward the process by which the Articles of the Union was adopted, and the legality of the Union today.

In order to clear up some terminology that will crop up time and again prior to the conclusion, a few words need to be said about the Union’s legal framing. To be succinct, “the

Articles of the Union” is actually a treaty between two sovereign states. As is common with treaty-making, the legality of the Articles would require some form of ratification by both contracting parties. The process for ratification is specified in the Articles, stating that in order to have the force of law the treaty needed to be enacted by the legislative assemblies in both countries, namely the Parliament of Tanganyika and the Revolutionary Council on Zanzibar

(Bakary, 2006).

There have been numerous disputes about whether or not such a process ever took place in the Revolutionary Council. What is known is that the drafting of the Articles was done in the most highly secretive fashion, limited to the minds and mouths of Nyerere and Karume, and possibly others high up in TANU and the ASP, such as Rashidi Kawawa, Oscar Kambona,

Abdallah Kassim Hanga, and Salim Rashidi. The contributions made by the lawyers themselves in drafting the document were suspect as well, undertaken exclusively by the Mainland Attorney

General Roland Brown, a British legal expert, and a Ugandan lawyer named Dan Nabudere21.

21 According to Haroub Othman, the Zanzibar Attorney General, Wolf Dourado, was not included in the drafting of the Articles of the Union because he was previously a member of the ZNP/ZPPP regime. Instead, he was sent away on a one-week leave while the Articles were being drafted (Othman, 2006).

141

Finally, the Union itself was given international support by Britain and the United States as a way of containing the radicalism within the Revolution and the corridors of the ASP22. Out of fear that discussion would thwart the Union formation, both countries encouraged the secrecy around the drafting and passing of the Articles of the Union (Othman, 2006).

In addition to the adoption process, contention also centers on the structure of government outlined in the Articles of the Union, which created a two tiered government, one specifically for Zanzibar and one for the Union and the Mainland together. Powers were divided up between them, leaving a total of eleven matters under the jurisdiction of the Union Government23, while the remaining powers were left to Zanzibar. This autonomy effectively meant that most affairs, including election management, judicial concerns, and the basic institutional design of the government, were left to the authorities of Zanzibar and the Zanzibar Constitution. It is the degree of this autonomy which has become a political hot potato, especially since the early

1980s as the number of Union matters grew.

Since the adoption of the Articles of the Union, most of the tensions over the question of the

1964 merger have been confined to institutional corridors of the ASP and TANU, and later the

CCM. Until the 1980s, the Union Question was quite simply swept under the carpet by patronage, broad commitments to Ujamaa throughout much of Tanzania, and outright repression on the Isles. Even as debates emerged in the 1980s, the language adopted by supporters and opponents of the two-tiered institutional arrangements, defined in the Articles of the Union, was

22 Western powers were quite concerned after the overthrow of the Sultan. For the United States and Britain in particular, Karume’s position within the ASP was far too weak when compared to the communist oriented Young Turks” like Abdulrahma Babu, Kassim Hanga, and Hassan Moyo. After British civil servants and American workers at a space tracking station were forced to leave the Isles, and as the ASP began to establish closer ties to East Germany, China, and Soviet Union, the Western press began to label Zanzibar as “The Cuba in Africa” (as cited in Harris, 2003, p. 18).

23 Today, the number of Union matters has been expanded to twenty-three, an expansion that has been seen by many Zanzibaris as an encroachment on the autonomy of the Isles.

142 often legalistic rather than reflecting ethnic, racial, or class tensions. Only since the early 1990s do we see a resurgence of sectarian discourses within the debates on the Union Question. A combination of long-standing beliefs and political entrepreneurship has reactivated these sentiments since the dawn of the new multiparty era and rendered CCM dominance on the Isles highly fragile.

Conclusion

By the dawn of independence in Tanganyika, which followed nearly a century of colonial enforced peripheral economic development, the consolidation of political power by the leftist urban petty bourgeoisie ideologues, and their abilities to capture the imagination of the peasants, was eased by the relative absence of social forces with ability to compete for peasant votes.

Rather than being hamstrung by widespread rural conservatism, TANU emerged with a cohesive commitment toward getting the British out, arresting control over the commanding heights of the economy, and setting the country on the path toward socialist development. As the data in the aforementioned tables illustrated, TANU’s dominance in early elections resembles, to a remarkable degree, the performance of CCM dominance today.

By contrast, class and racial tensions on Zanzibar gave rise to violent confrontations in the struggle for state power between the Arab aristocracy, Shirazi yeoman farmers, and the African workers recently migrating from the Mainland. Given the intensity of conflict on the Isles, the political dominance on Zanzibar by the ASP, and later the CCM, would prove far more tenuous than single-party dominance on the Mainland. Again, there is a remarkable degree of continuity between multiparty contestation on the Isles today to the nature of contestation at independence.

143

The next chapter links the reproduction of the structural conditions, which prevailed at the time of independence, to contemporary CCM dominance. We cannot simply accept that the structural conditions which prevailed at the time of independence determine the contours of contemporary politics in Tanzania. Instead, the process by which the economic and political elite attempted to implement policy objectives and reproduce political power will have a profound impact on party dominance and contention after the reform.

144

CHAPTER SIX AFRICAN SOCIALISM, IDENTITY AND EQUALITY

In an acquisitive society wealth tends to corrupt those who possess it. It tends to breed in them a desire to live more comfortably than their fellows, to dress better, and in every way to outdo them. They begin to feel they must climb as far above their neighbors as they can. The visible contrast between their own comfort and the comparative discomfort of the rest of society becomes almost essential to the enjoyment of their wealth, and this sets off the spiral of personal competition…

Julius K. Nyerere, 1962

The previous chapter concluded that rise of post-independence single-party dominance is most tenuous where colonialism sharpened sub-national loyalties and capabilities. A combination of well endowed political entrepreneurs and ethnic loyalties waiting to be stirred by the politically ambitious coalesced into highly divided yet polarizing forms of political competition, which challenged the calls for the continuation of national unity. In such circumstances, the collective beliefs about the immorality of, and cynicism toward outsiders, buttressed by powerful senses of exploitation and injustice from neighboring groups, inflated the distance between party factions, strained the durability of intra-party bargains, and seeded the soil for the growth of opposition parties committed to unseating the incumbent at any cost. As described in chapter four, on the African Continent and elsewhere, attempts to neutralize such potentially volatile tensions has been met with varying levels of state coercion, ranging from seemingly arbitrary arrests of opposition leaders, attacks against disloyal meetings and rallies, to outright eradication of entire communities.

In Tanzania, social plurality thrived. However, with over 120 different language groups, and relatively modest degrees of divide and rule tactics and land consolidation, the loyalties that did exist were far too fragmented to coalesce into rural-based national political movements with the capacity for challenging TANU. Only on Zanzibar do we find macro-structural conditions for rendering a tenuous position for the ASP.

My theoretical claim in this chapter is as follows: The rise of sustained popular contention in the lead-up to, and aftermath of multiparty reform is directly related to the production and reproduction of polarized social forces and well endowed political entrepreneurs during single- party rule. Stated differently, the micro-level politics during single-party rule define the macro- level parameters faced by would-be oppositionists when attempting to muster a material and ideational base for electorally competing against the incumbent.

In accordance with the discussion in chapter two, there are a couple of micro-level political characteristics which are especially relevant here. First, single-party rule associated with high degrees of political and economic exclusion, especially against particular social groups, will exacerbate polarization. Exclusion can be an effect of deliberate state policy that seeks to destroy the political and economic clout of particular sections of society. Exclusion can also take the form of policies which tolerate the accumulation of wealth among bourgeois elite at the expense of broader redistributive concerns. Secondly, where single-party rule fails to tie the interests of economic elites with the interests of incumbent survival, long-time ruling parties run the risk of finding well funded opposition parties once the flood gates of multiparty change are opened.

As this chapter will demonstrate, this pattern is nearly the exact opposite of what took place in Tanzania following independence. Thanks to the entrenched position of a leftist leadership stratum within TANU, post-colonial development policies stressed the articulation of a national identity along with economic and political equality rather than local collaboration and economic efficiency. Consequently, single-party rule in Tanzania reproduced the comparatively non-

146 divergent macro-structural environment which prevailed at the time of independence. In this chapter specifically, I will first describe the typical post-colonial African state as a contrast the case of Tanzania. I argue that some of the acute problems of consolidation faced by many

African regimes, which were forced to collaboratively pursue relatively conservative policies or coercively pursue socialist ambitions, were far less acute in Tanzania, giving TANU more latitude to pursue socialism with considerably more consent than found elsewhere on the

Continent. Next, I will move on to show how the TANU/CCM’s embrace of African Socialism, which planted the ideological seeds for state leadership in the economy and political and economic equality, impacted the nature of development policy. In order to simplify the analysis,

I will focus on describing five policy areas: language policy, educational policy, rural development, nationalization, and leadership recruitment and training. Each policy area broadly highlights TANU’s commitment toward African Socialism. Each policy area was also built out of the desire to undermine the growth of structural divergence often associated with development. In later chapters, I will examine the impacts each policy area had on the structural environment. These chapters will demonstrate that African Socialism, somewhat ironically, allowed the CCM to accommodate liberal reform pressures, without suffering the same fate as the Kenya Africa National Union, the Malawi Congress Party, or the United National

Independence Party in Zambia.

Collaboration and Coercion: The African Post-Colonial State

Derived from the Latin word status, meaning condition or standing, the state becomes a necessary focal point when accounting for political stability in general and regime tenure in particular. In bourgeois democracies, the state’s legal monopoly in the use of “coercion and

147 consent,” ensures political stability by enforcing cohesion among the dominant actors around policies which 1) undermine the dialectics associated with private accumulation and 2) organize a national community consenting to the virtues of the market economy and private accumulation

(Poulantzas, 1973; Gramsci, 1977; Jessop, 1982). Political competition is guarded through constitutional principles on rights to organize and freedoms to express political opinions, while the scope of competitive political debate is narrowed by the legal parameters defining a non- interventionalist state and the primacy of private property.

All states in the modern globalized world have some functional equivalence: namely exerting effective control over a given territory, and forming a constituent unit in international relations. However, African states differed from their bourgeois democratic counterparts in two important respects. First, African states, along with many Asian states, emerged institutions constructed by colonial powers for the purposes of linking the extraction and transportation of primary commodities and slaves, and disciplining local and largely rural labor, to the production demands in Europe and North America (Saul, 1974). While all capitalist states ensure the extraction of surplus value from one class to another, the colonial state did so through brutality scarcely found in other places throughout human history. This functional division of labor in the international world economy was replicated locally under the colonial state. Mahmood

Mamdani has identified this bifurcation as a separation between the world of civil law, which largely governed the whites living in urban centers, versus the system of customary rule, which empowered local notables as colonial collaborators, coercively maintaining peace and order throughout rural society1 (Mamdani, 1996).

1 An overemphasized distinction is sometimes made between the form of British colonial policy of indirect rule, where civilized society would be governed by civil law, while “traditional” society would be governed by “tribal” or “customary” law, versus the French colonial policy of direct rule, where all colonial inhabitants were, after being coerced into submission and assimilation, granted French citizenship and subject to one common French 148

Secondly, while independence transferred the seat of political power from the colonizer to the colonized, the post-colonial state essentially played the same economic role as it did under colonial rule. Given the experience of arbitrary colonial partition and bifurcation, along with the post-colonial state as a supplier of primary commodities into the regional and global economy, disciplining society could not take the coercion and consent form found in bourgeois democracies. In contrast to most of those democratic industrial powerhouses in Europe and

North America, these new African statesmen, despite their intellectual commitments to the nation-state ideal, found almost no social basis for dominance through consent once the energies of independence subsided.

However, customary rule also offered state leaders a readymade blue print for continued state dominance, and thus new leaders continued to collaboratively enlist the repressive character of the native authority system. The role of coercion has already been discussed in the previous chapter. Suffice it to repeat that the deportation and public security ordinances found throughout Africa during colonial rule, continued to buttress the position of the ruling sections far beyond the lowering the Union Jack or the Drapeau Tricolore.

In the years following colonial rule, the interests of local notables, whether the rural petty bourgeoisie or so called tribal chiefs, or both, were amalgamated to the interests of the urban bureaucrats and party cadre through patronage (Bayart, 1993; Boone, 1994; Englebert, 2000). In a striking irony, the relationship between local notables and the state after independence was

legal order. These distinctions did exist on paper and had some important ramifications, most notability differences in land tenure (Firmin-Sellers, 2000). However, Mamdani (1996) clarifies that due to shortages of European colonial administrators and the failures of assimilation policy, both French and British models relied extensively on an “association” between chiefly figures and customary authority on one hand, and civil authority in the metropole on the other hand (p. 83). In short, “decentralized despotism,” and thus bifurcation, “came to be characteristic of every colony” (Mamdani, 1996, p. 73). Earlier research by Clignet and Foster (1964) on colonial education clearly points out that the contrast between the British policies of indirect rule and French assimilation are significantly overstated, as both colonial authorities attempted to implement each policy in the education curricula. 149 much the same as before independence: powerful local elites would ensure popular passivity in exchange for access to the state’s potentially lucrative assets. Under this arrangement, rural and urban elites could agree on one thing: keeping popular pressures at bay through state institutions and the all encompassing political party.

At the same time, the necessity of keeping popular pressures at bay was a contradiction in relation to societies where local identities and resources could be easily mobilized into a viable political party sufficient to compete in early elections. Therefore, dominating the state required a winning message. In Africa, this message was that of a loosely concocted socialist idealism, which preached the popular virtues of equality and national unity, but with enough room for placating the interests of a diverse array of local elites, including chiefs and yeoman farmers.

Leaders from the capitals of Conakry to Lusaka repeatedly tried to align loosely defined concepts of socialism and humanism to depict a sense of commitment to the interests of all

Africans and the preservation of African ‘uniqueness’ as a counter to neo-colonial economic development (Le Vine, 1970; Okoko, 1987; Nugent, 2004). In short, championed by a single- party, socialism was an attempt to build the ideological superstructure for sustaining consent. A passage authored by Kwame Nkrumah in 1961 clearly illustrates this point:

We are the Party of the workers, the farmers and all progressive elements in our community….The CPP is a powerful force; more powerful, indeed, than anything that has yet appeared in the history of Ghana. It is the uniting force that guides and pilots the nation and is the nerve center of the positive operations in the struggle for African irredentism. Its supremacy cannot be challenged: THE CPP IS GHANA AND GHANA IS THE CPP2 (as cited in Zolberg, 1966, p. 58).

In the typical cases, concepts of socialism were loosely defined and generally contradictory in nature. Mohan (1966) charges that African leaders “use the rhetoric of Socialism, not as a guide to their actual policies and objectives, but as an ideological scaffold, among other devices,

2 Emphasis in original text 150 for their monopoly of political power” (p. 222). He goes on to cite a Kenyan white paper, which defined African Socialism in almost ideologically sterile and contradictory terms, as a flexible system that can benefit from proven policies, irrespective of the policys’ ideological underpinnings (Republic of Kenya, 1965, p. 2; Mohan, 1966, p. 220). In Zambia, Kaunda describes the blending of the traditional and the modern as part of a gradual “Zambianization of society.” But, Kaunda fails to specify a definition of what Zambianization is and a description of the process of getting there (as cited in Scarritt, 1971, p. 50).

In fact, in many cases it was nearly impossible to use actual policies as a way from distinguishing between regimes oriented toward capitalism and those oriented toward socialism

(Okoko, 1987, p. 12). Arrighi and Saul’s (1968) description is telling:

Even socialists, however, have tended to operate in terms of the conventional model of development based upon the expansion of cash crops for the export market, increased industrial capital formation in consumer-goods industries, and the import of foreign-generally private-capital, the requisite amount of infrastructural investment being the responsibility of the state (p. 154).

Thus, the term socialism would be used to describe policies that accentuated economic inequalities and exclusion, while enriching a class of petty capitalists who happened to be skilled or lucky enough to find themselves mediating the links between domestic and foreign capital through control over the state (Ake, 1976; Shivji, 1976; Mueller, 1980). Perhaps with the sole exception of some thoughts on the necessity of equality, Kaunda’s humanism for example, never clearly threatened the relatively bourgeois values of Zambian political and economic elites

(Scarritt, 1971). Humanism, which was effectively Kaunda’s version of African Socialism, emerged as a contradictory blend of ill-defined ‘African’ traditions of cooperation and community, and a “man-centered” concern for the primacy of the individual (as cited in Scarritt,

1971, p. 49).

151

In cases where socialist ideologies were more rigorously articulated and implemented, powerful bourgeois elements and ethnic interests coalesced to produce violent coups. The most notable example took place in Ghana. Here, where Nkrumaism was defined as the application of scientific socialism to Africa, the CPP, by the mid-160s, emerged as a vanguard of worker interests and required the “mobilization” and “unity” of social forces (Mohan, 1966, p. 257).

Through the institution of “democratic centralism” (Zolberg, 1966, p. 95), attempts were made to mould a “sense of destiny and purpose” through a crusade of sorts which celebrated

“exhortation, missionary work, equalitarianism, opportunism, [and] Pan-Africanism” (Apter,

1970, p. 306).

In summary, the African state emerged as an externally imposed institution for the extraction of labor and primary inputs into global economic supply chains. The modality of colonial rule was based on a mixture of collaboration and coercion between urban and rural society. In most cases, post-colonial rulers simply adopted these tendencies in order to govern societies whose relation to the state was strictly an exploitive one. In some cases, concerted attempts were made to mask repression through vague references to the promises of socialism.

Collaboration and the Post-Colonial State in Tanzania

In part, colonial rule and post-colonial politics in Tanzania exhibited patterns similar to those just described. As pointed out earlier, the coercive power of the state was at times called upon to stifle the expression of interests seen as threatening to TANU and the nation. Indeed, it was demonstrated in chapter four that, in aggregate terms, the use of coercion in Tanzania took on levels similar to the typical authoritarian African case, albeit coercion which generally did not result in politically related killings or violent demonstrations.

152

Collaboration, largely through accommodation, constituted the lifeblood for elite cohesion.

Similar to the process of consolidation described in other cases, the tentacles of patron-client networks extended from the TANU’s center into the leadership circles of some of the most powerful social forces. Hence, after the sidelining of the Tumbo faction3, the continued commitment from TANU collaborators within the Tanzania Federation of Labor4 was shored-up by granting labor representation in the TANU National Executive Committee (NEC) and promoting the union Secretary General Rashidi Kawawa and General Treasurer Michael

Kamaliza to the position of minister of labor at different times after independence (Hopkins

1971, p. 22; Chambua, 2002, p. 21). By 1962, the Trade Union Ordinance required all unions to affiliate with the TFL, under the management of the Minister of Labor and the Registrar of

Unions. The National Union of Tanganyika Act (NUTA) of 1964, passed in response to the suspected role played by labor in the Army mutiny, effectively banned all unions, leaving only

NUTA to represent the interests of workers (Chambua, 2002). The 1964 Act ensured tighter union-party ties by granting the country’s president the power to appoint the NUTA General

Secretary and President (Bienen, 1970). Additionally, NUTA was to function as a TANU agency for mobilizing workers to support party policies and encouraging party advocacy in union ranks.

3 It might be recalled from the previous chapter that that Kasanga Tumbo served as head of the railway union who pressed forward with a radical Africanization agenda. Shortly after independence, the Tumbo faction within the TFL was effectively neutralized. Tumbo himself was safely tucked away as the country’s ambassador in London, and by 1963, became one of the first victims of the Preventive Detention Act (Chambua, 2002).

4 While, in relative terms, labor presented a powerful social interest, in absolute term, the small industrial base of the Tanganyikan economy meant that labor would present only minor obstacles to TANU’s policy ambitions. At the time of independence, most of the wage earners worked on cash crop plantations. Aside from agricultural workers, some 12 percent were employed by the government, the second largest source for wage earning work. Less than 5 percent of the wage earners worked in the manufacturing sector (Bienen, 1970, p. 272). Nevertheless, labor cooperation would be needed if emphasis was going to be placed on industrial growth.

153

Likewise, by 1964, the Cooperative Union of Tanganyika (CUT)5, which represented the most powerful elements of the rural petty bourgeoisie, namely the export oriented cash crop farmers, was represented within the NEC (Okema, 1996) and converted into an institution for regulating the price of crops and collecting taxes (Bienen, 1970). When the effectual cooperative leader named Paul Bomani lost his 1965 parliamentary election bid, he was appointed by Nyerere as Minister for Economic Affairs and Development Planning (Bienen,

1970).

Be that as it may, the overall absence of social forces with the potential for dislodging

TANU’s hold on executive power, especially the weakness of ethnic and petty bourgeoisie impulses, meant that party leaders had the latitude to accommodate without fears of straining intra-party clientelistic control. Furthermore, it also meant that the leadership stratum had a freer hand in articulating socialism as a way for strengthening the party-peasant ties. The position of chiefs, as the conservative rural agents of colonial collaboration, and ethnic associations, as the principle opponents of the chiefly authority, simply lacked any broad base for mobilizing local cleavages against TANU’s ascendency. Likewise, the peripheral patterns of economic development, as identified in the previous chapter, did not bode well for consolidating economic resources as a basis for viably opposing TANU policy ambitions. Aside from a few large sisal plantations, the economy was almost exclusively based on small-holder peasant production.

5 Throughout much of east Africa, agricultural production was organized through locally or regionally-based cooperatives. Cooperatives are collective organizations formed by farmers primarily for the purposes of exerting control over prices, including the costs of loans, supplies, and the prices attained for crops at the market. These associations can also be a source of organizational strength for linking agrarian interests to local, regional, and national policy-making institutions. First appearing in the 1920s in Tanganyika, co-operatives such as the Kilimanjaro Native Cooperative Union (KNCU) and the Bukoba Native Cooperative Union played important roles in pre-independence politics (Agonafer, 1994). This was especially true for the Victoria Federation of Cooperatives Union, an umbrella union in the cotton belt region formed by Paul Bomani, which by the late 1950s, became the largest co-operative organization in Africa (Hyden, 1980) and effectively became a TANU operative after the banning of a party branch in the Lake Province (Bienen, 1970). Many cooperative leaders featured as prominent members in TANU (Okema, 1996). 154

Likewise, aside from a few mining and industrial operations, the manufacturing sector was one of the smallest in East Africa. For those few that could be classified as wage workers, most were employed in cash crop sisal plantations. In short, Tanzanian society simply lacked the antagonisms and class structures that, at the very least, forced leaders in other countries to moderate policy ambitions and engage in coercive action in an effort to thwart an emboldened opposition.

Consent and the Post-Colonial State in Tanzania

While collaboration by itself did not make Tanzania an exceptional case at the onset of independence and single-party consolidation, the relative weakness of conservative social forces meant that an ambitious ideology actually guided policy-making to a far greater extent in

Tanzania when compared to many of its neighbors. For example, whereas Jomo Kenyatta in

Kenya, Hastings Banda in Malawi, and Houphouet-Boigny in Côte d’Ivoire tended to emphasis the need to develop within the political and economic framework inherited from colonialism

(Foltz, 1970; Kaunda, 1998; Nugent, 2004), the leadership stratum within TANU vocalized a remarkable degree of dedication to social and economic transformation. The consensus within the circles closest to Nyerere wanted to rid the country of all the vestiges of colonial rule, including class and wealth inequalities, along with the highly unpopular institutions of customary rule. However, unlike the revolutionary vigor contained in the scientific socialism of

Kwame Nkrumah’s final years in Ghana, or the tenure of Sékou Touré in Guinea, the leadership stratum in Tanzania rejected the vanguard and proletariat language of Marxism and Leninism, opting instead for a more conciliatory, pragmatic, and perhaps contradictory notion of African

Socialism (Okoko, 1987).

155

Just as Tanzania’s progressivity was unique, so to was Nyerere’s commitment to ideological coherence. Kaunda’s humanism suffered from a lack of consistency, coherence, and concreteness (Tordoff, 1988) by locating social problems (and authoritarian rule) in natural human tendencies of “greed, selfishness, and the animal in man” (Kaunda, 1974; Gertzel,

Baylies, & Szeftel, 1984) In Kenya, Jomo Kenyatta, who had to “bargain to a much greater extent than Nyerere” with “established regional and ethnic leaders” with political ambitions of their own, never formally instituted single-party rule. Furthermore, despite attending school in the Soviet Union, Kenyatta never expressed any strong commitment toward ideology (Barkan,

1994). By the end of the 1960s, the more conservative political elite found in the Kenya African

Democratic Union (KADU) were co-opted by patronage and KANU’s pro-capitalist and pro- decentralization platform, while the renegade leftists who migrated to the Kenya People’s Union

(KPU) were neutralized through coercion and election manipulation (Throup & Hornsby, 1998).

Finally, while the Malawi Congress Party (MCP) “was the most popular, dynamic and successful political party in Central Africa – one that ranks alongside TANU…and the CPP”

(McCracken, 1998), the party contained far too many mutual hostilities for being conducive to a coherent ideology outside Banda’s absolutist style and personal ambitions in allying the country with the West.

For many, including the peasants during the first decade of independence, African socialism in Tanzania captured the imagination and became a focal point around which broad transformative ambitions could be pursued. It was the normative framework which steered state policy, and ultimately, the distribution of resources throughout Tanzanian society from the

1960s up through the early 1980s. Citing Arrighi and Saul (1968) at length: “Tanzania is, perhaps, the country in contemporary Africa where socialist aspirations figure most prominently

156 and interestingly in the development equation, and most powerfully affect the kinds of policies which are pursued” (p. 165).

Often dubbed as a form of utopian socialism (Nursey-Bray, 1980) similar to the visions of

Robert Owen and Charles Fourier, and the revisionist socialism in the traditions of Eduard

Bernstein and the Fabians, the philosophy sees capitalism as a form of economic organization ill-equipped to produce economic justice and social cohesion required for building nationhood.

Where the private market prevails, participants are socialized to be inward looking egoists engaged in the relentless pursuit of wealth maximization, all while economic policy becomes infatuated with aggregate economic expansion. Because of egoism and a highly agential sense of individual achievement, the capitalist “attitude of mind” operates as a basis for the rationalization of social inequality, poverty, and ignorance, which, according to African

Socialism, are every bit as important as aggregate growth (as cited in Hyden, 1980, p. 98).

Furthermore, for state leaders concerned about political stability, as related to social cohesion, forms of economic organization yielding greater degrees of social differentiation were treated with suspicion, as obvious threats to the reproduction of dominance and emergence of new nations.

At the same time, the philosophy of African Socialism rejected the application of classical

Marxist and Leninist thought to African conditions. In a frequently cited passage from his 1962 piece entitled “Ujamaa – the Basis of African Socialism”, Nyerere describes African socialism as:

opposed to capitalism, which seeks to build a happy society on the basis of the exploitation of man by man; and it is equally opposed to the doctrinaire socialism which seeks to build a happy society on the philosophy of inevitable conflict between man and man…We, in Africa, have no more need of being ‘converted’ to socialism than we have of being taught democracy (p. 15).

157

Often dubbed “scientific socialism”6, some of Nyerere’s later rivals like Oscar Kambona, along with other state leaders, such as Nkrumah and Touré, aspired to transplant to Africa the

Soviet experience of open class warfare. For the scientific socialists, history was driven by class conflict and Africa was far behind in that process. Traditional feudal structures, whose superstructure legitimated class exploitation in paternalism and superstition, were stubbornly persistent in Africa and needed to be eradicated (Saul, 1974; Raikes, 1975; Okoko, 1987).

By contrast, for Nyerere the structural problems in rural society throughout Africa were not related to traditional African life, but to the native authorities system and inward-looking

“tribalism” concocted and reinforced by colonialism. Whereas scientific socialism seen the necessity of social class and structural differentiation, as rooted in relations of production and as the necessary conditions for a transformation toward socialism (Saul, 1974; Raikes, 1975;

Okoko, 1987), African socialists were skeptical of structural differentiation as a source of fragmentation, conflict, and ultimately as a threat to nation-building. Nyerere (1962) draws an analogy between doctrinaire socialism and religious rituals when stating that “As prayer is to

Christianity or to Islam, so is civil war (which they call ‘class war’) to the European version of socialism – a means inseparable from the end” (p. 6). In short, not only was doctrinaire socialism incommensurable with national unity, but it was also seen as a neo-colonial threat to the Country’s recently won independence (Nursey-Bray, 1980). Instead relying on neo-colonial ideas, Africans had to find their development path outside the framework of doctrinaire socialism or capitalism.

6 The term “scientific” was used by Marx and Engels to denote their view of socialism not as the creation of ‘ingenuity’ or imagination, but the “necessary outcome” of class struggle. In support of a transition to socialism, the “task” of a socialist was to scientifically “examine the historic-economic succession of events from which these [bourgeoisie and proletariat] classes and their antagonism had of necessity sprung, and to discover in the economic conditions thus created the means of ending the conflict” (Engels, 2003).

158

Shaped by the necessity of nation-building in a context where societies were highly fragmented along lines of geography, the politics of nation-building was of certain central focus in Nyerere’s ideological formulations. Gone was the assumption that economic growth will automatically produce political and social progress toward a just and unified society (Barkan,

1979a, p. 33). Economic development should, according to Nyerere, first focus on laying the ethnical foundations for nationhood and economic development, rather than letting rapid economic development, which required rapid capitalization, drive the ethnical foundations of society. The years following independence were seen as a window of opportunity, albeit a narrow and rapidly closing one, for the laying of these ethical foundations (Pratt, 1999).

Based on a romanticized perspective of African peasant traditions, socialism in Africa, as conceived in the mind of Nyerere, was to emerge out of a theorized “organic” tie between the individual and the traditional African community (Mushi, 1971; Nugent, 2004). This sense of community, as bonds of brotherhood and familyhood rather than bonds steeped in feudal patron- client transactions, was the basis for articulating a version of a “classless” society in which adherents to Marxist and Leninist thought were utterly opposed to (as cited in Okoko, 1987, p.

13). For African socialism to materialize, shared experience and social obligation, as the moral foundation of the peasantry (Hyden, 1980), needed to be pulled out of its local sectarian settings, imbued with the norms and capital for modern production, and adopted as a bond for national unity, economic development, democracy, and socialism (Okema, 1996, p. 41). Hence, the term

Ujamaa, meaning familyhood, acquires its centrality in the concept of African Socialism and national unity.

The twin concepts of mass participation and democracy, normalized through a single party open to all, comprised the political framework by which African socialism would come to

159 fruition in Tanzania. While the single party was the ‘patriarch’ of African socialism and the beneficent father of the national family, Nyerere plainly rejected the Leninist model of a vanguard party composed of socialist intellectual elite (Pratt, 1979). Instead, Nyerere conceptualized the party is an institution where “people can and must express their desires and worries” to the government (Mwansasu, 1979). At the same time, party and government officials were supposed to provide ideological coherence to government policy, and communicate this ideological coherence back to the peasantry. Whereas the Leninist model was fundamentally authoritarian in nature, for Nyerere, the party should instead operate as a venue for the reciprocal exchange of ideas between elites and masses.

Additionally, this “two-way all weather road” (Nyerere, 1966, p. 158) called TANU, functioned as a central organ in accountability and community life not just during times of elections, but also during the times between them. Ceremonial events, such as the Uhuru Torch

Race7, Heroes Day8, Union Day9, and Independence Day, were intended to institute a participatory spirit and to bring people together for the ritualization of a shared political experience that would constitute a basis for nationhood. As Siri Lange (1995) states in her meticulous research on the uses of traditional dance as a “vehicle in the nation-building process,” (p. 27) songs and dance celebrations of national unity and political leadership were routinely conducted at schools, places of work, and in important ceremonial events. To be terse,

7 The Uhuru Torch Race has been carried out every year since 1964 by the Party’s youth wing, and, after multiparty reforms, by the National Service. Occasioned by a variety of other festivities, the torch is carried throughout every region of the country as a symbol of national unity. According to Michael Okema (1996), the “act of receiving the torch from another village, district or region is a reminder of a common heritage of the villages, districts, or regions” (p. 55).

8 Celebrated in September first, Heroes Day commemorates the 1964 inauguration of the Peoples Defense Forces.

9 Union Day, celebrated on April 24th, commemorates the Union between Tanganyika and Zanzibar.

160 political participation was not just a means for selecting representatives, but also as a way for instituting a real sense of equality and unity as the “political corollary of material equality”

(Barkan, 1979a, p. 29).

From the days before and after independence in Tanzania, the statements and proclamations radiating from policy ideas reflected a rhetorical commitment to the principles of African socialism. In 1962, Nyerere released a landmark work, entitled “Ujamaa, the Basis of African

Socialism,” where he forcefully argued that Tanzanians needed to be ‘reeducated’ to reject the

“capitalist attitude of mind,” which stresses exploitation of another for material acquisition, and accept “our former attitude of mind,” where people “took care of the community, and the community took care of [the people]” (1966, p. 28)

Likewise, the Constitution of 1961 was forceful on the need to eliminate the three enemies of the nation (poverty, ignorance, and disease), and to establish collective control over the means of production. By 1962, all land was nationalized. Thanks to the weak position of petty and commercial bourgeois classes, opposition to nationalization was minimal. In early 1963,

Nyerere announced the NEC’s intentions to pursue the formation of a “Democratic One-Party” state. Finally, Article Two of the 1965 Constitution was more direct by specifying that economic intervention was “responsibility of the State” in order to “prevent the accumulation of wealth to an extent which is inconsistent with the existence of a class-less society” (as cited in Hopkins,

1971, 20).

These formal proclamations and legal expression are not to suggest that commitments toward African socialism in Tanzania were shared by all within the leadership stratum, or that ideology was the final arbitrator in policy-making. Behind the scenes, the prevailing ideology in

Tanzania was widely criticized within the leadership stratum, mostly from those “scientific

161 socialists” more wedded to Marxist and Leninist strains of social thought and those pushing for closer cooperation with the Soviet Union. Rejecting Africa’s past and the apparently naive pacifism in utopian thought10, scientific socialists in Tanzania favored a concerted focus on rapid industrial development, which would raise overall output and generate the class-based social forces necessary for evolution toward socialism. The energy among these revolutionaries centered on a collection of ministers and administrative officials, most noteworthy being

Nyerere’s close associate and schoolmate at Edinburgh, Oscar Kambona. The base of this faction was found among youth and labor, led by the likes of union leader Kasanga Tumbo and among a cluster of Zanzibaris, such as A. M. Babu, Minister of Commerce and Cooperatives, and Kassim Hanga, Vice President of Zanzibar Revolutionary Council and Minister for

Industries, Mineral Resources, and Power (Bienen, 1970).

As the 1960s wore on, factional divisions within the party’s leadership stratum grew more pointed around what African socialist policies should look like and around the merits of African socialism altogether. For starters, African socialism was not immediately forthcoming in development policy. Inequalities of wealth and access to social services widened enough to alarm party leaders concerned about the growth of social divisions with the potential for upsetting national and party cohesion11. At the same time, private sector development was still

10 Marx and Engels committed similar attacks against what they called “Utopian socialism,” a belief and practice that wants “to improve the condition of every member of society…without distinction of class…they reject all political, and especially all revolutionary, action; they wish to attain their ends by peaceful means…and by force of example” (2000). This description of Utopian socialism accurately frames Nyerere’s perspective on African socialism.

11In rural society, wealth and access to social services disproportionately benefited the most progressive producers. This was especially true for the owners of sisal plantations, since buoyant sisal prices offered producers significant returns on investment (Bienen, 1970, p. 268). Likewise, programs intent on promoting model farms, such as agricultural education and the distribution of heavy farming machinery, generally did little more than enhance the economic position of surplus producing farmers relative to peasants (Hyden, 1980, p. 77). Additionally, the urban-rural duality in Tanzania increasingly reflected the trends found elsewhere in Africa. Income differentials between urban workers and rural peasants continued to grow, as did regional inequalities in general (Bienen, 1970, p. 414). By 1967, the annual pay for an urban worker rose by 80 percent since 162 tolerated, if not encouraged. Expatriates from Great Britain12, France, and the World Bank, whose overall commitment to socialism was questionable, still dominated a highly technocratic policy-making process, again upsetting those still pressing for Africanization13. As Nugent

(2004) points out, while TANU formally “signed up” to Nyerere’s ideas, the Party “was scarcely influenced by them at this early stage” (p. 141). Joel Barkan (1979a), in perhaps a bit of an over simplification, echoes this sentiment when stating that the policies pursued in Tanzania immediately following independence “were similar to those pursued by virtually every other

African state”14 (p. 4).

Alongside these concerns, scientific socialists continued to complain about the overall weakness and naiveté of African socialism as a political force for bringing about real structural change. Debates ensued over the logic of a mass party, as TANU was criticized for lacking a

independence. The wage differential between the highest and lowest paid government employees was 30 to 1 (Buchert, 1994, p. 99).

12 Nugent (2004) states that by 1966, the some 400 British officers remained in Tanzania, most of whom were colonial administrators prior to independence. To make matters potentially more offensive to Tanzanians, many of these officials were slated for pensions paid for by Tanzanians (p. 141).

13Following independence, the government continued to utilize the inherited three-year conservative development plan, inspired by the World Bank and drawn up by the British prior to independence. Far from a radical departure toward socialism, this plan stressed the importance of private sector agricultural development and industrial development based on import substitution, thus requiring a continued dependence on foreign capital. The new Five-Year plan, drawn up between 1963 and 1964 by a handful of high-level government and foreign technocrats operating within the Ministry of Development and Panning, which was latter converted to the Directorate of Development and Planning, was marginally more progressive then the plans inherited at independence. While it instituted an increased involvement of public sector investments in economic development, the catchword for the technocrats and the plan’s defenders was the “mixed economy” approach. The plan continued to encourage private enterprise, alongside planning in sectors controlled by the state and the cooperatives. Accordingly, around 53 percent of the investment expenditures were to be taken on by the public sector, while the remaining outlays were to come from private investments. Furthermore, industrial development, managed by the newly formed National Development Corporation (NDC), was directed toward those regions where incomes were high enough to consume the products produced. Therefore, the plan was subject to wide criticism from regional and local party leaders more loyal to Nyerere, but concerned about regional economic inequality (Bienen, 1970; Shitundu, 2002).

14 Barkan (1979a) goes on to point out that, out of the 58 “published speeches” that Nyerere presented between 1962 and 1967, the number that mention a connection between socialism as a model for Tanzania development does not exceed 5 (p. 7).

163 membership criteria and ideological vigor worthy of the name socialism15. Even more pointedly were the repeated calls for nationalization of the commercial sector and heavier investments in industry.

By the end of the 1960s, this impatience culminated in a series of expulsions, arrests, and trials. For starters, the increasingly radical Oscar Kambona, who had an almost cult like following among some corridors of society, was forced to flee the country in 1967 after accusing Nyerere of becoming a dictator. During the October 1968 NEC meeting in Tanga, nine

MPs were expelled from the Party. One was placed in preventive detention. In October 1969, a handful of Nyerere’s critics were arrested and charged with plotting to overthrow the government between the 10th and 15th of October. Among those arrested were close associates of Oscar Kambona, including Bibi Titi Mohamed, head of TANU women’s wing, Michael

Kamaliza, formers Labor Minister and head of NUTA, and Grey Mataka, former editor of the

Nationalist. While all were put on trial for treason, none received the death sentence. Two of

Kambona’s brothers and Bibi Titi Mohamed were pardoned in 1972.

Yet, the factional tensions within TANU also promoted Nyerere to take more seriously real commitments toward socialism, commitments which would have a lasting impact on the reproduction of regime tenure. In 1967, the developmental course that would be charted for the next decade to come was laid down. Bienen (1970) cites that the Arusha Declaration, which gave Tanzania’s development policies a much clear socialist direction, was “perhaps the most

15While the organizational expansion of the party throughout the 1960s was impressive, as the ties between local government and party were strengthened and the ten cell system instituted, the theoretical and practical party- government relations were far from satisfactory for scientific socialists. There was a clear want to see TANU not as a mass party, but as a party with a more defined and disciplined conception of socialism and a membership limited to those with clear dedication to this conception. While the Party Ideological Kivukoni College offered training in social sciences in order to develop a more educated cadre of middle-level TANU managers (Hopkins, 1971), the concept of a mass party also begged for a less rigorous and more nebulous sense of ideology (Mwansasu, 1979, p. 175).

164 dramatic attempt to assert a unique Tanzanian solution to developmental problems” (p. 406).

Officiated in Arusha during the January 1967 meeting of the NEC, the Arusha Declaration was most immediately conditioned by Nyerere’s inspirational visit to China in 1965, the deteriorating aid relations with the West that highlighted the need to break the country’s dependence on aid, and the political turmoil in other parts of Africa brought on by economic and ethnic tensions. Additionally, the accumulation of wealth, especially by a number of prominent party elite, illuminated the need to make a clean break from capitalistic and clientelistic behaviors. As a whole, the Arusha Declaration solidified the philosophy of African Socialism, couched the development strategy in terms of rural transformation (i.e. peasant production), and reified into an iconic reference point for future social and economic policies (Hyden, 1980;

Okema, 1996; Nugent, 2004). Let us now look at some of those policies.

Language Policy

In the lead-up to independence and the years immediately following, TANU leaders routinely displayed their aversion toward politicking on the basis of ethnicity and often treated it as an issue of internal subversion and national security. Nyerere (1968) so forcefully expressed this sentiment during a 1967 speech at a Zambian UNIP party conference, where he underscored the underhanded tactics pursued by neighboring Apartheid regimes.

It is for this very fundamental reason what the enemies of Zambia, and of Africa, still make every possible attempt to destroy unity the unity of Zambia. And to do that they will have to destroy the unity of UNIP. Their methods will be insidious, underhand -and clever. In particular, they will play upon our traditional loyalties to tribes and men (p. 332).

While leadership in Tanzania at the time of independence benefited from highly fragmented ethno-linguistic loyalties, representation within the state clearly favored those groups hailing

165 from the cash crop producing regions of the country, largely because educational opportunities were far more widespread in these areas. As table 6.01 demonstrates, the Haya, Chagga, and

Sambaa were disproportionately represented within the civil service and, to a lesser extent, within the legislative assembly.

Table 6.01: Ethnic Representation in Government Ethnic Group Nation Administrators Legislators Sukuma 12.5 4.3 10.2 Nyamwezi 4.1 3.4 5.8 Makonde 3.8 0.9 0.9 Haya 3.7 20.7 5.1 Chagga 3.6 17.2 3.6 Gogo 3.4 1.7 2.2 Ha 3.3 0.0 2.9 Hehe 2.9 0.0 2.2 Nyakusa 2.5 7.8 3.6 Luguru 2.3 0.9 0.0 Bena 2.2 2.6 2.2 Turu 2.2 0.0 0.7 Sambaa 2.2 4.3 3.6 Zaramo 2.1 1.7 2.2 Iramba 2.0 0.0 2.9 Yao 1.6 0.0 1.5 Mwera 1.6 0.9 0.7 Mbulu 1.5 0.0 0.7 Zigua 1.5 2.6 2.9 Pare 1.4 2.6 1.5 Other 51.0 29.3 44.3 Source: Hopkins, 1971, p. 77

At the same time, a correlation study performed by Raymond Hopkins during the 1960s indicated that ethnicity is not a highly significant factor for organizing elite cleavages (1971, p.

82). The leadership stratum within TANU would go to considerable lengths to keep it that way.

This was especially true for the approach taken by party leaders toward the usage of Kiswahili.

By some accounts, language forms a basic, almost essential component of national identity

(Papstein, 1989; Davison, 1997). At the very least, language constitutes a basis for socially

166 identifying national and ethnic groups (Forster, Hitchcock, & Lyimo, 2000). There is little doubt that language can function to smooth network transactions, act as a basis for distinguishing between insiders and outsiders, and become a valued symbol national familyhood. Therefore, one might expect language to be a sufficient, although not a necessary, condition for reducing the level of divergence within a given polity.

After Tanzania’s independence, the usage of Kiswahili continued to be promoted by the government, not only as a tool for efficient communication, but also as an engine for national integration and a symbol of national identity. As stated earlier, the breadth in which Kiswahili is spoken throughout Tanganyika greatly assisted unity in the drive for independence. While

Kiswahili is widely spoken in Uganda and Kenya, the language is strongly associated with particular regions – the southern portions of Uganda and the coastal regions of Kenya.

Furthermore, while considered as a national language in Kenya, many Kenyans have admitted to me that fellow countrymen have not really mastered Kiswahili and that, to learn it, it is best to stay in Tanzania. By contrast, given the language’s level of diffusion throughout Tanganyika, thanks to pre-colonial trade routes that linked the coastal areas to the hinterland, Kiswahili in

Tanzania had the advantage of being widely spoken and disassociated with any particular ethnic group (Okema, 1996, p. 57).

Even prior to independence, TANU encouraged the usage of Kiswahili through literacy and adult education programs. As the language of the Party’s mass organization, Kiswahili grew to symbolize national unity and the spirit of the revolution against colonial rule (Askew, 2002, p.

182). After independence, TANU leaders took a number of strides to ensure that Kiswahili was central to Tanzanian life, starting with the 1964 adoption of Kiswahili as the official national language. Compared to Kenya, where local vernaculars are the language of instruction in

167 primary schools, while English the principle language of Secondary schools, in Tanzania

Kiswahili was made compulsory in all secondary schools in 1965 (Miguel, 2004). During that same year, TANU adopted a policy requiring candidates to use Kiswahili in all political campaigns. Only in conditions where locals could not adequately understand Kiswahili, was an interpreter permitted. In order to carry out research on Kiswahili, the Kiswahili Institute at the

University of Dar es Salaam was opened in 1964. By 1967, Kiswahili became the medium of instruction in all primary schools. Adult illiteracy was to be eliminated by 1975. In order to assist in these ambitions, two language councils were created, the Baraza la Kiswahili la Taifa

(BAKITA), which was to promote the language throughout the United Republic of Tanzania, and Baraza la Kiswahili la Zanzibar (BAKIZA), which was charged with promoting the use of

Kiswahili on the Islands16 (Bienen, 1970; Okema, 1996; Askew, 2002).

As will be demonstrated more thoroughly in Chapter 12, the continued emphasis on the diffusion and perfection of Kiswahili paid huge dividends for peace and unity in Tanzania.

Kelly Askew (2002) supports this claim when noting that “many credit the impressive sense of

Tanzanian national unity in large part to Nyerere’s decision to institute and enforce Kiswahili as the national language as well as the sole instructional medium in primary school” (p. 183).

As will be illustrated in subsequent chapters, the diffusion of Kiswahili also paid dividends for the ability of the CCM to defend its tenure after the onset of multiparty change. First, as suggested above, the diffusion of Kiswahili acted as a centripetal mechanism for reducing polarization along ethnic lines. All members of the Tanzanian community share in a common etymological, phonological, and morphological mode of expression, one that is not tied to any

16 On Zanzibar, Kiswahili was more heavily influence by English and Arabic. Therefore, the objectives of BAKIZA were much more open to the exploration of Kiswahili and promotion of dialectical differences. BAKITA on the other hand, was preoccupied with the construction of a more pure Kiswahili (Askew, 2002, p. 183). The fear was that applying the BAKITA objectives to language policy on Zanzibar would have been offensive and damaging to the Union 168 particular Tanzanian ethnic group. As repeated time and again, reduced levels of polarization translated into fewer sub-national identities with intensified senses of insecurity and distinctiveness which are the bases for mobilizing people into political action in so many parts of the world.

Secondly, while the diffusion of Kiswahili constituted a mechanism for the solidification of a national identity, defense of the nation today constitutes the raison d’etre of the CCM. Defense of the nation against internal and external threats has become a campaign expression, one that constructs pressing threats to the nation as outgrowths from the divisive nature of the opposition. Defending against this divisiveness is, at the same time, constructed as the central role of the CCM. In short, nearly every political battle is defined as an issue of national security, where the CCM can rightfully claim to have a good track record. These points will be taken up more explicitly in chapter twelve.

Education for Self-Reliance

African socialists within TANU expressed sincere reservations about the growth of “social differentiation” brought about by modernization and the educational objectives prior to Arusha embodies in what was referred to as Education for Manpower Development (Buchert, 1994). At the time of the Arusha Declaration in 1967, less than half of the eligible children were being serviced through primary schools (as cited in Bienen, 1970, p. 428). Concerns over differentiation and grumblings within the corridors of the Party over growing regional inequalities in education became the main motivations for the adoption in 1967 of new policy objectives embodied in what was called Education for Self-Reliance (ESR) in 1967.

169

The approach taken toward education in Tanzania was highly unique. As Cooksey, Court &

Makau (1994) point out, “ESR was without doubt one of the most systematic attempts in the

Third World to mold cooperative, non-elitists, adaptive (to local economy and culture), and socially leveling education structures” (p. 216). To be sure, educational pursuits in neighboring countries were more laissez faire. In Kenya for example, despite the stated support within the

Kenyan Ministry of Education for utilizing the school system as a nation-building institution

(Court & Ghai, 1974), primary schooling generally taught provincial geography and history, and, as Miguel (2004) suggests, “probably serves to exacerbate regional and ethnic divisions, especially among the many Kenyans who drop out of primary school before grade 5” (p. 336).

By contrast, education in Tanzania under ESR was more clearly directed toward national integration not just in a materialist sense through creating a more inclusive and egalitarian environment, but also in a more distinct ideational sense through the shaping of popular discourses over the meaning of the Party and the Nation and the roles played by citizens

(Buchert, 1994, p. 115).

In pursuit of self-reliance, the need of hard work, as a social responsibility to the community and nation, were emphasized rather than relying on market incentives for the encouragement of labor (Buchert 1994). Kiondo (2001) cites that the promotion of national culture through a strong dose of political education in schools, the Pioneers, and the TYL, which celebrated the “country’s nationalist struggles against colonialism and narrow nationalisms,” along with “sports, traditional dances, stories, songs and other cultural practices,” were among the most important contributions to “molding a national identity in Tanzania” (p. 262). Civics lessons, current affairs, and political training were supplemented by the frequent participation in

170 local development projects and TYL activities designed for strengthening commitment to the

“national political process, the national party and its ideology” (Buchert, 1994, p. 115).

Likewise, higher education was tailored toward fostering a shared sense of national comradeship and responsibility. When celebrating the opening of the University College campus in 1964, Nyerere made the logic of higher education clear:

The annual per capita income in Tanganyika is £19. The cost of keeping a student at this college [the University College of Dar es Salaam] will be about £1,000 a year. That is to say that it takes the annual per capita income of more than 50 of our people to maintain a single student at this college for one year. It should not be necessary to say more (as cited in Lema, Mbilinyi, & Rajani, 2004, p. 19).

Hence, in order to enter higher education, one year national service was required. Between 1974 and 1984, entry into the university required an additional two-year work experience, as part of an ‘attitudes toward work’ program (Buchert 1994).

While it is clear that the socialist ambitions of Ujamaa were treated with the utmost suspicion by many sections of society, especially among those university students who, for two years, would apply their newly acquired skills as manual laborers in the remote some of the remote parts of the country. Yet, a study during the 1970s showed that Tanzanian primary school students tended to be solidly acquainted and committed to the “official ideology”

(Buchert, 1994, p. 116). More critically, country’s educational efforts went a long way in reducing ethnic and regional tensions and fostering a real sense of egalitarianism. As

Mwakikagile (2004) described his personal experience:

At our school, students came from all parts of the country and from many different tribes. We were not encouraged to attend – except at the primary school level – in our home districts, which were usually inhabited by members of our own tribes. We were in fact, assigned to schools and jobs after graduation far away from our tribal homelands in order to love and work with members of other tribes. It was a deliberate effort by the government to break down barriers

171

between members of different tribes and races in order to achieve national unity (p. 110).

Rural Development

For the leaders of Africa’s newly independent states, one of the most pressing problems was launching a path toward economic growth from a starting point characterized by rural, agrarian- based production, with varyingly low degrees of capitalization. For some, moving past the starting point simply meant reforming colonial development trajectories of the past, but leaving the foundational aspects intact. Hence, in Malawi, the 1961 MCP Manifesto portrayed the eagerness to facilitate the growth of large-scale cash-crop estates, alongside smallholder production. After the emergence of single-party rule in 1966, MCP and state officials were encouraged to buy-up commercial farms so as to replace European settler domination of commercial agriculture. Together with the institutional encouragement from the Agricultural

Marketing and Development Corporation (ADMARC), the National Rural Development

Program (NRDP), and the Malawi Development Corporation (MDC), the petty bourgeoisie benefited from low tax and land rental rates, along with access to bank credit and cheap labor

(Kaunda, 1998).

Patterns of rural development were similar in Kenya, albeit far less centrally administered than what was found in Malawi. Because of the decentralized nature of party formation, largely due to British prohibitions against nation-wide independence movements and the entrenched position of local notables promoted under the Native Authority System, The Kenya African

National Union (KANU) emerged from independence as a highly fragmented party organization, with no clear channels for strong social control from Nairobi (Barkan, 1979a).

Given the weak capacity of the party paired with the economic clout of rural cash crop farmers,

172 the pragmatic and skillful Jomo Kenyatta was not about to violate the entrenched economic interests among the rural petty bourgeoisie through major political and economic reforms similar to those which infolded in Tanzania. He instead, circumvented the Party altogether, and built authority around what has come to be labeled as “Patron-client capitalism”, a form of social control which linked political ambitions of local notables to the national center’s ability to attract development funds. These clientele relations spanned from common peasants, through local ethnic elders and businessmen, members of parliament, culminating in the office of the presidency (Barkan 1979b).

While the emerging party ideologues and bureaucrats in Tanzania faced similar challenges from local notables, the peripheral development path ensured that these challenges would be far less acute in comparative terms. At the beginning of independence, TANU leaders generally agreed on one important thing: The institutions by which rural politics was steered, namely the

Native Authority System, had to be dismantled.

National political leaders quite simply distrusted institutions and officials which spoke the language of kinship and ethnicity. At the time of independence, Nyerere made it clear that the system of chieftaincy and the Native Authority System generally, would play no role in post- independence Tanzanian politics. As Nyerere stated in 1960:

We tell the Chiefs quite frankly that their authority is traditional only in the tribes, which were traditional units. Tanganyika is not a traditional unit at all, and if the Chiefs want to have a place in this thing we call Tanganyika, they have got to adapt themselves to this new situation. There is nothing traditional in the Central Government of Tanganyika today (as cited in Bienen, 1970, p. 66).

Therefore, in October of 1960, the tribal unions17, which were so important for the penetration of the TAA into rural Tanzania, were put under the authority of the Ministry for

17 As a reminder, tribal unions grew out of opposition to colonial rule and the chiefs which were empowered to administer what the colonialists saw as tribal entities.

173

Local Government and Housing, headed by Rashidi Kawawa. The institution of chieftaincy was also attacked. While chiefs were at first simply encouraged to resign from their posts by regional and area commissioners (Maguire, 1969), by 1962 the Local Government Ordinance repealed the 1953 African Chiefs Ordinance (Mukandala, 1998). The position of the chief was reduced to little more than an “honorary title” (Hopkins, 1971, p. 25) with the permission to perform only traditional ceremonial roles (Maguire, 1969, p. 336)18. Their formal powers were passed onto elected district councils, which in turn were supervised by TANU commissioners and the

Minister of Local Government (Buchert, 1994). While some chiefs obtained positions in local government, evidence show that only a handful found their way into the national assembly

(Bienen, 1970). Most of them simply retired from politics without launching serious complaints

(Maguire, 1969).

Fast forwarding seven years or so; a new set of policy ambitions entered the fray, ones that would be forced to confront the material base of the petty bourgeoisie directly. For African

Socialism in Tanzania, the pressing problem of economic development looked something like this: How is it possible to develop a strong economic base out from those resources which the country could produce on its own (i.e. production based largely on low capital intensive smallholder production), without generating a corresponding growth in economic inequality and an egoistic landowning class seeking to maximize returns from land and labor (Barkan, 1979a)?

The policy consequences of this issue, which ultimately ensured the continued predominance of peasant production at the expense of expanding privately owned, capital intensive agricultural production, would have lasting impacts on rural class structures and, more pointedly, the ability

18 Few chiefs continued to perform traditional roles at the local level due to fear of retribution by TANU officials. As Maguire (1969) notes, “many chiefs feared that anything they might undertake in the way of traditional functions – especially where large public gatherings were customary – might be misinterpreted by TANU and the government” (p. 336). 174 of political entrepreneurs to find the necessary capital inputs for opposition party building following the multiparty reforms of 1992.

As cited repeatedly in this section, prior to the 1967 Arusha Declaration, rural development greatly favored the expansion of the market economy. World Bank inspired village settlement schemes were constructed to act as ‘demonstration centers’ for encouraging peasants to work with greater attention to producing for the market by allowing market mechanisms to drive incentives (Raikes, 1975; Migot-Adholla, 1979). However, rural development up to 1967 also threatened to undermine the structural basis that ensured TANU’s ease of consolidation. Most notably, rural development yielded greater regional inequalities of wealth, and did little to undermine the economic position of the rural petty bourgeoisie. As the World Bank data illustrates in table 6.02, income inequality actually grew in Tanzania between 1959 and 1969.

This is especially true for the wealthiest 20 percent, whose share of income rose from 58.1 percent in 1959, to 63.3 percent in 1969, and for the poorest 20 percent, whose share of income fell from 5.2 percent in 1959, to 2.3 percent in 1969. Likewise, the Gini Coefficient also rose during the same time span (Jain, 1975).

The effects of the Arusha Declaration on rural development were to disrupt the growth of class and sectarian tensions and to reproduce an agrarian economy dominated by the peasant mode of production (Hyden, 1980). While the initial rural development fallout from Arusha appeared conservative19, over the course of the following decade, TANU engaged in some of

19Some figures have been cited by a number of authors as evidence of this conservatism. Government expenditures to rural development never increased from FYP1 to FYP2. At the same time, parastatal investment policy since Arusha appeared nearly blind to the emphasis on Ujamaa and rural development, highlighting the lack of a general strategy in developing relations between agriculture and industry (Clark, 1978; Boesen, 1979). While, by 1972, the parastatal sector was usurping some 71 percent of credit available from domestic financial institutions in order to develop capital intensive enterprises, almost no attention to promoting rural industry. Clark (1978) illustrates several cases, such as a fertilizer plant and a tire factory which imported raw materials rather than strategically exploiting domestic possibilities as a long-run promoter of rural capitalization. Finally, while the 175 the boldest attempts at “capturing” peasant loyalties and production, while keeping the growth of agrarian class contradictions in check. Clark (1978) points out that despite the lack of increase in government expenditures to rural development thanks to an increased reliance on foreign aid, the Second Five-Year Plan “has made real progress in the provision of social service to rural areas” (p. 321). The Second Five-Year Plan quadrupled the proportion of national expenditures dedicated to primary education, increased the funding to rural water projects by two-fold, and greatly expanded access to preventive healthcare throughout rural Tanzania.

These and similar expenditures led Bienen (1970) to conclude that “unlike many other African countries where leaders tell their people to ‘go back to the land,’ Tanzania has accepted the policy implication of this exhortation” (p. 422).

Table 6.02: Growth of Income Inequality in Tanzania Year 59 69 Population Percentile Household income Household income 0‐10 2.3 0.8 10‐20 2.9 1.5 20‐30 3.5 2.3 30‐40 4.2 3.2 40‐50 5.2 4.3 50‐60 6.2 5.8 60‐70 7.8 7.9 70‐80 9.8 10.9 80‐90 13.5 16.6 90‐100 44.6 46.7

95‐100 35.1 33.5 Gini Coeff 0.5282 0.5973 Source: Jain, 1975

The defining documents for rural development shortly after the formal Arusha Declaration, namely Ujamaa Vijijini (Socialism and Rural Development) and Education for Self-Reliance

share of industrial output in GDP composition was slated to rise from 10.7 to 13.7 percent during the Second Five-Year phase, subsistence production was projected to fall from 26.4 to 22.4 percent (Roe, 1970).

176 ushered in the most ambitious and widely studied attempt to restructure rural Tanzanian society from one characterized by scattered, family-based shambas, to one characterized by communal, production-based settlements.

The establishment of communal farmers following Arusha took place in two distinct phases.

The first phase was effectively a policy to encourage peasants to voluntarily move into Ujamaa village schemes or transform their existing settlements into Ujamaa villages. However, because peasant responses were slow20, in November 1973, Nyerere announced the government’s intensions to relocate the entire rural population into village settlement schemes by 1976.

Hence, between 1973 and 1976, the second resettlement phase emerged, largely referred to as

Villagization. In contrast to earlier Ujamaa policies, Villagization was compulsory and often quite brutal.

The actual number of people relocated during forced collectivization is striking. Hyden

(1980) notes that Villagization was perhaps the largest resettlement program in contemporary

African history, involving some 5 million people (p. 130). Barkan (1994) points out that between 1973 and 1976, some 80 percent of the rural population (those not living in villages) were resettled (p. 20). According to Buchert (1994), around 85 percent of the population was registered as members of an Ujamaa farm by the end of 1976 (p. 100). Mwakikagile (2004) cites that by 1981, some 91 percent of the Tanzanian population lived in villages (p. 68). Although official statistics may be somewhat misleading, in 1970, some 530 thousand were reportedly living in the nearly 1,900 villages, By 1975 there were a reported 9.2 million villagers living in the 6,400 villages (Mascarenhas, 1979, pp. 153-54).

20 There were a few successful Ujamaa efforts, most notability those carried out by the TANU Youth League, which transformed existing settlements into Ujamaa schemes. There were a few other cases where local petty bourgeois farmers were forced by peasants to vacate their land to make way for collective farms (Hyden, 1980, p. 101). 177

The vision of the ideal form of rural development among the TANU leadership stratum was the establishment of communal farms “where people live together and work together for the good of all” (Nyerere, 1966, p. 368) Specifically, communal farmers would modernize agricultural sector through cooperation and participation, and the utilization of the country’s most voluminous assets (land and labor) in ways that did not result in the emergence of a well- entrenched landed capitalist class and gross economic inequalities (Mascarenhas, 1979, p. 150).

The logic of Ujamaa Vijijini and Education for Self-Reliance was to reject private capitalization and the creation of differentiation in relations to production.

While resistance to Villagization was widespread, the most entrenched resistance was found in those regions were the yeoman peasants and petty bourgeois elements were the most consolidated. As a comparatively weaker class in Tanzania, the petty bourgeoisie were found primarily among the cash crop producing regions around Lake Victoria and Kilimanjaro. This class became the single biggest organized opposition to Ujamaa, and eventually the CCM during the onset of multiparty change. Boesen (1979) cites the case in Karagwe District (West Lake

Region), where peasants cultivating perennial cash crops complied with the demands for resettlement only after authorities threatened to call in the army (p. 137). Table 6.03 indicates that Ujamaa tended to be weakest in regions where surplus producing classes were strongest

(Shivji, 1976; Mascarenhas, 1979), most notable being the West Lake, Kilimanjaro, and

Morogoro regions.

From the perspective of the party-state, whether from the eyes of the ideologically committed members of the leadership stratum, or the materialist ambitions of the nascent

178 bureaucratic bourgeoisie21, the rural economic interests and bases of commercial and yeoman farmers had to be disciplined, if not altogether eradicated. Following Arusha, diverse interests within the top echelons of TANU and the state converged, which brought the need for dismantling local power structures into urgent focus. In comparative perspective, and from the perspective of reproducing regime tenure, the party-state dismantling of local power structures were perhaps one of the most successful policy endeavors in post-independence Tanzania.

Table 6.03: Number of Villages by Region 1969 1970 1971 1972 1973 1974 1975 Arusha 20 25 59 92 95 110 180 Dar es Salaam 25 53 Dodoma 40 75 246 299 336 354 388 Iringa 60 350 651 630 659 619 464 Kigoma 14 34 132 129 129 123 193 Kilimanjaro 7 9 11 24 24 14 16 Lindi 148 185 572 626 589 339 315 Mara 19 174 376 376 271 111 303 Mbeya 22 91 493 713 715 534 933 Morogoro 15 19 113 116 118 96 397 Mtwara 264 465 748 1088 1103 1052 773 Mwanza 10 28 127 211 284 153 106 Pwani 46 56 121 185 188 238 238 Rukwa 121 385 Ruvuma 26 120 205 205 242 180 315 Shinyanga 6 98 150 123 108 134 369 Singida 12 16 201 263 263 317 258 Tabora 41 52 81 148 174 156 324 Tanga 37 37 132 245 245 255 302 West Lake 21 22 46 83 85 77 72 Total 808 1856 4464 5556 5628 5008 6384 Source: Mascarenhas, 1979, p. 153

21I use the term bureaucratic bourgeoisie interchangeably with other terms like state bureaucrats and high level civil servants. Operationally speaking, the bureaucratic bourgeoisie are a wealth accumulating class occupying positions of stately authority at the intersection between international “neo-colonial” interests and domestic sources of surplus production. Because of their bourgeoisie ambitions, this class is cast in opposition to socialism, although in the case of Tanzania, is in alliance with the party leadership stratum in their mutual opposition to the rural petty bourgeoisie for the control over peasant loyalties and agricultural production (Shivji, 1976; Parker, 1979; Okoko, 1987).

179

Attempts at dismantling local power structures were pursued under the Decentralization reforms on 1972. Decentralization effectively extended TANU local control by, as described in a 2005 World Bank Report, “concentrating civil servants and party functionaries to the regimes and districts below” (p. 25). District Councils, which grew into institutions of entrenched rural petty bourgeoisie interests, were abolished and replaced with larger District and Regional

Development Councils with far less scope for action independent of the central government. At the same time, in order to neutralize regional and district developmental inequalities, the financial independence of these development councils was constrained by an increased reliance on funding from the central government (Hyden, 1980; Havnevik, 1993; Mukandala, 1998).

Attempts were increasingly directed toward neutralizing the rural bourgeoisie influence in national politics. Similar to trends found elsewhere in Africa, by the early 1970s in Tanzania, it was increasingly normal to expel critical members of parliament who challenged the fundamentals of Ujamaa and Villagization. More importantly, the National Executive

Committee took an increasing interest in screening slates of potential state and party officials.

Elections were more closely scrutinized to ensure that selections did not end up placing petty capitalists in positions within the parliament and local government.

Perhaps the single most important economic institution underpinning the rural petty bourgeoisie was various cooperative societies which proliferated under the latter stages of colonial rule. As described by Hyden (1980), since cooperatives controlled “significant amounts of capital” and thus “provided a base for petty bourgeoisie power independent of the state”, cooperatives “made up the most important middlemen in rural areas” (p. 132). Destroying the material base for this independence was seen as a key to neutralizing potential political threats throughout rural society. The 1976 Operation Maduka (Shops) attempted to replace privately

180 owned shops, including cooperative shops, with communal Ujamaa shops (Hyden, 1980). The bureaucracy was increasingly charged with the task of hiring personnel formulating cooperative policies (Holmquist, 1979, p. 142). By 1976, these semi-independent cooperative authorities, such as the Bukoba Native Cooperative Union and the Kilimanjaro Native Cooperative Union, were abolished altogether, being overtaken by state marketing authorities and CUT22 (Barkan,

1994; Okema, 1996).

At the same time, the growth of parastatals following the Arusha Declaration increasingly interceded between local producers and the market in order to slow the exodus of profits out of the country and expand investment in production (Coulson, 1982; Havnevik, 1993). Price controls on agricultural products and the co-operative retail shops were established. Marketing authorities (cotton, tobacco) and livestock, fishery, and small industry authorities bought and sold to and from producers. Whereas in 1964, new government investments were largely confined to privately controlled manufacturing and tourism parastatals, by 1969, new government investments were directed toward a mixture of privately and state controlled parastatals in manufacturing, tourism, commerce, agricultural, and finance (Clark, 1978; Okoko,

1987). All-in-all, parastatals and price controls acted to undermine any lucrative returns that may have been attained by the rural petty bourgeoisie..I Instead they became a principle economic force for enhancing the relative position of the party ideologues and state bureaucrats

(Boesen, 1979).

22 At least two additional policies are worth mentioning. First, many entrepreneurial families which migrated to Ujamaa villages supplemented their incomes through petty trade, sewing, and tool-making. After 1973, private individuals could no longer engage in small-scale industries on Ujamaa villages. Secondly, the Villages and Ujamaa Villages (Registration, Designation and Administration) Act (1975) placed all village lands under the ownership of Village Councils, which assumed a sort of “corporate personality” to regulate economic affairs, primarily the usage of land (Tenga, 1996) 181

As a summary; rural development policies in Tanzania generally eradicated the economic and institutional basis that sustained peripheral power centers in other countries, while

Villagization utterly failed to capture the imagination and productive capacity of the peasants and weakened the overall agricultural production output. This had profound consequences for governance in Tanzania generally, and CCM dominance in particular. First, by the 1980s, the peasant mode was still the primary engine of economic growth, thus reducing the foreign earning capacity of the Tanzanian economy, and increasing the country’s aid dependence. This point will be taken up more directly in the next chapter. Secondly, rural development policies following the Arusha Declaration ensured that, by the time multiparty change was in the pipeline, elements from the rural petty bourgeoisie were in no position to consolidate sufficient resources for building opposition parties with the wherewithal to challenge the CCM. Indeed, as evidence of the importance of the petty bourgeois impulse in opposition politics, those parties that did emerge during the early 1990s were largely confined to those regions around Lake

Victoria and Kilimanjaro where the economic base of the petty bourgeoisie was the most developed.

Nationalization

Within a few years following independence, African leaders almost universally under the auspices of socialism sought to exert state control over major sources of production and accumulation, thus providing the critical material base for constituting a rising bureaucratic bourgeoisie, relative to other social forces. Meillassoux (1970) describes nationalization in Mali, where

Socialism permitted [the bureaucrats] to put the bureaucracy into the position of a managerial board of a kind of State corporation. The creation of a public

182

economic sector thus brought the bureaucrats into competition with private business, particularly with the native traders (p. 106).

Tanzania was no exception, although the rationale behind nationalization was more directly an attempt to address the dialectics associated with the growth of a powerful commercial class.

Indeed, differences in rationale within TANU did not center around whether or not to nationalize, but over the justification for nationalization. For doctrinaire socialists, state control was vital for giving a vanguard greater control for steering development policy in the direction of indigenously owned industrial development. For African socialists, the most important premise behind nationalization was the growth of disruptive class antagonisms, which would likely ensue between labor and the commercial bourgeoisie where private ownership prevailed.

Despite this disagreement, both African and scientific socialists could agree that post- independence trends toward private capitalization dominated by foreign investments needed to change.

As pointed out earlier, while these ideological tensions within the leadership stratum were easily patched over after the sidelining of the scientific socialists in the late 1960s, a more important conflict was emerging between the party leadership stratum and the nascent bureaucratic bourgeoisie. While these more technocratically oriented senior civil servants could agree with the party ideologues over the necessity of reigning in the bases for the reproduction of the rural petty bourgeoisie, they did fundamentally disagree over the ideological justification for nationalization. For the bureaucratic class, nationalization would be seen as a source for wealth accumulation similar to their capitalist counterparts throughout the developed world. For the African Socialist ideologues, nationalization was quite simply a way to contain the formation of class structures, which would upset national unity and socialism. However, both could agree on the need for nationalization

183

Similar to the patterns of rural development shortly following independence, the development of Tanzania’s commercial sector within the first and second development plans were relatively business friendly. Prior to 1967, foreign private interests were guaranteed protection through the Foreign Investment Protection Act of 1962. With these assurances, foreign investors rolled in. An oil refinery was established by the Italian government firm, ENI, the newly erected Kilombero Sugar Plant was managed by a Dutch firm, and a cement plant was opened by the London-based Associated Portland Cement Company (Shivji, 1976, p. 164). With the exception of the addition of a “People’s Plan” amendment in the spring of 1962, which emphasized grass roots participation in economic development through a hierarchy of regional and openly elected local level development committees (Maguire 1969, p. 325), development plans up to 1967 stressed the importance of private sector agricultural development, industrial development based on import substitution, infrastructure investment, and a heavy reliance on foreign revenue sources (Shitundu, 2002). The catchword for the technocrats within the

Directorate of Development and Planning was the “mixed economy” approach, since the plan continued to encourage private enterprise, alongside planning in sectors controlled by the state and the cooperatives (Bienen, 1970, p. 299).

Changes to development strategy followed from soured aid relations with Britain and West

Germany, and only lukewarm support from the United States23, all of which brought into sharp focus, concerns about the political merits of a reliance of foreign and private investment. First, the concept of “self-reliance” gained momentum, stressing the importance of development

23 Generally, the west was increasingly irritated by Tanzania’s failure to declare itself as a committed ally of the West and its increasingly friendly relations with communist countries. The acceptance of Chinese military assistance was especially troublesome. The opening of an East German mission in Dar es Salaam resulted in the suspension of aid from West Germany. Tanzania’s principled stand against apartheid, and respectable assistance for liberation movements in Mozambique and Rhodesia were met by the suspension of aid from some of Tanzania’s most important contributors, especially Britain. 184 through the maximization of local resources, most importantly land and labor, while minimizing the need for state assistance and national development by minimizing dependence on external support (Bienen, 1970, p. 408).

Secondly, aid portfolios shifted from a heavy reliance on western, pro-capitalist countries, toward ones more socialist friendly. From fiscal year 1965-66, the United Kingdom dispatched some 17.5 million TSh in development assistance, while the Federal Republic of Germany contributed slightly more than 4.3 million. By FY 1968-69, no assistance was offered by the

UK, while West Germany’s contribution sank to 0.6 million. The only pro-capitalist country to show continued (actually increasing) desire to fund Tanzania’s development was the United

States. At the same time, Sweden provided 1.6 million shillings in 1965-66, and Denmark provided none. By 1968-69, each social democratic country was providing 20.8 and 3.3 million respectively. China also stepped up its commitment, although the outlays vary considerably from year to year. Between FY 1961-62 and 1964-65, China dispersed just under 6.0 million in development assistance to Tanzania. By contrast, from FY 1965-66 to 1968-69, China’s contributions swelled to 75.9 million. By 1974, China, which charged no interest on its loans to

Tanzania, was owed nearly 2 billion shillings by the United Republic, while the next largest creditor, the World Bank, was owed just over 800 million (as cited in Okoko, 1987, pp. 190-98).

Finally, and most importantly for this piece, nationalization was politically palatable. Not only did nationalization coincide with the class interests of the bureaucratic bourgeoisie, whose technocratic expertise was increasingly important for those within the corridors of decision- making power of the Party’s National Executive Committee. Nationalization was also certain to garner public sympathy among workers and those proposing policies of Africanization, since the

185 net losers of nationalization was the nascent commercial class dominated by Asians and

Europeans (Hyden, 1980; Okoko, 1987).

Hence, the “major means of production, distribution, and exchange”24 were brought under state control25. Taking place within a few weeks in the first half of February 1967, nearly all private sector manufacturing enterprises were taken over by the National Development

Corporation (NDC), which was established in 1964 as a holding company for state-owned enterprises (Hyden, 1980). Banks, except for the Cooperative Bank, were nationalized, along with food processing companies and sisal firms. Under the Acquisition of Building Act, the

President may acquire any building worth more than 100,000 shillings, to be vested with the

Registrar of Buildings without the burden of compensation.

The general aversion to the private sector was clearly visible in other policy domains. For starters, in the Second Five-Year Plan, roughly 16 percent of industrial investment was to be undertaken by the private sector, a stark contrast to the development plans in the First Five-Year

Plan (Bienen, 1970, p. 412). Other policies, such as the National Industrial Licensing and

Registration Act (No. 10 of 1967), Tourist Agents Licensing Act (1969), and the Business

Licensing Act (No. 25 of 1972), greatly increased the regulatory environment and the red-tape involved in license acquisition. Equally important, regulations increased the ability for

24 The definition of “major means of production, distribution, and exchange” following the second five-year plan included “land, forests, mineral resources, water, oil, electricity, communications, banks, insurance, import- export trade, [and] wholesale businesses”. It also included manufacturing, such as “steel, machine-tool arms, motor-car, cement and fertilizer factories”, along with the textile industry (Nyagetera, 1995, p. 7).

25 While most of the major means of production were nationalized, a few sectors remained in public-private partnerships, including food, beverage, furniture, paper, printing, and chemical manufacturing and the production of metal and petroleum-based products (Nyagetera, 1995, p. 7). In cooperation with foreign partners, the NDC maintaining half the holdings of Williamsons Diamonds, Portland Cement, and the Tanzanian Publishing House, and just over half the holdings of firms such as Tanzania Breweries and Tanganyika Packers, and British American Tobacco Company (Nursey-Bray, 1980, p. 65). A handful of other sectors remain predominantly private, including textile and apparel manufacturing and the production of non-metallic and non-petroleum mineral products (Nyagetera, 1995, p. 7).

186 bureaucrats to grant favors in exchange for loyalty (Nyagetera, 1995, p. 8), constituting a major mode of wealth accumulation among the nascent bureaucratic bourgeoisie.

In Malawi, Kenya, and Zambia, the post-colonial developmental trajectory nurtured the growth of a vibrant commercial capitalist class (Barkan, 1979a) that, despite clientele connections to the state, could (and indeed did) muster together resources for launching opposition parties. In Tanzania by contrast, the net effect of this less business friendly environment generally had two consequences. First was the eradication of the development of a commercial bourgeoisie, dominated by a section of the population in which Africans were hostile to, namely the Asian community. At the same time, the major means of production were now placed under the direct supervision of the party, and would effectively weaken class-based or racially defined cleavages, the economic base for constituting a politically viable social force.

Secondly, nationalization effectively meant that those bureaucrats which controlled the regulatory leavers also had an enormous potential for accumulating wealth. Indeed, such wealth accumulation was an obviously complicated problem among many Tanzania’s neighbors, serving as an object for opposition among those entrepreneurial politicians outside the spatial reach of state clientelism. In short, an extravagant bureaucratic lifestyle was a serious source of social contention among the poor and among those regime outsiders wanting in, most notability seen in the anti-corruption campaigns of the late 1980s and 1990s. The potential for public sector wealth accumulation was exactly what the Leadership Code, to be discussed in the next section, targeted.

187

Party/State Leadership

Despite the moves to nationalize the economy, and undermine the material base of commercial and petty bourgeoisie formation, there were clear tendencies toward the bourgeoisie development in other areas of society, ones perhaps more threatening to African Socialism.

First, following independence throughout Africa, bureaucratic power increased significantly, and with this, public offices became a more prestigious source of personal accumulation. High placed government officials acquired expensive cars, additional properties, and covered their bodies in expensive suits made from the finest European tailors. Arrighi and Saul (1968) cite a study on Ghana, Guinea, and Mali which demonstrates the “heavy weight of bureaucratic expense and conspicuous urban consumption” in state budgets (p. 160). Mamdani (1975) discusses nationalization in Uganda under Obote as a road for private accumulation among the bureaucratic bourgeoisie, in the form of “corruption” rather than “profit” (p. 51).

As mentioned earlier, in Malawi, bureaucrats were encouraged to engage in commercial agricultural enterprises. Banda, in a series of acquisitions through the Malawi Press, retained control over a conglomerate of some 45 companies, which, by 1981, constituted the “largest and most extensive private sector enterprise in Malawi” (Kaunda, 1998, p. 55). In contrast, Nyerere had always displayed a strong aversion to the type of habits found among African bureaucratic elite. During a 1959 declaration in the Legislative Council, Nyerere predicted the arrival of post-

1967 reforms in the following passage:

We are not going to enter the government to make money. We are condemned to serve, to wage a war against poverty, disease and ignorance. I warn our future civil servants that they must think in terms of our country and not compare themselves with anyone outside his country…We shall slash the salaries of local people, if necessary, we shall slash them hard (as cited in Pratt, 1976, p. 218).

188

While personal and class aggrandizement were far less prevalent in Tanzania, the nationalization of the economy would surely have strengthened the hand of this nascent bureaucratic bourgeoisie in relation to party ideologues, yielding space for bureaucrats to use political appointment as a source strengthening personal political loyalties and personal wealth.

In short, not only did the prospects of lavish bourgeois lifestyles appear contrary to the spirit of socialism, but, like class development in the commercial sector, capitalization within the public sector constituted a source of political power through patronage with the capacity to threaten the very core of African socialism.

These trends were most evident in rural society, as the pre-Arusha development agenda strengthened the relative mutual position of the rural petty bourgeoisie who assumed public offices and party positions at the local level and inside the Parliament. These bourgeois impulses meant that elitist tendencies were infiltrating the party, a fact that certainly contributed to the high turnover rate in the 1965 election (Harris, 1967; Maguire, 1969). This popular rejection of the tendency among politicians to treat public office as a means toward economic security meant that the relationship between TANU and the peasants was potentially threatened without the appropriate checks against excessive consumption habits among party officials and bureaucrats

(Bienen, 1970, p. 433).

The Leadership Code was the African Socialist solution. As a component of the Arusha

Declaration, the Leadership Code was a protocol limiting the bourgeoisie tendencies increasingly found throughout the party and bureaucracy. The writing was on the wall when in

1966, Nyerere and his ministers took a 20 percent pay cut, district and regional party and government heads took between a 10 to 15 percent pay cut, while higher paid civil servants

189 received 10 percent less pay. At the same time, a number of perks which benefited MPs were also scratched, including the low-cost loans for cars (Bienen, 1970).

After the Arusha Declaration, leadership restrictions were formally legislated. The

Commission Act No. 6 of 1973 established the Commission for the Leadership Code26, which was charged with assisting the NEC with enforcement (McHenry, 1994). In 1969, the Code became part of the TANU Constitution; in 1977 it was made part of the CCM Constitution and the Union Constitution. In 1975, The Leadership Code was made applicable to all TANU members, while the CCM Constitution of 1977 made it clear that prospective and current party members would have to demonstrate knowledge and adherence to party principles (Hyden,

1980; McHenry, 1994).

The Leadership Code was drafted with a least two broad goals in mind. First, the Code intended to stamp out capitalism, minimize structural differentiation resulting from economic development, and put the brakes on tendencies for rural petty bourgeoisie to turn to the party and state as a source for wealth and tenure (Holmquist, 1979, p. 142). As the document states,

“Every TANU and government leader must be either a peasant or a worker, and should in no way be associated with the practices of capitalism or feudalism.” Hence, party and government

26 The provisions of the leadership code were as follows: 1) Every TANU and government leader must be either a peasant or a worker, and should in no way be associated with the practices of capitalism or feudalism. 2) No TANU or government leader should hold shares in any company. 3) No TANU or government leader should hold directorships in any privately owned enterprise. 4) No TANU or government leader should receive two or more salaries. 5) No TANU or government leader should own houses which he rents to others. 6) For the purposes of this resolution the term “leader” should comprise the following: members of the TANU National Executive Committee; ministers, members of parliament; senior officials of organizations affiliated to TANU; senior officials of parastatal organizations; all those appointed or elected under any clause of the TANU constitution; councilors; and civil servants in the high and middle cadres (In this context ‘leader’ means a man, or a man and his wife; a woman, or a woman and her husband).

190 leaders were not permitted to privately own profit-making properties, including rental properties or company shares. Saul (1974) notes that real restrictions were placed on lavish lifestyles among civil servants and party leaders, making bureaucrats and leaders in Tanzania notoriously envious of their Kenyan counterparts. He goes on to cite that the Leadership Code represented an initial victory for a “progressive wing of the petty bourgeoisie” (p. 362).

Additionally, the aspirations of the Leadership Code cannot be fully appreciated outside an understanding of the corresponding context of the growth of a more extensive and socially active party committed to social transformation and enforcing, sometimes coercively, the commitments were laid down after the Arusha Declaration. For starters, TANU leaders were always committed to making the Party the center of political life throughout the countryside.

The infamous Ten-Cell system, whereby each house in the United Republic belongs to a unit of ten houses, is to constitute the basic organizational unit and provided a mass membership base, with little or no restrictions on party membership.

By 1967, the leadership stratum within TANU increasingly understood the purpose of the party, with its elaborate cell system, as instilling a broader sense of national identity and commitments to socialism. As pointed out previously, primary and secondary education was transformed into institutions for citizenship and political education, rather than simply education for market consumption. The National Youth Corps (National Service), which was established in 1963 by the Ministry of National Culture and Youth for assistance in road-building, agricultural project labor, and civil defense and surveillance, was increasingly used to instill a sense of egalitarianism, socialism, and patriotism and help articulate a sense of common purpose against the enemies of imperialism and apartheid. At the most extreme, the Youth Corps served as a source of loyal recruits for the Jeshi la Mgambo (People’s Militia) and violence against

191 those acting in ways to threaten socialism27 (Brennan, 2006). Albeit a sentiment not shared by all, Mwakikagile (2004), who gives accounts of his personal experience in Ruvu National

Service, concluded that national service was “one of the most successful policies which fostered egalitarian values among the elite, including us, many of whom felt they were better than the poor and illiterate peasants and workers…” (p. 29).

More specific to party leadership, changes were also reflected in membership criteria. While the original focus of TANU was to attract as many members as possible in order to stave off challenges from opposition during earlier elections, the leadership code was an attempt to spread socialist values among party leaders by providing a minimum standard for conduct, and hence move away from the model of a party with no membership requirements (Bienen, 1970;

McHenry, 1994).

In short, the Leadership Code was to predict the eventual ideological training of all party leaders, to make the party a model for socialism, to instill a broader commitment toward socialist transformation, and to ensure that class divisions never became a point of contention in

Tanzanian politics. More important for the longevity of the CCM, the leadership code meant that popular resentment against lavish lifestyles on the eve of multiparty change were more minimal than resentments found elsewhere. At the same time, it ensured that the growth of economic bases within the party and the state with the capacity for independent action were curtailed. Whereas incumbents in Kenya, Malawi, and Zambia that defected to an opposition position were previously allowed to amass considerable personal wealth during their tenure,

27Operation Vijana, which took place in October 1968, was a public disciplinary campaign inspired by some of the more radical socialists, such as the Minister of Youth and Culture, Chedial Mgonja, and TYL chairman, Lawi Sijaona. Reflecting some of the sentiments of the Chinese Cultural Revolution, Operation Vijana was an effort to advance socialist transformation by enlisting the TYL to police against “lumpen and renegade practices.” This included a ban on bourgeois clothing, such as mini-skirts, wigs, and tight trousers, attacks on laziness and corruption, and political education as a means for staying vigilant against the forces of imperialism (Brennan, 2006, p. 238). 192 those political entrepreneurs who broke ranks with the CCM in efforts to stake out their own political ambitions as opposition party leaders, quite simply, had little only modest amounts of financial resources to take with them.

Conclusion

This chapter has demonstrated two broad themes in relation to post-independence Tanzanian development policy. First, at the time of independence TANU was not in short supply of ideologues committed to ambitions of a comparatively articulated and somewhat pragmatic brand of socialism. Secondly, in conjunction with those structural conditions defined in chapter five, during the 1960s TANU was in a position by which to implement African Socialism. The interests of state bureaucrats, which at the time were a relative nascent class, and party ideologues converged on the need to break the institutional and economic base of the rural petty bourgeoisie. Given the fact that peripheral economic development left the rural petty bourgeoisie comparatively weak, most simply decided to join TANU’s ranks rather than risk an all out confrontation and isolation. Likewise, the scientific socialists found almost no popular basis for constructing their opposition to African Socialism.

This chapter has also demonstrated that policy products of African Socialism were constructed from a world view which emphasized unity and familyhood rather than social division. Language and education policies were directed toward fostering a powerful sense of national identity and egalitarianism. Rural development, nationalization, and leadership code, directed toward the production of a socialist society where wealth differentials were minimal and where the state emerged as the undisputed singular economic actor. In short, African

Socialism had the ambitions of undermining the growth of macro-structural divergence.

193

This chapter has however, raised as many questions as it answered. First, just because party ideologues had the ambitions of undermining structural divergence does not mean they were successful in doing so. Most of the chapters in part three empirically examine the question of the macro-structural consequences of policy action under African Socialism. Secondly, one might expect TANU/CCM’s overt commitments toward African Socialism and state economic leadership to dialectically interface with the economic liberalization demands of the 1980s and the political liberalization demands of the 1990s. This indeed was partially true. However, as the next chapter demonstrates, the pursuits of African socialism also meant that policy adjustments could more easily be made without a corresponding reaction from supporters and opponents of reform inside or outside the regime.

194

CHAPTER SEVEN AID AND THE REFORM STIMULUS

“It is the reformer who is anxious for the reform, and not society, from which he should expect nothing better than opposition, abhorrence and even mortal persecution”

Mahatma Gandhi , 1927

”Who can say how much is endurable, or in what direction men will seek at last to escape from their misfortunes?”

John Maynard Keynes, 1920

Within a few days of the start of my first trip to Tanzania, I was struck by the level of economic decay throughout nearly every part of Dar es Salaam, indeed, almost literally so. On my return from watching a rage of flames devour an old German colonial building along the Dar es Salaam water front, a power cable broke free from its dilapidated post, only to land in the road where I had past only moments ago. After returning to the scene the following day, I realized that there was nothing extraordinary about the condition of the dilapidated poll.

Looking around Dar es Salaam in the late 1990s, a large minority of the lighting and electric polls could be seen with perhaps a 15 degree plus lean, resulting in overly stretched or overly slack cables hovering over similarly dilapidated roads and sidewalks.

Like most sub-Saharan countries, by the late 1970s and throughout the 1980s, macro- economic conditions in Tanzania were in a serious recessionary state, only beginning to recover in the 1990s. Aside from a global system of trade which relegated the Third World to the supply of labor intensive primary outputs, in Tanzania specifically, the socialist path carved an environment less friendly to the accumulation of capital beyond the reaches of the state, resulting in the movement of private capital to neighboring countries and the overall decline of capital utilization and productivity (Shivji, 1976). Production of textiles, fertilizers, ropes, batteries, and beer declined precipitously during the early 1970s (Hyden, 1980, p. 174).

Likewise, food production also dropped, largely as a consequence of drought. By the 1974-75, maize was imported at a rate 20 times greater than in 1966-67, while rice imports increased 10- fold. Wheat imports also stood some 3.4 times greater in 1974-75 than in 1967-68 (Hyden,

1980, p. 141).

Figure 7.01: Gross Domestic Product in Selected Countries

Source: International Monetary Fund, 2007

Throughout Africa, the overall macro-economic decline was far more acute during the

1980s due to continued recession in global commodity prices and the fallout from rising oil prices in the 1970s. Figures 7.01 above and 7.02 below describe Tanzania’s relative decline throughout the 1980s and into the 1990s. According to the World Bank 1987 World

Development Report, while overall production in agriculture, industry, and manufacturing expanded during the 1970s, between 1980 and 1985, agricultural production grew less than 1 percent, while industry and manufacturing shrank 4.5 and 4.6 percent respectively. The long term decline in export earnings was even more pronounced in the 1980s. While the country

196 exported slightly more than one dollar for every dollar in imports in 1967, by 1985, Tanzania was exporting only 29 cents for every dollar in imports (McHenry, 1994, p. 171). In short,

Tanzania, not unlike most other African countries, was rapidly running up a balance of payments deficit.

Figure 7.02: Annual Percentage Change in GDP per Capita

Source: International Monetary Fund, 2007

The ideal of self-reliance, one of the hallmark catchwords of African Socialism, quickly lost any sense of relevance as the reality of economic decline pushed Tanzania into a greater state of reliance on international aid. This reliance in turn became an asset for international lenders and a liability for the ruling elite in more ways than one. Not only did debt service constitute a growing economic obligation, but the continuation of aid increasingly required economic policy adjustments. By the end of the Cold War, the international community turned its attention to political reform.

The scope of the third part of my work, which spans from this chapter through chapter twelve starts here with an exposé of the macro-structural economic consequences of

Tanzanian’s Arusha Declaration development trajectory and the juxtaposition of those

197 consequences to the domestic and international pressures for multiparty reform. The goal of this chapter is two-fold. First, I will demonstrate that, while economic crisis and a correspondingly greater reliance on aid subjected the CCM to greater potential for international reform pressures, policies undertaken during the single-party era also yielded macro-structural patterns with comparatively fewer tensions between social forces, thus making it easier to engage in political reform without the risk of unleashing a wrath of latent discontent. Secondly, I will briefly demonstrate that the rising materialist ambitions of the younger and more technocratic bureaucratic bourgeoisie relative to the ideals of the older socialist ambitions of party ideologues place the CCM in a position where it could easily shed its previous socialist commitments and credibly engage with international economic and political reform pressures.

Indeed, what I will demonstrate throughout the remaining sections of my work, is that the CCM, since the late 1980s can hardly be considered a champion of African Socialism but instead serves as a vote getting machine for the reproduction of bureaucratic bourgeoisie class interests.

The Reform Stimulus

With economic decline and high volumes of debt only becoming more acute throughout the

1980s, the post-independence development trajectories were brought to bear on the prospects of continued incumbency. In Tanzania, while the hegemony of the CCM appeared to depend largely on their relative ability to manage the country’s resources and limit the scope of political competition, the increasingly strident pressures to liberalize under the threat of aid stoppages, translated into a choice between accepting reforms and a correspondingly diminished capacity to manage resources and limit competition’s scope, versus accepting the domestic status quo of a state run economy and single-party rule along with a correspondingly diminished bequeathment

198 of aid revenues which would have posed an equality compelling threat to the survival of the

CCM.

The source of the contradiction between these choices was long in the making. Despite the rhetoric of self-reliance, the period between 1970 and 1980 has been classified as the “first aid boom” as donors, including the Scandinavian countries, China, and the World Bank, were attracted to Tanzania’s ideological or experimental ambitions (Clark, 1978, p. 193; Nyagetera,

1995, p. 15). This renewed interest, when paired with the reproduction of a non-monetarized peasant sector unable to provide goods or foreign earnings to keep pace with domestic consumption, resulted in enormous trade imbalances. As a result, direct foreign aid in the

Second FYP increased five-times over aid flows prior to Arusha. Furthermore, despite the policy of Africanization, by the end of the 1970s, Tanzania’s dependence on the technical inputs of foreign experts actually increased (Freund, 1981, p. 486). Table 7.01 provides a glimpse of the growth of aid in relation to expenditures and investment. The overall picture is that, while a sharp decline in aid during the period of 1980 to 1984 resulted from policy disagreements with lenders such as the IMF and World Bank, the overall dependence on foreign aid has increased right up to the period of political liberalization (Nyagetera, 1995, p. 15).

Table 7.01: Foreign Financing During Single-Party Era Foreign aid as a Percentage of: Period Total Development Total Government Parastatal Fixed Capital Expenditure Expenditure Formation

1965‐69 30.4 7.1 23.5 1970‐74 41.0 13.0 23.5 1975‐79 45.5 18.5 30.4 1980‐84 33.5 13.4 46.5 1985‐90 48.0 24.6 51.7 1991‐93 48.0 8.9 25.6 Source: Nyagetera, 1995, p. 16

199

Assuming that aid dispensations are a measure of dependence, and that a greater level of dependence on international lenders equals a greater source of international leverage over reform agendas, based on the data provided below, it is logical to conclude that leverage international actors had over the CCM appears to have been abnormally high. In terms of outstanding debt, table 7.02 indicates that Tanzania was among Africa’s most heavily indebted countries during the 1970s and 1980s. In fact, from 1971 to 1975, Tanzania owed more debt to international aid providers than any other African country. Between 1970 and 1989, Tanzania owed on average 4.3 billion dollars1, compared to 2.8 for Kenya, 3.2 for Zambia, 0.7 for

Malawi, and 1.5 for the average African country (World Bank, 2007b). When accounting for the relationship between aid and development, a similar picture emerges. In 1990 in Tanzania, aid inflows were greater than the rate of gross capital formation (GCF). The ratio of aid to GCF stood at 105 percent.

Table 7.02: External Debt - Percentile Rank 1985 1980 1975 1970 Botswana 28% 15% 38% 21% Gabon 51% 80% 82% 51% Kenya 84% 88% 85% 90% Malawi 47% 59% 59% 56% Tanzania 95% 95% 100% 72% Togo 44% 66% 41% 41% Zambia 86% 85% 92% 97% Source: World Bank, 2007b

Similarly, in Malawi, the ratio of aid to GCF stood at 115 percent. By contrast, the same measure in Kenya for 1990 was 57 percent, while Zambia scored 84 percent2. During the same year, nearly 29 percent of gross national income (GNI) in Tanzania was accounted for by aid.

1 Current U.S. Dollars.

2 It is noteworthy that the ratio of aid to GCF in Zambia drew dramatically by 1991, as lenders felt more comfortable about rewarding docility. 200

The same ratio was comparable to Malawi, at 27 percent, but far higher than Zambia and Kenya, which stood at 16 percent and 14 percent respectively. The average GNI to aid ratio for Africa stood at 17 percent (World Bank, 2007c). In short, all other things being equal, the leverage that international actors had over political and economic policy in Tanzania were higher than just about any other sub-Saharan country.

With a dwindling tension around international Cold War rivalries, the actual extent to which this leverage was largely depended on the degree to which regimes resisted reform in the face of rising domestic discontent. As Moi’s repression against venues of dissent grew ever more naked by the late 1980s, protests among elites and throughout the streets of Kenya’s major urban areas also grew. As described by Chege (1994), the contention which provoked protests centered around objections to the President’s “personal rule, his government’s ethnic oriented policies, and the patronage network of the regime” (p. 48). As well placed leaders, such as Kenneth

Matiba, Charles Rubia, Paul Muite, and James Orengo, along with the organizations like the

National Council of Churches of Kenya (NCCK) and the Law Society of Kenya (LSK), chimed their anger, the agenda among Moi’s opponents rapidly moved from one of reforming the open

“queue-voting” system which blocked so many of these leaders during the rigged 1988 election3, toward a broader goal of repealing Section 2A of the Constitution (Wanjohi, 2003).

Confrontations began to climax after high-level government officials were implicated in the

February 1990 murder of Minister of Foreign Affairs and International Cooperation and anti- corruption investigator, Robert Ouko, along with the June 1990 arrest and detention of

3The 1988 election was based on a system whereby KANU party members voted by openly queuing behind the representative of their chosen candidate, thus exposing the voters’ preferences and subjecting them to fear of reprisal for voting for the “wrong” candidate. As a consequence, the 1988 election was widely dubbed as rigged and many of those outspoken members of Parliament, such as Charles Rubia, Kenneth Matiba, and Martin Shikuku, lost their seats (Throup & Hornsby, 1998, pp. 42-4).

201 multiparty supporters such as Kenneth Matiba, Charles Rubia, and Raila Odinga, and the shooting deaths of 28 dissidents during a pro-democracy rally in July 1990. In response, at a meeting in Paris in November of 1991, Kenya’s bilateral aid and trading partners resolved to suspend balance of payments support for half-a-year (Kanyinga, 1998; Throup & Hornsby,

1998). The sheer size of the protests, and organizational skills and financial resources of behind this self-described pro-democracy movement, when backed by the suspension of aid following the Paris meeting, forced Moi to unenthusiastically agree to the revocation of Section 2A of the

Constitution in December of 1992, thus paving the way for multiparty elections (Akivaga,

1992).

As described by Posner (1995), aside from a rejection of Banda’s continued authoritarianism, the principle grievance in Malawi was directed toward “Banda’s estate- focused and patronage-based development strategy,” which led to an increased “disparity between the middle classes and the poor” (p. 137). When, in March 8, 1992, eight Catholic

Bishops penned a letter calling on Banda to recognize that Malawian society was in desperate need of political change, the door had flung open to a flood of mass demonstrations, as students and Unions displayed their support for the Bishop’s calls. Whatever veneer of legitimacy or credibility the MCP and Banda still claimed was burned up when 40 textile workers were killed in a May, 1992 strike. In what was at one time a cozy relationship between the West and

Banda’s pro-Western rhetoric, were soured as human rights organizations drew attention to the suppression of political demands and harsh prison conditions under the regime’s autocratic rule.

As mass opposition continued, and opposition media flourished, Banda finally called for a referendum on multiparty change in October, 1992 (Posner, 1995; Ihonvbere, 1997).

202

The domestic push for change in Zambia was arguably even more intense, undoubtedly due to the gravity of Zambia’s economic misfortunes. As described by Mbao (1996), “spear-headed by the labor movement, university and college students, pressure groups, the churches and professional bodies such as the Economic Association of Zambia and the Law Society of

Zambia” (p. 9), the massive popular energy directed toward repealing Article 4 of the 1972

Constitution was fueled by years of rapid economic decline and resentfulness toward one-party rule (Baylies & Szeftel, 1992). The coup attempts of 1988 and 1990, along with the massive food riots in 1991, in conjunction with international pressures, forced Kaunda to move quickly toward multiparty reforms.

By contrast, the color of the atmosphere in Tanzania as these events unfolded in neighboring countries was quite calm and far less rancorous. Less evident corruption thanks to the Leadership Code, a lack of political prisoners, and weak ethnic allegiances ensured that those tensions which played out in neighboring countries, failed to transpire in Tanzania.

(Asingo, 2003, p. 25) Thus, donors felt little pressure to threaten the CCM (Chege 1994). While calls for greater openness and more political competition were constantly biting at the edges of contention, calls which broadly rejected existing political leadership were comparatively less prevalent in Tanzania. As Chege (1994) points out, Mwinyi and Nyerere generated far less

“opprobrium” from the opposition when compared to countries like Kenya (p. 48).

During the early days of Tanzania’s reform, the opposition was an unambiguously exclusive affair. The 1990 Steering Committee for Multiparty Reform, organized by Fundikira, was a rather small clique of well-to-do individuals with no solid grassroots following. The almost anti-climactic formation of the first opposition party by James Mapalala in November,

1991 had few of the energies found around the formation of Kenya’s Forum for the Restoration

203 of Democracy (FORD), Zambia’s Movement for Multiparty Democracy (MMD), or Malawi’s

United Democratic Front (UDF). Instead, it was Nyerere, when announcing his concerns about the failures of single-party rule, which brought discussions of multiparty change into a popular political discourse. Shortly afterwards, a constitutional commission was formed and charged with investigating the merits and popular support for multiparty change. Despite the fact that the

Commission found 77 percent of the surveyed Tanzanians in support of the continuation of one- party rule, the Commission and the National Executive Committee of the CCM, announced their support for multiparty elections.

By comparison, Zambians, Kenyans, and Malawians stood at a precipice overlooking the convergence of material and ideational tensions which were nurtured under single-party rule. As the logic of patron-client capitalism gave way to growing wealth differentials, which, coincidentally mapped onto territorial, linguistic, and religious cleavages, the coercive apparatus was increasingly seen as the most viable route for buttressing social control amid rising social tensions. The price of this contradiction was registered when the existing logics of social control could not easily be reconciled with international and domestic calls for multiparty reform, nor could existing patterns of social control be managed once the green light for reforms was given.

In short, the anterior regimes simply lost the capacity to contain or manage society’s contradictory social forces, as pressures for change exploded into waves of expressed hopes, aspirations, and ultimately, new political parties.

In Tanzania on the other hand, the interposition of a few threats from international lenders and the winds of change sweeping the post-Cold War world, each of which provided the casual force for change, failed to coincide with a similar convergence of material and ideational forces.

When compared to many other African neighbors, the constellation of social forces which

204 challenged TANU in the past, were not fundamentally more vocal by the early 1990s. Whereas conceding political space in Zambia, Malawi, and Kenya eventually resulted in change through the ballot box, Nyerere could, without concern, open the door for multiparty change and proceed to manage the transition by reinvigorating the CCM throughout the electorate. It would appear that the intensity the divergence between those demarcations which defined the constitution of social forces discussed in the previous chapter, acted as the primary impetus shaping the nature of the transition for single-party to multiparty politics.

The CCM and the Rise of the Bureaucratic Bourgeoisie

In Africa, the state’s political hegemony within society is largely buttressed by its position as the mediator between domestic and international society. Therefore, the degree to which a political party or social cleavage dominates the state, impacts the relative capacity for long-term incumbency. To make a rather long digression short, by the 1980s, after instituting a doctrine of

Party Supremacy, the CCM and the state were seemingly indistinguishable, allowing the party not only to reproduce dominance through the coercive power of the state, but also to benefit exclusively from the state’s international network. Furthermore, being situated between the domestic and the international economy and the increased need for technical expertise to administer state ownership of the economy’s commanding heights, the state became less an institution for the reproduction of socialism and more for the reproduction of the tenure and consumption style of a continually growing bureaucratic class.

Baregu (1997) defines party supremacy as “One-party dominance over the society as a whole at the macro-political level and centralized party bureaucratic control over policy making and implementation at the micro level” (p. 58). By definition therefore, the legal process of

205 party supremacy was born in 1965, with the emergence of a single-party constitution as a hedge against social divisiveness and political instability. However, over the ten-years that followed, the National Assembly was still a vibrant, often an undisciplined source of political debate and sometimes bordering on the torment of ideological dissent which regularly found in multiparty assemblies. Likewise, mass membership organizations proved to be most troublesome to contain, and, by their sectionalist nature, posed as the biggest continual threat to national and party unity.

The mid-1970s ushered in a more clearly articulated commitment to the supremacy of the party over the state and society. Party Supremacy over the National Assembly was originally enshrined in a National Executive Committee meeting in Shinyanga in 1975 (Okema, 1996, p.

140). The drafting of the CCM Constitution of 1977, following from the merger of TANU and the ASP, continued the process. The CCM evolved out of this merger with greater degrees of centralization and exclusivity, adopting an overt vanguard function which made it clear that the era of mass membership as a primary goal was over. Instead, party members would have to demonstrate knowledge and adherence to party principles, with the goal of implementing an already constructed socialist ideology. Mass organizations, such as NUTA, were brought under direct party control by making them CCM departments under the party constitution (Hyden,

1980; Barkan, 1994; Chambua, 2002). Supremacy over mass organizations coincided with these name changes. Thus, NUTA became Jumuiya ya Wafanyakazi wa Tanzania (JUWATA) and the

TYL became the Umoja wa Vijana wa CCM. Finally, throughout the 1960s and 70s, TANU managed to secure finances through membership fees, party publication sales, party business ventures like SUKITA4, and voluntary donations. As the CCM bureaucracy grew, the party

4 SUKITA, or Shirika La Uchumi na Kilimo, was one out of three companies owned by the CCM. As a commercial arm of the ruling party, SUKITA raised revenues through rent-seeking activities and a variety of business 206 increasingly relied on the state for financial support. Revenue data is telling. In 1973/74, TANU depended on the government’s subvention for some 39 percent of the party’s annual budget.

Eleven years later (1984/85), nearly 93 percent of the annual budget of the CCM was derived from government funding (Kapinga, 1995).

Ironically however, despite greater vocalized commitments to socialism, party supremacy was also part of a larger political shake-up, which loosened priorities to the degree that adapting to the economic and political adjustments of the 80s could more easily be negotiated without producing major splits and disintegration. First, a major turnover in leadership ensued in 1977 which managed to retire some of the older generation of party leaders, and replace them with younger and more educated officials. Hyden (1980) goes as far as to define the 1977 turnover as a “change of regime” comparable to what might be achieved in multiparty systems (p. 138).

Secondly, while the CCM was, on paper, more rigorously committed to socialism, the party focused less on heady ideological articulations and more on technicalities of policy implementation. Over time, this also meant that emphasis would be placed on technocratic competence in task execution, without an overarching ideological commitment among the bureaucratic managers. Okoko (1987) for example, finds the National Development Corporation

(NDC) lacking in a “clear national industrial strategy” or guidelines as a basis for the

“implementation of its task of socialist construction” (Okoko, 1987, p. 222). In an earlier study,

Parker (1979) found a similar trend within the NDC, and the failures of the Party to serve as an adequate ideological guide (p. 69). In short, Party Supremacy appears to have strengthened the hand of a more technocratically oriented bureaucratic bourgeoisie ideology, which was generally more committed to public service as a means for accumulation rather than socialist

ventures. The organization also borrowed heavily from the National Bank of Commerce and was eventually placed under receivership due to insolvency. 207 development. Therefore, by the mid-1980s, commitments to socialism were loose enough to allow space for the rationalization of economic reforms, and by the 1990s, political reforms.

From the perspective of the party organization, Party Supremacy in Tanzania by itself was not a unique condition. Similar terrain was staked out in Zambia. Through appointment powers,

UNIP’s dominance of national and local government institutions expanded enormously throughout the 1970s, formalized through the 1975 Constitutional Amendment which instituted the supremacy of the Central Committee over the Cabinet (Gertzel, 1984). While KANU under

Kenyatta was little more than a name for a hallow organization once the center of Kenya’s independence movement, KANU was reinvigorated under Moi, who learned from the experiences of the CCM that party supremacy offered the clearest path by which to discipline unruly parliamentarians (Widner, 1992). In Malawi, the MCP ranked as one of the most dominating parties in Africa. As one of the fastest growing parties during the late 1950s and early 1960s, the MCP quickly “developed as an absolutist body,” both internally as externally

(McCracken, 1998, p. 242). As cited by McCracken (1998), one of Banda’s close associates once commented that ‘there is absolutely nothing wrong in dictatorship, and, therefore, if we, through the Malawi Congress Party, established a dictatorship here, we can only be described as the loyal servants of God’ (p. 245).

This section has demonstrated two important points. First, with the increasingly high demands for technocratic competence and the state supervision of nearly every facet of economic life, the position of those most committed to the ideals of socialism were descending while more technocratically oriented bureaucrats were ascending. Secondly, while the consolidation of party power was more personalistic with respect to Banda and Moi, the common trend between all four cases was the increased central control through a party

208 organization that left little room for political competition. In other words, differences in the degree to which various parties dominated the state will probably not go very far in accounting for different patterns of political contestation in the lead-up to multiparty reform. Instead, the single biggest difference seems to lay in variations in the levels of macro-structural divergence, which in turn, had consequences on the centrifugal pull within incumbent regimes and the prospects for rising contention from long-time regime outsiders. Stated differently, relative to the bureaucratic bourgeoisie in neighboring countries, the ruling class in Tanzania at the onset of multiparty reform was far more cohesive due to a macro-structural milieu which offered few attractive opportunities for would-be entrepreneurs to break free in order to establish their own political organizations. These same opportunity structures also negatively impacted the growth of popular contention or the ability entrepreneurial regime outsiders had to build sustainable party organizations.

New Complications

In a number of ways however, the lead-up to the first multiparty election in 1995 was an uphill battle for the CCM, despite the dominant position of the bureaucratic bourgeoisie relative to other social forces. While popular protests were, without a doubt, a weak impetus for multiparty change, as the CCM increasingly lost its ideological way due to the difficulties of fitting the nature of economic reform to the rhetoric of Ujamaa, and the increasingly dominant position of the consuming spirit of the bureaucratic bourgeoisie within the party, the rhetorical justification for CCM dominance was harder to find. Moreover, by the early 1990s, it was clear that party leadership could not find an adequate set of ideals by which to rest a party platform.

While Ujamaa increasingly came to symbolize economic failure, the welcome mat for

209 liberalization was not placed at the doorstep by those that bore the brunt of reforms. Currency devaluation and labor retrenchment left many more Tanzanians living on the margins of existence, while the dividends of privatization promised to follow the established patron-client networks. As one reporter notes on the eve of the 1995 election campaigns, “there is a common agreement that CCM is very unpopular among the youth, elite, civil servants, students and urban dwellers, many of who have lost their jobs through retrenchment (Kilimwiko, 1995).

By the onset of the election nomination period, it would appear that the threat to the continuation of the CCM’s dominance over the state was very real. A Tanpoll survey conducted in the spring of 1995 revealed the CCM’s increasingly tenuous position, where overall support for the incumbent dropped to 36.6%, down from 63.2% in late 1994. It would also appear that the defection from the CCM of the renegade Augustine Mrema, was a net gain for the NCCR-

Maguezi, which polled some 53.3% in the same survey (Godwin, 1995a). The popularity of

Mrema and the NCCR-M was especially notable in relation to the CCM’s slate of presidential nominees. While Mrema scored a 56.3% approval rating in the Tanpoll survey, the highest scoring CCM candidate was Edward Lowassa at 12.1% (Semwaiko, 1995). A separate six region poll conducted in July of the same year by the Department of Political Science and Public

Administration at the University of Dar es Salaam, showed that Lowassa was preferred by 21% of the sample, while Mrema polled 33%. Mkapa was never even mentioned by the respondents

(Mpinga & Cushnie, 1995)!

Yet, in spite of these survey results, the CCM trounced the opposition in the first and subsequent multiparty elections. In the chapters that remain, questions over the CCM’s comparatively wide election victories will be the central focus. The framework adopted in chapter two clearly points toward two topics that will be included in the subsequent analysis.

210

First, the development trajectory establish by the Arusha Declaration had clear macro-structural consequences, which in turn, shaped the material and ideational opportunities available for generating spontaneous protests and for political entrepreneurs wanting to launch popularly supported political parties with some organizational sustenance. Chapters eight through twelve will specify the causal consequences of Arusha and, in turn, relate these consequences to the rise of opposition to single-party rule and the CCM. Secondly, the opportunities and constraints posed for organizing and mobilizing contention do not however, determine the contours of multiparty competition. Instead, actions at the micro-level, which include strategies and tactics adopted by political entrepreneurs, have causal force in their own right. In chapter twelve, I will demonstrate how, when compared to the opposition parties, the CCM managed to craft its campaign in ways that maximize message impact, thus overcoming the dwindling support reported in the aforementioned Tanpoll survey.

211

CHAPTER EIGHT THE MATERIAL BASIS FOR ORGANIZING CONTENTION

“It should not be necessary for TANU leaders to understand statistics before they realize that Tanzania is poor.”

Julius K. Nyerere, 1968

“Men's ideas are the most direct emanations of their material state.”

Karl Marx, 1863

During the course my various engagements in Tanzania between 1999 and 2006 with dozens of leaders from six of the most notable opposition parties, I was universally struck by the rudimental quality of party offices, the small staff sizes, and the total absence of a single computer. In my first meeting in Dar es Salaam with a few courteous heads of the now nearly defunct Tanzania Labor Party (TLP), I expected to find a headquarters which might rival some of my previous visits to CCM establishments, especially considering the TLP was chaired by the enormously popular Augustine Mrema. In fact, I was taken on a brief tour of the office spaces, but never once managed to catch a glimpse of even a typewriter.

Some of the other parties appeared to have discernibly better financial luck. In 1999, the headquarters of the United Democratic Party (UDP), thanks to the personal financial fortunes of its founder John Cheyo, had a relatively vibrant staff and at least two computers. As for

CHADEMA, the encounters of my first visit during 1999 were similar to my encounters at the

TLP headquarters. During my more recent visits, CHADEMA headquarters in Dar es Salaam contained a computer and a somewhat professional staff, and it was clear that the party was able to benefit from its lucrative networks into the cash crop producing regions of the Mainland. In this chapter, I argue that, when thinking about the manifestations of macro-structural resource distributions at the unit-level, e.g. the relative capability of a particular actor, heavier concentration of resources outside the spatial reach of the state should translate into greater capacities for political entrepreneurs to mobilize resources for party building. Success in political competition is often decided by success in rallying together finances, often done by tapping lucrative networks to powerful interest groups or utilizing technologies for building a pool of resources out of small contributions from many sources.

In societies found throughout Africa, where the bulk of the population is poor, the latter tactic is simply not a realistic option. Instead, political entrepreneurs must rely on partnerships with a variety of wealthy actors for sources of finance, most of which are probably involved in business activities. As is often the case, these financially endowed actors are the very political entrepreneurs trying to capture state power. In short, this chapter argues that, political contestation is intensified when regimes are faced with a society whereby resources can be easily mobilized by a few political entrepreneurs seeking to dislodge the incumbent’s control over the legislative and executive branches. As the opening words of this chapter suggest, mobilizing economic resources in Tanzania is an almost insurmountable feat which continues to dog the opposition today.

Economic Development and Capitalization

Had the development patterns found in Kenya, Zambia, and Malawi unfolded in Tanzania, then it would have been more likely that cross-regional political alliances would have found the financial capital and popular appeal for launching a sustained attack against the CCM during the first, and subsequent multiparty elections. The post-independence development trajectory set by

213 the Arusha Declaration, quite simply generated a macro-structural climate on the Mainland whereby the commanding heights of the economy were under state control, while the most productive asset, namely land, remained in the hands of the highly dispersed peasantry. The successful attempt to discourage private accumulation slowed or negated the growth of landlessness and land concentration. Consequently, Ujamaa also undermined the formation of class structures endowed with the financial wherewithal for underwriting opposition party formation. In short, opposition parties cannot hope to pay rents for office spaces, salaries for professional staffers, prices for office supplies, or the cost for recruitment drives and election campaigns in environments where political entrepreneurs are not able to round up finances. This problem is compounded by the fact that, where the bulk of the population is poor, they often look to political organizations as potential suppliers for their daily bread.

The data for backing up these claims are abundant. In Kenya, since initial state consolidation was based on a sort of pact between the assorted sections of the rural petty bourgeoisie and the state executive, regional notables were able to retain an enormous amount of political and economic clout when cast in comparison with Tanzania. Those political bonds that formed between state bureaucrats and the petty bourgeoisie constituted the network ties for increasing agricultural capitalization and land consolidation while simultaneously increasing the rate of landlessness (Migot-Adholla, 1979, p. 154). At the same time, the Harambee self-help system, which linked local notables and parliamentarians to the distribution of executive patronage, tended to favor those previously developed regions, whose political backing was essential for national political stability. Furthermore, the post independence pursuit of private sector growth also generated the development of sizable urban commercial and professional classes, with the wealth and international networks for organizing against Moi’s increasingly autocratic style.

214

The qualities of Malawi’s reform environment are explained largely by the developmental differentials nurtured under single-party rule, which amalgamated into tripartite divisions between the northern, central, and southern regions. Disparities within the agricultural sector grew under single-party rule, as large estates were given preferential treatment over smallholder plots (Kydd & Christiansen, 1982; Mhone, 1987; Mkandawire, 1992). In fact, up to 1990, smallholders were outright prohibited from growing export produce. To make matters more acute for smallholders, the real producer price of tobacco, a principle export crop, rose some 10 percent between 1975 and 1990, while the prices of smallholder crops, such as maize, rice, and groundnuts, declined considerably during the same period (Kaunda, 1998, p. 58).

Further to the west, in Zambia, the differentials between wealthy famers in the Southern

Province, versus poor tenant or smallholder farmers throughout the remainder of the countryside, helped provide the financial impetus for organized political contention in the pre and post-reform environment (Sichone, 1996; Du Toit, 1999). Even more importantly, the growth of copper mining provided the occupational conditions for the rise of a powerful labor union movement, whose membership base within the Zambian Congress of Trade Unions eventually outstripped UNIP’s membership (Gertzel, 1984).

Additional data supports the claim that capitalization and class developments in Tanzania were exceptional in Africa. When compared to Kenya for example, the annual average growth rates of manufacturing and commercial agriculture during the single-party era were considerably lower in Tanzania (Barkan, 1979a, p. 17). The sectoral composition of employment in Tanzania at the onset of multiparty reform is indicative of modest levels of economic differentiation, and in turn, capitalization. Whereas in Kenya, some 20 percent and 61 percent of the working population was employed in the industrial and service sectors respectively, and in Zambia,

215 around 8 percent and 24 percent worked in the industrial and service sectors respectively, in

Tanzania industrial and service sector labor represented some 4 percent and 12 percent of the total workforce1. All-in-all, over 84 percent of those employed in Tanzania earn their daily bread from agricultural work, making Tanzania one of Africa’s most agrarian economies at the time of political reform, alongside Uganda, Rwanda, and Burkina Faso (World Bank, 2007a)

Just as significant, survey data from Afrobarometer (2005b, c, d, e) finds some 24.5 percent of the Tanzanian respondents as subsistence farmers and another 36.4 percent as peasant farmer.

By contrast, 13.9 percent and 14.8 percent of the Kenyan respondents and 13.4 percent and 7.8 percent of the Zambian respondents reported their occupation as subsistence and peasant farmer respectively. The responses for Malawi were somewhat similar to Tanzania, as 30.2 percent reported subsistence farming, while another 23.3 percent reported peasant farming as primary occupations.

The fragmentation of rural production is captured in a variety of additional data. It is safe to assume that the greater the extent to which technology is incorporated into the production process, the greater the level of capitalization, and by extension, the greater the prospects for political entrepreneurs to find lucrative sources of funding. For example, at the onset of multiparty reform pressures, fertilizer use in Malawi and Kenya was well above the average for

Africa, at 36 and 24 percent respectively, while Tanzania, along with Zambia, were slightly below the average, at 11 and 12 percent respectively (United Nations Food and Agriculture

Organization, 2006). In Tanzania, it is the smallholders and traditional agro-pastoralists that use about 85 percent of all the arable land for productive purposes. While there are 44 million hectors of arable land, medium and large scale farming exploits only 100 thousand hectors

1 Numbers for Kenya and Tanzania are reported for 1991, while figures for Zambia are taken from 1996. 216

(Tanzania Bankers Association, 2005). Traditional production methods go to produce 93 percent of the domestically consumed milk and 99 percent of the domestically consumed meat, while the rest is produced within the commercial sector (Sarwatt, 2000).

Data on enterprise size also supports the claim that, whereas capitalization in neighboring countries has led to the proliferation of wealth concentration and powerful commercial classes, in Tanzania almost all business venturers are self-employed or small business operators. As reported by the International Finance Corporation, between the years of 1994 to 2005 there were

22,014 micro, small, and medium enterprises (MSMEs) operating in Kenya, and 747,296 in

Malawi, and a whooping 2,700,000 MSMEs in Tanzania (International Finance Corporation,

2006)2. When discounting for population differences, Malawi and Tanzania are similar, at 18 and 14 people for MSME respectively, while Kenya stands at 1677 people per MSME.

In short, African Socialism in Tanzania, when compared to the openly capitalist development trajectories laid out in Kenya and Malawi, or the ambiguities of Humanism in

Zambia, resulted in patterns of development that were considerably less capitalized and more agrarian. As specified at the onset of this section, the significance of this uniqueness is that capitalization also equates to a greater capacity for organizing resources beyond the spatial reach of the party-state nexus. While the same process also means that undercapitalization weakens the absolute capacity of the state, which is forced to rely on the low yield production of the peasantry and greater quantities of international aid, the very fragmented nature of peasant production also means that collective action in opposition to the state will be almost impossible.

2 An MSME is a business which employees no more than 250 workers. 217

Class and Political Divisions

Theoretically speaking, class structures in Tanzania lack the capacity for translating economic production into viable political challenges. In order to give this theoretical claim some empirical substance, I will examine how existing class structures relate to the array of political actors which emerged out of the 1992 political reforms. Since comparative data, such as party balance sheets and funding sources, among these cases is non-existent, I will cast my empirical net broadly by looking at a number of surveys which contain potential indicators for identifying politically motivated allocation of resources. These include a discussion of class, occupation, and educational standing of party founders and members. Where relationships between these indicators and party leadership and membership are tight, we can safely conclude that had post- independence development policy in Tanzania placed a greater emphasis of macro-economic growth and private sector capitalization, the size of bourgeoisie and labor classes would have been infinitely more difficult to contain during the transition from single to multiparty politics.

In a nutshell, the overwhelmingly large size of the peasant sector, and the fragmented nature of the petty bourgeoisie and commercial classes has generally translated into opposition parties with little or no financing beyond the paltry revenues from party founders. As partially captured by the data in table 8.01, at the onset of multiparty reform in Tanzania, the most obvious political expression of the commercial interests were found in the Party for Democracy and

Development (CHADEMA) and the United Democratic Party/Union for Multiparty Democracy

(UDP/UMD). Formed by Edwin Mtei, a former finance minister, Bank of Tanzania Governor,

IMF director, and coffee grower, CHADEMA was organized in the early 1990s largely by business owners and commercial farmers in Kilimanjaro and Arusha (Erdmann, 1995). Perhaps due to Mtei’s history of vocalizing opposition to Ujamaa, roughly half of the founding members

218 were engaged in private business. Not surprising, the overall profile of the party’s platform tends to be economically conservative, with a principle focus on private enterprise and cash crop production (Whitehead, 2000). Similarly, the UMD, and later, John Cheyo’s breakaway faction, the UDP, generally favor liberalization, individual freedoms, and, increased foreign investment

(Whitehead, 2000). Incidentally, these ambitions reflects the personal ambitions and class interests of the party founders which, as defined by table 8.01, appear to hail from the commercial and petty bourgeoisie sections of society. Cheyo is a prominent businessman, often referred to as “bwana mpesa” (money man), while Chief Abdallah Fundikira, who resigned from his cabinet post in 1963 out of protest to the growing prospects of single-party rule, was a

Nyamwezi Chief prior to the dismantling of the native authority system and private business owner. Other younger opposition leaders, such as James Mapalala, one of the founding members of CUF, and Maberere Marando, founding member of the NCCR-Mageuzi, hailed from a professional class of talented lawyers and security agents, but lacked the financial capacity for building party organizations similar to their youthful counterparts in Kenya, Malawi, and

Zambia (Chege, 1994).

Table 8.01: Occupation of Founding Members NCCR CHAD CUF UMD TPP NLD PONA YPDP Total

n= 1703 2000 3000 2000 2036 1998 2000 2080 16817 Peasant/Farmer 58.7% 35.0% 45.0% 47.5% 61.4% 50.2% 48.6% 47.6% 48.1% Public Servant X Private 29.5% 50.1% 26.1% 52.2% 23.1% 45.0% 42.8% 48.5% 39.0% Business Politician/Other 11.7% 15.0% 28.9% 0.4% 15.4% 4.8% 8.7% 3.9% 12.1% Source: Mmuya & Chaligha, 1994, p. 49

REDET survey data compiled by Mushi, Mukandala, and Baregu (2001) finds a considerable degree of political divergence across occupational categories, although the sample sizes of a few of the categories leave a lot to be desired. When considering the question over the

219 level of attention paid by government officials toward citizen complaints or concerns, and when discounting for non-committed responses, we find that politicians (n=16) and farmers (n=3152) to be the most inclined to feel that government officials pay attention to their concerns (70 percent and 53 percent respectively), and are the least inclined to feel ignored (0 percent and 19 percent respectively).

In contrast, students (n=95) and self-employed (n=192) tend to be the least inclined to feel government officials listen (31 percent and 35 percent respectively), and the most inclined to feel ignored (26 percent and 30 percent respectively). The same pattern of divergence holds within the category of “little attention.” Politicians and farmers were the least inclined to report little attention (30 percent and 28 percent respectively), while students and self-employed were the most inclined (43 percent and 35 percent respectively).

Turning to the question on the positive impact that national government has on daily life

(see table 8.02), when discounting for non-committed responses, farmers (n=2918), along with the unemployed (n=4519), had the highest propensity to see national government as a benefit to daily life (58 percent and 57 percent respectively), while students (n=127) and self-employed

(n=72) had the lowest propensity to see the same benefits (40 percent and 46 percent respectively). Likewise, self-employed and students, were the most likely to feel that national government produces small benefits (51 percent and 49 percent respectively), while farmers and other (n=320) were the least likely to feel the same way (37 percent for each). Finally, while less than 6 percent of the committed sample found national government to have no benefit on daily life, students and politicians (n=15) were the most skeptical (10 percent and 13 percent respectively), while self-employed, employed, farmers, and business owners were tied for the least skeptical (5 percent each).

220

Table 8.02: Impact of National Government on Daily Life E P F B S SE O Un T Great Benefit 55 % 50 % 58 % 53 % 40 % 46 % 53 % 57 % 57 % Small Benefit 40 % 38 % 37 % 42 % 51 % 49 % 37 % 38 % 38 % No Benefit 5 % 13 % 5 % 5 % 9 % 5 % 10 % 6 % 5 % Notes: E=employed, P=politician, F=farmer, B=business owner, S=student, SE=self-employed, O=Other, Un=unemployed, T=total Source: Mushi, Mukandala & Baregu, 2001.

Moving on to table 8.03, the question over the private benefit of local government reveals patterns similar to those described above. When controlling for the non-committed, politicians

(n=15) and farmers (n=2922) appear to have the most positive assessment of local government policy. Some 50 percent and 49 percent respectively returned a “great benefit” response, while farmers, along with the employed (n=682), were also the least likely to attribute local government with no private benefit (9 percent and 7 percent respectively). At the same time, business owners (n=252) and the students (n=90) had the clearest negative assessment of local government policy. Slightly over 31 percent of the students and 44 percent of the business owners reported great benefits from local government policy, while 23 percent and 13 percent respectively reported no benefit from government policy.

Table 8.03: Impact of Local Government on Daily Life E P F B S SE O Un T Great Benefit 45 % 50 % 49 % 44 % 31 % 45 % 49 % 46 % 48 % Small Benefit 48 % 38 % 42 % 43 % 46 % 43 % 32 % 42 % 43 % No Benefit 7 % 13 % 9 % 13 % 23 % 11 % 20 % 11 % 9 % Notes: E=employed, P=politician, F=farmer, B=business owner, S=student, SE=self-employed, O=Other, Un=unemployed, T=total Source: Mushi, Mukandala & Baregu, 2001.

In short, the data compiled by Mushi, Mukandala and Baregu (2001) indicated some political tensions along lines of occupation. For starters, students were the most skeptical toward the political status quo. They were the least inclined to feel listened to and to see positive benefits from government policies. Likewise, the self-employed and business owners, which

221 may be described as a sort of petty and commercial bourgeoisie respectively, were also somewhat more skeptical politically. By contrast, farmers, the bulk of which are undoubtedly peasants, were the least skeptical. Overall however, most of the respondents in all categories attribute some benefit from government policy.

Additional data further refines the picture of the political orientation and class ties in

Tanzania. Correlations between the 1995 election results and data from the 2001 Household

Budget Survey (National Bureau of Statistics, 2002) produced a number of interesting findings.

Somewhat surprisingly, income from cash crop production was significantly and positively correlated with support for the CCM, with a Pearson Correlation of 0.518 for the 1995

Parliamentary election and 0.490 for the 1995 Presidential election. In other words, regions which consisted of larger numbers of individuals earning income from cash crop production

(Ruvuma, Mtwara, Lindi, and Shinyanga) were far more likely to be pro-CCM regions. Indeed,

Mtwara, Lindi, and Ruvuma were the three regions most committed to the CCM during the

1995 elections, while Shinyanga turned out to show strong support for the UDP.

At first glance, this association seems to contradict the suggestion that cash crop production is inversely related to CCM support. However, this data attains some clarity when accounting for land consolidation. Greater levels of land concentrations indicate greater levels of cash crop production from commercial estate farmers rather than smallholders. This association was born out in the 1995 elections. The greater the percentage of individuals owning farmland, the more likely the region was a pro-CCM one, with a Pearson Correlation of 0.3913 for the

Parliamentary election and 0.473 for the Presidential election. Three out of the four regions

3 The correlation between the 1995 Parliamentary elections and percentage owning farmland was significant at the 0.10 level (2-tailed). All other correlations are significant at the 0.01 or 0.05 levels, unless otherwise noted.

222 where the percentage of people owning farmland was lowest (Kagera, Kilimanjaro, Mbeya) were also strong opposition supporters, the exception being Iringa.

Correlations between data from the Register of Establishments and the 1995 elections produce similar findings. Regions with higher concentrations of commercial and professional services, such as financial and public administration establishments, are less likely to vote for the CCM4. While correlations on the number of agricultural and industrial establishments were not statistically significant, it is worth noting that both were substantially and inversely related to CCM support. Three (Mbeya, Arusha, Kilimanjaro) out of the four regions with the highest number of industrial establishments were also strong opposition backers in the 1995 election.

Arusha and Kilimanjaro also registered as the first and fourth highest number of agricultural establishments respectively, although Iringa and Tanga, both relatively pro-CCM regions, ranked as number second and third respectively.

A closer inspection of the spatial distribution of election supports the claim that, in

Tanzania, the weakness of petty and commercial bourgeoisie classes is a robust explanation for the inability of opposition parties to pose as sustained threats to the CCM. Stated differently, because there are definite zonal patterns of opposition support which coincide with greater levels of bourgeoisie class development, had these class structures been larger in size, opposition strength would also have correspondingly greater.

As identified in appendix C, in 1995, the Arusha Region, which borders Kilimanjaro, proved to be a valuable Mrema backer. The Great Lakes Regions of Kagera and Mara proved to be his third and fourth most committed support bases. Together, all four regions (Arusha, Kagera,

Kilimanjaro, and Mara) account for 43.7 percent of the votes cast for Mrema, while all four

4 Statistical significance was obtained only with respect to the 1995 Presidential Elections. The number of financial establishments yielded a Pearson Correlation of -0.399, while number of public administration establishments yielded a slightly higher correlation of -0.420. Both are significant at the 0.10 level (2-tailed test). 223 account for only 22.8 percent of all votes cast. Stated differently, Mrema’s combined support in these four regions in northern Tanzania was 25.55 percentage points above his national support level. Tucked along the southern border with Zambia, Mbeya proved to be Mrema’s fourth most committed region. Traveling to the north west of Mbeya, the region of Rukwa also demonstrated its support for Mrema. In fact, a total of 9 out of 25 regions (Arusha, Kagera, Kigoma,

Kilimanjaro, Mara, Mbeya, Morogoro, Rukwa, and Tabora) committed levels of support above the candidate’s national support level. In geographical terms, aside from Morogoro, Mrema’s most loyal regions extend west from Kilimanjaro, following the Tanzanian border around the

Great Lakes and turning southward, ending in Mbeya. These regions are also considered the country’s most vibrant larger-holder regions. By contrast, his support is lowest among those regions which show low degrees of class differentiation, mostly concentrated in the central regions and in the south and south east.

When aggregating the spatial distribution of support for the opposition as a whole, ’zonal’ patterns of support reflect patterns of support for Mrema. According to the results from the 1995 presidential election, movement from northern to southern Tanzania corresponds to an increased probability of finding CCM supporters. As depicted in table 8.04, support for the CCM was roughly 4 times greater in the south than the north, northwest, and northeast. Even Zanzibar was less of an opposition stronghold than the Northeast, North, and Northwest zones.

Undoubtedly, part of the popular vigor behind opposition support stemmed from the personal connections that individual candidates had with their place of birth. This is most evident with Cheyo’s support in Shinyanga and Mrema’s support in Kilimanjaro. However, one cannot ignore the breadth of Mrema’s spread throughout the northern sections of the Mainland,

224 and conclude that, since class developments in the north have been far more differentiated historically, the petty bourgeoisie emerged as the principle basis of the opposition.

Table 8.04: “Zonal” Variation in Incumbent Support Zone CCM as % of Opposition Northwest (Kagera, Mwanza, Shinyanga) 123% North (Mara, Arusha) 119% Northeast (Kilimanjaro, Tanga) 89% East (Coast, Dar es Salaam, Morogoro) 154% Central (Dodoma, Singida, Tabora) 184% West (Rukwa, Kigoma) 135% Southwest (Iringa, Mbeya) 161% South (Lindi, Mtwara, Ruvuma) 415% Zanzibar (Pemba N, Pemba S, Unguja N, Unguja S, Urban W) 187% Source: National Electoral Commission, 1997

Correlations between educational background and party support also serves as an indicator of the generative capacities that class structures have on political party formation. Educational attainment is likely to result in greater pay, expanded social networks, and higher levels of respect, all of which are assets for party building. The assumption here is that, if education levels correlate with political orientation, had post-independence education policies focused less on equality and more on education for market consumption, opposition parties would be far more capable today than they currently are.

I will provide a brief overview of the educational distributions among opposition party founders, as depicted in table 8.05. Quite clearly, CUF and the TPP have the most educated profile out of all the opposition parties at the time of multiparty reform. Only 38.7 percent of the

CUF and 30.5% of the TPP founding members had nothing beyond a primary school education, compared to 75.5 percent for CHADEMA, 88.8% for TYPD, and 55.2 percent for the distribution total. CUF and the TPP as well show higher rates of founding members with university level education, although CHADEMA shows high numbers here as well. What this

225 distribution shows is that, had education levels been higher in Tanzania at the time of multiparty reform, all other things being equal, CUF and the TPP would have been the net beneficiaries.

Table 8.05: Education of Founding Members NCCR CHAD CUF UMD TPP NLD PONA TYDP Total Primary 50.0 % 75.5 % 38.7 % 42.5 % 30.5 % 60.2 % 51.5 % 88.8 % 55.2 % Secondary 35.2 % 20.4 % 34.7 % 48.5 % 45.2 % 30.0 % 25.0 % 6.4 % 30.1 % Post‐Secondary 14.1 % 0.6 % 21.4 % 8.0 % 20.1 % 9.1 % 20.0 % 3.9 % 12.0 % University 0.7 % 3.4 % 5.3 % 1.0 % 4.2 % 0.8 % 3.5 % 0.9 % 2.7 % n 1420 2940 3000 2000 2087 2000 2000 2118 17565 Source: Mmuya & Chaligha, 1994, p. 49

Table 8.06: Education versus Party Identification None Std 1‐2 Std 3‐4 Std 5‐6 Std 7‐8 F 1‐2 F 3‐4 F 5‐6 Degree CCM 92% 92% 94% 88% 87% 88% 74% 42% 22% CHADEMA 1.3% 0% 2.1% 0% 2.2% 6.8% 4.3% 14% 0% DP 0% 0% 0% 0% 0.6% 2% 5.8% 3.2% 28% NCCR 0% 0% 0% 0% 0.6% 0% 0.8% 0% 0% TADEA 1% 0% 0% 0% 0.6% 0% 0% 0% 0% UMD/CUF 0% 0% 0% 0% 0.6% 0% 0% 0% 0% Don't Know 2.9% 7.3% 0.8% 5.3% 1.9% 2.4% 1.1% 0% 0% None 2.3% 0% 2.3% 5.9% 5.9% 0% 13% 40% 48% N= 108 18 91 44 533 27 157 29 9 Source: Erdmann, 1995

However, a more interesting data set would include measurements for the CCM as well as the opposition. While suffering from inadequate sample sizes, Erdmann’s 1995 survey provides some interesting comparative data. As table 8.06 above indicates, greater levels of education correspond with greater probabilities of identifying with an opposition party or no party at all.

While 92 percent of the respondents with no formal education report supporting the CCM, that level falls to 74 percent for form 3-4 achievers. At the same time, the level of support for

CHADEMA increases from 1.3 to 4.3 percent and the DP from 0 to 5.8 percent. Those reporting no party identification increased from 2.3 to 13 percent. Therefore, while respondents with higher education levels gravitate toward the opposition only at the margins, respondents with higher levels of education are overwhelmingly repelled away from any political party, including

226 the incumbent party. It appears that at the very least, had educational policies in Tanzania approached those found in Kenya, a larger volume of political loyalties during the multiparty reform period would have been up for grabs.

Evidence of varying levels of education as a force associated with varying political tastes is found elsewhere as well. Data collected by REDET (1999b) in December of 1999 revealed quite convincingly that lower levels of education corresponded with higher levels of satisfaction with government. Of those with no formal education, 76.4 percent were satisfied with the handling of the economy, 85.4 percent with the level of political trust, 71.9 percent with government responsiveness, 76.4 percent with overall government performance, and 85.4 percent with political leadership of the CCM. At the same time, only 39.3 percent of the respondents with no formal education were satisfied with opposition leadership. These numbers decline as one moves up the education ladder, including satisfaction with the opposition. Instead, the bulk of those with the highest level of educational attainment were only somewhat satisfied with the government and the ruling party. Among those with the highest level of education, only 12.5 were satisfied with the handling of the economy, 31.3 percent with the level of political trust,

18.8 percent with government responsiveness, 18.8 percent with government performance, and

25 percent with the CCM’s leadership. Interestingly, only 18.8 percent of most educated respondents were also satisfied with the opposition’s leadership, while 37.5 percent were not satisfied. These trends were duplicated in another REDET (1999a) survey conducted in March of 1999: The greater the level of education, the less satisfaction with political institutions and government performance generally. Again, political loyalties appear more up for grabs as one move up the education ladder.

227

Additional REDET data duplicates the finds of Erdmann and the two REDET surveys above. Cross tabulating the responses in the Mushi, Mukandala, & Baregu (2001) report reveals that opposition party membership is slightly weighted toward those with higher education (see table 8.07). Whereas 1.9 percent of the CCM respondents (n=2916) reported having a higher education, 8.5 percent of the opposition party members (n=165) reported the same education level. Again, 11.4 percent of the CCM respondents also reported having a secondary education, versus 17.6 percent for the opposition. By contrast, 76.9 percent of the CCM respondents cited the no education or the primary level as the highest level of educational attainment, versus 61.2 percent for the opposition.

Table 8.07: Educational Status According to Party Membership CCM Opposition None Other No Answer Total Madrassa 1.4% 9.1% 0.8% 0.0% 1.4% 1.5% Adult Ed 6.0% 2.4% 3.6% 0.0% 4.1% 5.0% Primary/None 76.9% 61.2% 74.2% 62.5% 79.7% 75.5% Secondary 11.4% 17.6% 15.8% 25.0% 12.2% 13.1% Higher 1.9% 8.5% 3.7% 0.0% 2.7% 2.8% Other 2.4% 1.2% 1.7% 12.5% 0.0% 2.1% Source: Mushi, Mukandala & Baregu, 2001

Reversing the tabulations in this data set yields the figures reported in table 8.08 and provides a slightly different insight. The probability of selecting an opposition member out from those with no more than a primary school education is 2.8 percent, compared to the 3.4 percent probability of picking an opposition supporter out of the entire sample. Similarly, the probability of selecting an opposition supporter who has an education level beyond the secondary level is

10.4 percent, well above the probability of picking an opposition supporter out of the entire sample. These numbers are distinctly different for the CCM, where the probabilities of picking a

CCM supporter proportionately decline as one moves up the education level. Furthermore, this

228 data set also indicates that as one increases the level of education, one also increases the likelihood of finding people with no political affiliation.

Table 8.08: Party Membership According to Educational Status CCM Opposition None Other No Answer Madrassa 57.7% 21.1% 19.7% 0.0% 1.4% Adult Ed 72.3% 1.7% 24.8% 0.0% 1.2% Primary/none 61.6% 2.8% 33.9% 0.1% 1.6% Secondary 52.2% 4.6% 41.5% 0.3% 1.4% Higher 41.8% 10.4% 46.3% 0.0% 1.5% Other 68.6% 2.0% 28.4% 1.0% 0.0% Total 60.4% 3.4% 34.4% 0.2% 1.5% Source: Mushi, Mukandala & Baregu, 2001

While nearly all of the surveys reviewed thus far suffer from inadequate sample sizes, when taking the survey data in total, several consistencies emerge. First, greater levels of education are inversely related to CCM support, while lower levels of education are positively associated with CCM support. Secondly, while higher levels of education appear to relate positively to opposition support, the more educated respondents appear even more inclined to not identify with any party at all and are the most likely to be pessimistic toward political institutions altogether.

What do these patterns mean? For starters, declining tendencies to identify with a party among the more educated respondents might indicate a greater appreciation of the risks associated with identifying with the opposition. This interpretation would support the repression thesis discussed in chapter four. Yet, these same individuals also expressed unreserved discontent with political institutions, expressions which would likely be subject to the same risk assessment that restricts identifying with the opposition. A more plausible explanation is that, for a variety of reasons, the political milieu which shapes the actions of those disenchanted with the status quo simply lacks any credible alternative. Perhaps opposition parties are not seen as worthwhile since they lack the wherewithal for seriously challenging the incumbent.

229

Whatever the basis may be for the positive association between higher levels of education and greater skepticism toward political institutions and greater tendencies to, at the very least, not identify with a political party, what is more clear is that had policy fallout from Arusha

Declaration resulted in greater levels of education in the aggregate, a greater number of political loyalties would be up for grabs. This proposition alone would be enough to weaken the electoral performance of the CCM.

While stepping back from education to a more general discussion of the relationship between class and political support, it would be of interest to assess these correlations in comparative perspective. More specifically, do we find similar tendencies for indicators like occupation to correlate with party support among some of the other cases discussed thus far? To enable this assessment, I will turn to some of the more systematized data provided by

Afrobarometer.

For starters, cross tabulating Afrobarometer data on occupation and party support, the

Tanzania sample reveals that out of the 82.9 percent of the respondents which supported the

CCM, some 63.8 percent were peasants or subsistence farmers, versus 43 percent out of the 6.1 percent which supported an opposition party, and 51 percent of the 11 percent uncommitted respondents. By contrast, 19 percent of the opposition supporters were business owners, compared to the 12 percent which supported the CCM. This was especially true for hawkers and traders, which consisted of 22 percent of CUF’s supporters, compared to 10.4 percent of the

CCM’s. Likewise, commercial farmers were slightly more likely to be found among the opposition supporters, but only marginally so. The only surprise was the category of professional employees, which slightly favored the CCM. However, the sample sizes for

230 commercial farmers and professionals, when paired with party preference responses, were far too small to draw any conclusions.

Wage workers were also somewhat more inclined to favor the opposition, or express no commitment at all. Slightly over 15 percent favored the opposition, another 14.7 percent were uncommitted, and 10.1 percent favored the CCM. This was especially true for retail workers, unskilled manual laborers, miners, and farm workers, which tended to flock to CUF,

CHADEMA, and the TLP.

In Malawi, it is widely known that class differentials were the seeds for opposition party development. Hence, the Alliance for Democracy (AFORD) boasted the strongest level of support from northerners, which, on average, were more educated and overrepresented in the civil service at the onset of reform, forming the backbone of the Malawi Congress of Trade

Unions (Dzimbiri, 2005). The economic wealth of the central region by contrast, depended largely on the work of small-holder tobacco producers, who generally remained loyal to Banda and the MCP. Finally, the UDF hailed largely in the south, where commercial farming resulted in considerable land and wealth consolidation and a vibrant business environment. Indeed, as

Posner (1995) clarifies, the UDF, which emerged as the victor of the 1994 election, “was literally born in the Chamber of Commerce” (p. 137).

Afrobarometer data for Malawi does indicate that class background is associated with party affiliation to similar degrees found in Tanzania. Relationships between occupation and party support were quite weak in some areas. While peasants and pastoralists made up 53.1 percent of the sample, 49 percent of the Democratic Progress Party (DPP)5 supporters hailed from the same occupational category, versus 56.6 percent for the opposition. By contrast, commercial

5 The DPP, the current ruling party in Malawi, was formed in 2005 by President Bingu wa Mutharika after he broke away from the UDF over a corruption dispute. 231 farmers and wage workers were slightly more represented in the DPP. When breaking the data down into the three largest vote-getting parties, we still find subtle, but notable differences.

Most noteworthy, subsistence and peasant production is more clearly associated with support for the former incumbent, the MCP, at 62.4 percent of the respondents, versus 49 percent for the

DPP and 52 percent for the UDF. Business owners, which constituted 11.5 percent of the sample, tended to be most strongly represented in the UDF (14.1 percent) and least represented in the MCP (9.3 percent). Wage workers were most strongly represented in the DPP (13.8 percent) and least represented in the MCP (8.8 percent).

In the face of a serious economic decline and inflation, the organizational strength of labor in Zambia acted as a critical mass for the popular support and leadership for the Movement for

Multiparty Democracy (MMD), and ultimately, the downfall of UNIP. Recent Afro barometer data supports the claim that differences in occupational category, including labor, is associated with differences in party support. While 21.1 percent of the sample reported being a peasant or and subsistence farmer, this same occupational class made up some 27.4 percent of the MMD’s support. Breaking this support down into the three most supported parties, and distinguishing between subsistence and market oriented peasant farmers, reveals that the former category is highly represented in the MMD (18.8 percent) and the UPND (17.7 percent), but least represented in the PF (8.3 percent), while the latter most represented in the UPND (12.5 percent), but less so in the MMD (8.6 percent) and the PF (5.1 percent). The composition of opposition support also slightly favored workers, business owners, and professionals. Some 28.8 percent of the PF’s support came from wage workers, compared to 21.4 percent for the UPND and 20.4 percent for the MMD. Likewise, business owners were more strongly represented in the PF, at 13.5 percent compared to 10.5 for the UPND and 9.2 percent for the MMD.

232

Moving on the Kenya, class foci were common elements among a wide range of discussions over the rise of opposition in Kenya. Wider (1992) for example cites that:

To the extent that the party-state did not and has not assumed the range of functions and powers in Kenya that is has in other African countries, it has shown restraint largely because of the existence of interest groups who economic power independent of government favor or whose international ties have given them leverage over the Office of the President (p. 201).

Michael Chege (1994) more specifically point to class development in Kenya and its relation to organized opposition when citing that, in Kenya a propertied, self-assured, educated middle class and dejected professional and business groups – bigger and more organized than those of

Tanzania – played a critical role in mobilizing the opposition masses (p. 73).

More recent data suggests that, in Kenya, associations between occupation and political support are less pronounced than in Zambia, but slightly more than Malawi and Tanzania. The biggest gap was found among peasants, subsistence farmers, and pastoralists, which, while composing some 29.6 percent of the sample, made up 34.8 percent of the National Rainbow

Coalition (NARC) support, 25.6 percent of the opposition support, and 26.9 percent of the uncommitted. Breaking this down clarifies that the Liberal Democratic Party (LDP) was the least likely to get support from peasants (17.3 percent), when compared to KANU and NARC

(31.5 percent and 34.8 percent respectively). It is noteworthy however that market oriented peasants are comparably more represented within the LDP (11.1 percent) versus subsistence peasants (5.6 percent). Instead, the LDP was disproportionately represented by wage workers, business owners, and professionals. Slightly over 29 percent of the LDP backing comes from wage earners, 22.8 percent from business owners, and 13.4 percent from professionals. By contrast, only 15.8 percent of KANU’s backing comes from wage workers, with another 17.2

233 percent from business owners, and 10.8 percent from professionals. For NARC, these numbers are 20.2 percent, 16.3 percent, and 8.1 percent respectively.

Table 8.09: Variation in Party Support According to Occupation In percentage points, Std W(AbV) min = 0.00, max = 100.00 Malawi Incumbent Opposition Uncommitted Total Total AbV 11.72 9.10 8.00 28.83 Total W(AbV) 4.06 3.94 1.76 9.77 Std W(AbV) 5.40 5.25 2.34 12.99

Kenya Incumbent Opposition Uncommitted Total Total AbV 13.04 12.11 10.69 35.84 Total W(AbV) 5.25 4.52 2.39 12.17 Std W(AbV) 6.56 5.65 2.99 15.21

Zambia Incumbent Opposition Uncommitted Total Total AbV 12.58 14.54 25.92 53.03 Total W(AbV) 3.29 6.03 8.39 17.71 Std W(AbV) 4.38 8.02 11.15 23.55

Tanzania Incumbent Opposition Uncommitted Total Total AbV 5.62 40.08 24.50 70.20 Total W(AbV) 4.66 2.43 2.69 9.78 Std W(AbV) 6.20 3.23 3.58 13.01

In order to provide a clear and concise comparative summary, I standardized and weighted the Afrobarometer data reported above. The results are posted in table 8.09. Standardizing the data on a scale from 0 to 1006, and weighting it according to overall party support among respondents7 provides a useful summary of the comparative association between occupation and

6 Here, I took the weighted average of the absolute difference (W(AbV)) between party occupational composition and the occupational composition of the sample and multiplied it by the total possible variation. For Kenya, the standardized weighted difference equaled W(AbV)×1.25, and for the remaining three W(AbV)×1.33.

7 The weighted difference equals the absolute value of the difference between party occupational composition and the overall sample composition multiplied by the percentage of respondents in three political categories: incumbent, opposition, and uncommitted. The following is a list of the weighted values for each of the four cases: Malawi 4.1=11.7(.346) for the incumbent, 3.9=9.1(.433) for the opposition, and 1.8=8.0(.220) for the uncommitted; Kenya 5.2=13.0(.403) for the incumbent, 4.5=12.1(.373) for the opposition, and 2.4=10.7(.224) for the uncommitted; Zambia 3.3=12.6(.262) for the incumbent, 6.0=14.5(.415) for the opposition, and 234 party support. What we find is that, when discounting for the overall volatility of political support, which is naturally higher in Tanzania due strong support for the CCM and weak support for the opposition, and when standardizing the data so as to make each case comparable, the total weighted and standardized variation of support is not that strong in any of the cases.

The highest variation of support is, not surprisingly in Zambia, at 23.55 percent, while the lowest are statistical ties between Malawi, at 12.99 percent and Tanzania at 13.01 percent.

Conclusion

What does all this data mean in terms of the persistence of incumbency? For starters, all other things being equal, had capitalist class structures developed in Tanzania to the same intensity found elsewhere, the relationship between political elites at the onset of multiparty reform would have been more competitive. An increased number of well financed but bold business elites of the likes of John Cheyo and Edwin Mtei would have certainly translated into stronger leadership for organizing opposition against the CCM. This is most evident from the tabulations posted in table 8.09. Given the tendencies of those engaged in commercial activities to orient their support away from the CCM, the fact that Arusha stifled the growth of the commercial sector can also be linked to a stifled growth of opposition support from one of the most financial lucrative sections of society.

There two additional highlights to the data presented in this section. First, it would appear that correlations between occupation and political party support in Tanzania, as well as Malawi, are considerably less than what was found in Zambia and slightly less than Kenya. Indeed, in

Zambia, the role of the Zambian Congress of Trade Unions, as an outgrowth of the copper

8.4=25.9(.324) for the uncommitted; Tanzania 4.7=5.6(.829) for the incumbent, 2.4=40.1(.061) for the opposition, and 2.7=24.5(.110) for the uncommitted. 235 mines, provided a powerful and somewhat politically independent force for dragging workers away from UNIP. In no other case do we find an occupational or class organization in a similar institutional position.

Secondly, the ability of class structures to enhance the financial capacity of opposition parties says more about organizational growth and sustenance of opposition and less about the level of contention within society as a whole. While well funded political parties, along with vocal political entrepreneurs are important for cultivating and mobilizing discontent, the spontaneity of popular discontent in the lead-up to reform is difficult to explain through the existence of well financed opposition leaders. Spontaneous political protests, while often fading quickly without appropriate leadership, does not necessarily require well financed leadership in order to take to the streets. In short, how can we explain the sheer lack of popular protests against single-party rule, or the fact that so many Tanzanians rejected the need to go multiparty in the first place? The answer I argue is that, in comparative perspective, the Tanzanian polity offers comparatively less ideational space for legitimating discontent.

236

CHAPTER NINE INEQUALITY AS THE BASIS FOR CONTENTION

“The essence of socialism is the practical acceptance of human equality. That is to say, every man’s equal right to a decent life before any individual has a surplus above his needs; his equal right to participate in Government; and his equal responsibility to work and contribute to the society to the limit of his ability.”

Julius K. Nyerere, 1967

“Poverty is the parent of revolution”

Aristotle, circa 350 BC

The train ride from Harrisburg, the small state capital tucked in at the foothills of

Pennsylvania’s Appalachian Mountains, to the historic city of Philadelphia, provides about a 90 minute panorama of life in Northeastern United States. From the Amish farmlands of Lancaster

County and the cozy but crammed boxes and lawns of middle-class delight, through the aching remains of the once booming steel town of Coatesville and the spectacular amenities of the six- bedroom pads of the Mainline, the ride comes to an end at 30th Street Station just after a 5 minute tour of the run-down, seemingly inhabitable row houses in West Philadelphia. The awe of life’s asymmetries is constantly at the mind’s doorstep with each passing moment. Such are the scenes throughout so many sprawling urban and suburban areas in North America and

Western Europe.

A 30 minute rib-jabbing daladala ride from the Bahari Beach suburbs of Dar es Salaam to

Ubungo elicits similar awe, albeit to unimaginably greater extremes. Riding past tightly gated mansions with state of the art security systems and well manicured palm trees, only to squeeze out of the steam and sweat of the an overloaded van amid some of the most dire living conditions on the planet, is really a breathtaking, if not heart wrenching experience. This same tour can be experience in nearly any urban sprawl in Africa, from Dakar to Durban.

The severity of these extremes is, without a doubt, the most significant challenge facing

Tanzanians today. Yet, the severity of inequality is a more recent phenomena in Tanzania, scenes of which today are far more intense than those I found during my first visit to the country back in 1999, or those stories I’ve heard about the 1970s and 80s. Back then, Tanzanians were almost universally poor, save for a relative few who managed to flout their circumvention of the

Leadership Code.

The scope of this chapter is to explicate a broader understanding of the political ramifications of economic development in Tanzania. Whereas those economic factors defined in the previous chapter speak to those material capacities for mobilizing and organizing dissent, i.e. a unit-level property of sorts, economic factors in this section are treated in a more systemic fashion, as an objective source for shaping the perceptions of fairness and justice, which, in turn, translates into varying grievance levels which in turn raises the likelihood of protest. Stated differently, whereas the ability of an opposition party to tap into labor unions or chambers of commerce might say a great deal about the party’s long-term organizational prospects, perceptions of economic fairness and justice speak more directly to the size or intensity of political protests, which may or may not have long-run organizational viability. Perceptions of economic injustice can certainly raise the likelihood of an explosion of political unrest directed against the regime, state, or dominant class. Drawing on the data from the previous section, and the theoretical understanding that greater inequalities of income and wealth leads to greater probabilities of contention, it can be expected that perceptions of economic performance in

Tanzania should have weaker associations with political orientation.

238

The CCM’s Populist Adversary

If the hegemony of the CCM in the lead-up to political reform had an aura of inevitability, with the sinking popularity and an ideological staleness discussed previously, the road to victory for the CCM during the year of the first multiparty election seemed far more tentative. While the popular energies around political reforms left a lot to be desired, the atmosphere around

Augustine Mrema’s presidential candidacy in 1995 was filled to the brim with hope and jubilation.

Mrema, who was trained at the Kivukoni Political Education College near Dar es Salaam, has been an active politician and CCM stalwart since his first political bid for a parliamentary seat in his home region of Kilimanjaro. He would however, prove to be a distinct source of political trouble for the bureaucratic bourgeoisie due his continual outspokenness against high level corruption. Due to his anti-corruption campaign, in 1995, he was sacked from his position as Minister of Labor. Disgruntled with the leadership within the incumbent party and his loss of status, Mrema left the CCM just prior to the first post-reform multiparty election. He took up a position as Chairman and Presidential candidate for the NCCR-M before factional struggles between him and Secretary General Mabere Marando convinced Mrema in 1999 to yet again defect, this time taking up a position in the Tanzanian Labor Party (Whitehead, 2000).

The source of Mrema’s popularity lay first, in his history of being tough on crime while sitting as the Home Affairs Minister and opposing what were perceived as corrupt party “big- wigs” during his tenure within the CCM. And after his defection from the CCM, his popularity was given renewed life when he charted his political position as the ‘class’ enemy of the ruling bourgeoisie (CCM delivers Mrema, 1995). In what was clearly an attempt to capture much of the broad peasant base which upheld their support for the CCM in the past, the main thrust of

239 his campaign while in the NCCR-Maguezi, and later, the TLP, was characterized as an active

“liberation” from the mismanagement of the CCM (You’ll live to regret, 1995), and restoration of the past by “recapturing the country’s lost glory” (Thomas, 1995). Indeed, Mrema was single- handedly trying to reinvigorate some of the earlier socialist debates that took place within the corridors of TANU as a tool for delegitimizing the hegemony of the bureaucratic bourgeoisie.

To this end, Mrema framed an argument in a 1995 debate by suggesting that socialism was never a reality in Tanzania. Instead, poverty was brought about by wealthy party and commercial cliques, who lived off the work of the poor (Socialism crucified, 1995). Mrema has consistently maintained that the danger for the country was not found in the possibility for ethnic strife, but the dangers of a growing gap between haves and have-nots. “It is very dangerous when you have a class of people who are not certain of their lives and another of people literally swimming in wealth and leading extravagant lives” (as cited in Thomas, 1995).

Secondly, Mrema’s rhetoric on Africanization, now reframed under the term

“indigenization,” threatened to reopen the social fissures that, at the time of independence, acted as a major sticking point between labor unions and TANU, and part of the thrust behind the

January, 1964 mutiny. The electoral advantages of evoking indigenization were clear. While indigenization might scare away the sizable Indian population living along the coastal areas in the north, the appeals to racialized class hierarchies found in the indigenization debate was believed to be capable of attracting support from the larger community of urban Africans, whose only relation to the Asian communities took place as the consumer of basic home and job supplies, and in some cases as employees. Given the history of the disproportionate participation of Asians in commerce, it would appear that Mrema’s class and racial rhetoric would certainly have broad appeal.

240

Comparatively speaking, there was nothing unique about Mrema's efforts to couch political conflict as a battle between the rich and the poor. Kenneth Matiba, the Forum for the

Restoration of Democracy-Asili’s (FORD) 1992 presidential candidate in Kenya, built support by appealing to the poorer sections of Kikuyu society through a popular discontent against commercial and state elites. More recently, Raila Odinga, the presidential aspirant from Kenya’s

Orange Democratic Movement (ODM) has managed to draw considerable support outside his ethnic base by defining elections as contests between haves and have-nots. In Zambia, Michael

Sata, the 2006 presidential contender for the Patriotic Front (PF) cried out on behalf of the poor against economic adjustment policies, the growth of economic gaps, and the increased Chinese dominance of the economy.

The final chapter of Mrema’s political saga however, was a disappointment, a point captured well by the pie charts in appendixes D through I. While he polled an impressive 27.8 percent of the vote in the 1995 Presidential election, in the 2000 election he polled a paltry 7.8 percent, and by the 2005 presidential race, he polled an embarrassingly insignificant 0.8 percent. Even within his home region of Kilimanjaro, where he polled 77.7 percent in 1995, and 41.3 percent in 2000, by the 2005 election he polled a mere 3.5 percent.

As shown in appendix I, Matiba’s performance in the 1992 election was only marginally better than Mrema’s and, with the exception of his minor come back in the 2007 election, has been notably absent from the Kenya’s political scene after the 1992 election. But, other politicians have carried the class message forward, most notably Raila Odinga, and have had considerable luck with it. By contrast, in Tanzania, others have attempted to carry this message of populism against the wealthy commercial and bureaucratic bourgeoisie elites, but with no success. When considering that Mrema’s popularity once made up nearly 73 percent of the

241 opposition support in the presidential contests, and that the electoral gains made by CHADEMA and CUF since the 1995 election do not compensate for Mrema’s losses, the decline of Mrema marks the decline of the opposition generally. What happened to Mrema and why has no other party filled the void?

An obvious answer is repression. Without a doubt, his 1995 campaign took the brunt of the

CCM onslaught. In Dodoma, Sumbawanga, Dar es Salaam, Tanga, and twice in Mushi, Mrema was tear gassed by the police and by what some described as “CCM zealots” (Dismas, 1995b).

Ultimately, the build-up of toxic residues required the candidate to seek medical attention as a remedy for his degraded eye sight (Mwanjisi, 1995). In a separate complaint, the NCCR-M campaign chairman Professor Mwesiga Baregu claimed that Edward Lowassa allocated some 50 million shillings to have Mrema killed (Mbogora, 1995).

However, repression still leaves unanswered as to why, despite Mrema’s persistent attempts at the presidential bid, and the persistent attempts by the opposition as a whole, does his electoral support decline so rapidly? After all, can we really expect that, given the discussion in chapter four, placing Mrema and his closest allies in the sights of the state’s coercive arm will result in such a rapid and significant decline of support?

Economic Inequality

The rise of popular discord and popular electoral support requires focal points or themes for establishing a framework that defines the causes and perpetrators of society’s ailments. Corrupt state leaders for example, was defined by Mrema as the source of society’s divisions between haves and have-nots. Frameworks with the saliency to compete with other campaign themes are not however, invented out of thin air by a few candidates. This is not to suggest that corruption

242 or wealth inequalities in Tanzania are not pressing issues. Instead, it suggests that successful campaign frameworks will generally reflect a socially defined hierarchy of threats and ailments.

Let us now begin to undercover the reality of these ailments by taking a cue from Mrema: the divisions between haves and have-nots.

While most regimes in post-independence Africa, even among those espousing African socialism, did little to relieve radically inegalitarian distributions of capital and income, egalitarian considerations in Tanzania were placed alongside the need to consider things such as rapid economic growth and development (Okoko, 1987, p. 20). Nyerere repeatedly expressed a deep concern over the political instability that would be caused by long-term inequalities of wealth and polarization of politics.

Nyerere’s concern was undoubtedly a reflection of the economic reality in Tanzania during the early 1960s. As depicted in table 9.01, at the time of independence, income inequality in

Tanzania was among the highest in the region. At independence, the income share among the top 5 and 10 percent was highest in the sample, and income concentration only worsened by the end of the 1960s. The same pattern emerges from the Gini Coefficient data. In short, not only was Tanzania one of the most inegalitarian countries in the sample below, but was growing more inegalitarian following independence.

The policies set forth by the Arusha Declaration altered this pattern of inequality. In the late

1970s, Nyerere boasted that the salary ratio shrank from 50:1 in 1961 to 9:1 in 1976 (as cited in

Nugent, 2004, p. 146). Equally important, post-Arusha policies were successful in strengthening the rural standard of living relative to the standard of living for urban workers (Nursey-Bray,

1980, p. 57), a trend not commonly obtained in Africa. Even those employed in the state sector and with political connections, were comparatively constrained in their ability to accumulate

243 wealth. In 1977 the World Bank cited that, since 1967, the lot of peasants increased faster than the lot of urban workers (as cited in Okoko, 1987, p. 117), again a trend quite uncommon in

Africa. When commenting on the standard of living within the bureaucratic bourgeoisie,

Nursey-Bray (1980) states that “those privileged to indulge in luxury goods consumption have had to do so, by virtue of the prevailing ideology, somewhat more circumspectly and in fewer numbers than, say, in neighboring Kenya” (p. 63).

Table 9.01: Household Income Inequality at Independence Kenya Malawi Botswana Tanzania Tanzania Zambia (68‐69) (69) (71‐72)₁ (59) (69) (59) 0‐10 1.5 2.3 0.2 2.3 0.8 2.6 10‐20 2.4 3.4 1.4 2.9 1.5 2.8 20‐30 3.4 4.3 2.5 3.5 2.3 3.5 30‐40 4.4 5 3.5 4.2 3.2 4.1 40‐50 5.7 6 4.9 5.2 4.3 4.9 50‐60 7.4 7.1 6.4 6.2 5.8 6.2 60‐70 9.6 8.5 8.8 7.8 7.9 7.7 70‐80 12 10.5 12 9.8 10.9 10 80‐90 18.7 14 18.2 13.5 16.6 14.2 90‐100 33.9 38.9 42.1 44.6 46.7 44

95‐100 20.2 29.5 28.1 35.1 33.5 33.7 Gini Coeff 0.479 0.4696 0.574 0.5282 0.5973 0.5226 Source: Jain, 1975

Whereas the provision of social services in Kenya greatly favored those regions with the ability to exert leverage through political networks, in Tanzania, development assistance and social services were focused more stridently on uplifting those regions historically excluded from development. For example, while the expansion of private secondary sector education during the 1980s highlighted the incomplete nature of socialist transformation and acted to

244 undermine some of the leveling effects of public education institutions and policies1, educational achievements following independence were “most impressive in those parts of the country which had historically received minimal access to education” (Nugent, 2004, p. 148).

Efforts directed toward universal primary education and eradicating adult illiteracy was positively linked with declines in regional socio-economic asymmetries. Free secondary school, plus a quota system which granted secondary school slots in proportion to the number of primary school leavers in each region, further reduced of asymmetries (Buchert, 1994, p. 113).

The conclusions reached by Barkan (1979a) on the state of inequality in Tanzania are perhaps less enthusiastic than those conclusions stated above. He cites that, while the taxation rate in Kenya is far less progressive than the rate found in Tanzania, and while the disparity between the highest and lowest paid government employee fell from 20:1 to 9:1, “members of the bureaucratic class in Tanzania receive a variety of ‘perks’…which result in an overall pattern of inequality that is not substantially different from that in Kenya” (p. 23). Furthermore, despite the leveling efforts in access to education, those regions in Tanzania with a history of cash crop production managed to rally enough capital to establish a “web of institutional structures that reinforce their dominant position,” with the result that individuals from cash crop regions are disproportionately represented in higher education and the civil service (p. 25).

Nonetheless, Barkan does conclude that the level of inequality in Tanzania during the 1970s was “not increasing, as in Kenya, or, if it is increasing, it is not rising at as fast a rate” (p. 23).

Policies such as nationalization, tightly regulated commerce, and the Leadership Code, have simply “forestalled the emergence of a commercial petty bourgeoisie in Tanzania comparable to

1 Buchert (1994) reports that the five specific regions of Kilimanjaro, Iringa, Kagera, Mara, and Mbeya, where the formation of the petty bourgeoisie were strongest, “accounted for nearly 60 per cent of all form V students, but less than 30 per cent of the relevant national cohort” (p. 114).

245 that in Kenya” (p. 23). And, because bureaucrats and party officials in Tanzania have less disposable income than their counterparts in other African countries, their consumption habits have been far more modest in comparative terms. Barkan concludes that, given this modesty, government and party officials “are not perceived as ‘ripping-off’ the rest of Tanzanian society…” (p. 23).

By-in-large, this comparative equality has carried over into the post-political reform era, albeit with a clear reversal in the direction toward growing inequality. Taking a look at the

World Bank (2007e) data on income inequality for sub-Saharan Africa between 1989 and 2004, we find that the poorest 20 percent in Tanzania controlled 7.3 percent of the income, compared to 5 percent for the rest of the African sample (n=29). Similarly, the top 20 percent in Tanzania controlled some 42.4 percent of the income, compared to average of 53 percent for the remaining cases.

Table 9.02: Today’s Economic Inequality Percentile Rank (1989‐2004) Country Lowest 20% Second 20% Third 20% Fourth 20% Highest 20% Lesotho 7.1% 100.0% 3.5% 17.8% 96.4% Botswana 17.8% 10.7% 10.7% 3.5% 92.8% Namibia 3.5% 0.0% 0.0% 0.0% 100.0% S Leone 0.0% 67.8% 57.1% 50.0% 85.7% Ethiopia 100.0% 96.4% 100.0% 71.4% 0.0% Benin 96.4% 89.2% 92.8% 71.4% 7.1% Tanzania 92.8% 92.8% 96.4% 89.2% 3.5% Malawi 89.2% 82.1% 78.5% 53.5% 21.4% Cameroon 60.7% 50.0% 50.0% 42.8% 50.0% Kenya 67.8% 3.5% 60.7% 57.1% 39.2% Zambia 25.0% 21.4% 25.0% 60.7% 71.4% Source: World Bank, 2007e

Some simple statistical work with the World Bank data puts Tanzania in broader perspective. As posted in table 9.02 above, which catalogues the comparative ranking of

246 incomes earned among each quintile, the poorest 20 percent of the Tanzanian population owns more share of the country’s income than poorest segments of the population in nearly 93 percent of the remaining countries (n=29). Meanwhile, the highest 20 percent earned less of the

Tanzanian national pie than the richest 20 percent in any other African country, save for

Ethiopia. Among those cases of comparative interest, Zambia appears as highly inegalitarian,

Kenya as modestly inegalitarian, while Malawi more closely resembles the egalitarianism in

Tanzania.

A comparative glimpse at the Gini coefficient is just as telling. Presented in table 9.03, we see that income dispersion in Tanzania is far more egalitarian than what is typically found on the

Continent. In fact, Tanzania ranks as having one of the most equally distributed incomes throughout Africa (2nd only to Ethiopia), while Namibia, Lesotho, and Botswana are the most unequally distributed. Among those cases of most comparative interest, Zambia ranks as highly inegalitarian, Kenya as moderately inegalitarian, and Malawi as relatively egalitarian.

Survey data on perceptions of inequality does not exactly match what might be expected to evolve out of the Gini coefficients reported in table 9.03, especially in relation to perceptions in

Kenya and Malawi. Yet, data from Afrobarometer surveys do indicate quite clearly that perceptions of inequality are comparatively low in Tanzania. For example, when asked to rate their personal standard of living, the respondents from the Tanzanian sample, along with Kenya, were far less likely to rate their positions as worse or much worse compared to cohorts from the

Zambia and Malawi sample. As presented in table 9.04 below, respondents in Malawi and

Zambia were far more likely to give a negative evaluation of their economic lots when compared to fellow country folk, suggesting that respondents in both cases were more likely to feel a sense of economic inequality than respondents in Kenya and Tanzania. Most notably, over

247

43 percent of the Tanzanian respondents rated their living conditions as the same as others, versus 39 percent for Kenyans, 18 percent for Malawians, and 17 percent for Zambians.

Table 9.03: Gini Coefficient for Income Country GINI (1989‐2004) Percentile Rank Lesotho 63.2 96.5% Botswana 60.5 86.2% Namibia 74.3 100.0% Sierra Leone 62.9 93.1% Ethiopia 30 0.0% Benin 36.5 6.8% Tanzania 34.6 3.4% Malawi 39 13.7% Cameroon 44.6 44.8% Kenya 42.5 37.9% Zambia 50.8 79.3% Average 47.2 Source: World Bank, 2007d

Table 9.04: Your own Standard of Living Compared to Others Kenya (2003) Malawi (1999) Zambia (1999) Tanzania (2001) Much Worse 3.8% 18.5% 23.9% 8.0% Worse 21.7% 34.6% 34.7% 28.0% Same 39.3% 17.7% 17.2% 43.1% Better 31.6% 24.9% 20.2% 18.2% Much Better 3.6% 4.3% 4.0% 2.7% n 2282 1205 1169 2185 Source: Afrobarometer, 1999a, b, 2001, 2003

Additional data goes to support the claim that perceptions of economic equality in Tanzania are quite real but comparatively less so. The most noteworthy collections of numbers, which reflect respondent ratings of the government’s success at narrowing income gaps between rich and poor, clearly demonstrate that the Tanzanian cohorts are far more likely to register positive assessments. As table 9.05 shows, the majority of the respondents in Kenya, Malawi, and

Zambia rated government performance as very badly, while slightly over one-third of the

Tanzanians gave their government the same rating. While 35 percent of the Tanzanian sample gave the government a “fairly well” or “very well” rating, only 17.9 percent of the Kenyans,

248

19.5 percent Malawians, and 8.4 percent of the Zambians gave their government similarly positive ratings.

Table 9.05: Ratings of Government efforts at Addressing Income Inequality Kenya Malawi Zambia Tanzania Very Badly 51.3% 69.0% 61.4% 36.5% Fairly Badly 30.7% 11.5% 30.1% 28.2% Fairly Well 16.5% 13.4% 8.0% 27.1% Very Well 1.4% 6.1% 0.4% 8.2% n 1233 1144 1171 1091 Source: Afrobarometer, 2005b, c, d, e

Inequality and Political Orientation

How do perceptions of inequality translate into political divisions? Table 9.06 below lists those Afrobarometer economic variables which were found to have statistical significance when correlated against indicators of political orientation. In the three categories where significant associations were found across all four cases, Tanzania consistently ranks among the lowest.

Tanzania had, by far, the weakest association between evaluations of the government’s success in job creation versus trust in the ruling party (0.071), the weakest association between the evaluation of the government’s success at economic management versus trust in the ruling party

(0.177), and second weakest association between evaluations of the government’s ability to meet educational needs versus trust in the ruling party (0.138).

For Tanzania, the strongest two associations found during the correlation study were perceptions of economic management versus trust in ruling party (0.177) and perceptions of the state of education versus trust in ruling party (0.138). In other words, those respondents which had positive perceptions of economic management tended to place more trust in the CCM. But negative perceptions on the other hand, had no significant relationship with perceptions of the opposition. Likewise, the positive assessments of education were positively correlated with

249 greater levels of trust in the CCM. Again however, no significant relationship was found with the opposition. Similar to the tendencies found in the previous chapter, where greater education levels were associated with weaker party affiliation, negative assessments of government policy performance here appears to have no clear political direction.

Table 9.06: Correlation Coefficients for Significant Economic Indicators Kenya Malawi Zambia Tan Economic conditions v trust in ruling party 0252 0.118 0.118 Economic conditions v trust in opposition ‐0.065 Economic conditions v close to party 0.069 Your econ conditions v trust in ruling party 0.165 Your econ conditions v trust in opposition ‐0.091 Creating jobs v trust in ruling party 0.331 0.153 0.248 0.071 Creating jobs v trust in opposition ‐0.060 ‐0.063 Creating jobs v close to party 0.066 Educational needs v trust in ruling party 0.320 0.197 0.093 0.138 Educational needs v trust in opposition ‐0.136 0.076 Educational needs v close to party 0.120 0.105 Managing economy v close to party 0.107 0.061 Managing economy v trust in ruling party 0.417 0.214 0.261 0.177 Managing economy v trust in opposition ‐0.113 ‐0.064 Econ policies help/hurt v. close to party ‐0.091 ‐0.114 Econ policies help/hurt v. trust in ruling party ‐0.228 ‐0.1 ‐0.086 Econ policies help/hurt v. trust in opposition Present living conditions v close to party ‐0.055 Sum of absolute values 2.540 0.837 1.115 0.651 Source: Afrobarometer, 2005b, c, d, e

The correlation coefficients for the three other cases are, save for Malawi, substantially stronger than Tanzania. In Kenya for example, the more positive the assessment of economic management, the greater the trust in the ruling party (0.417), the lower the trust in opposition parties (-0.113), and the closer respondents felt to a political party (0.107). In Zambia, positive assessments of economic management and job creation are associated with greater levels of trust in the ruling party (0.261 and 0.248 respectively). Finally, in Malawi, positive evaluations of economic management, meeting educational needs, and job creation are positively associated

250 with trust in the ruling party (0.214, 0.197, and 0.153 respectively), while being negatively associated with trust in opposition parties. When summing up the absolute values of all the significant associations, Tanzania scores the lowest, followed closely by Malawi. Thus, the data above clearly indicates that perceptions among Tanzanians of their personal economic standing and the general standing of the economy do not strongly correlate with political orientation.

When they do, positive evaluations usually go to benefit the CCM while negative evaluations have no significant correlation with opposition support. Again, this could be a product of the comparative weakness of opposition parties to construct a counter-hegemonic explanation of the economic plight faced by so many Tanzanians. But, it could also explain the inability of the opposition to successfully sell an alternative account in the first place. More specifically, latent economic grievances are, quite simply, less powerful of a political motivator, and thus more difficult to frame and mobilize as a basis for viable political competition. The overall comparative breadths of popular protests since the appearance of the winds of multiparty change suggest that the latter interpretation carries more weight.

Conclusion

There were three data clusters presented in this chapter that suggest quite clearly that the economic leveling policies pursued under auspices of African Socialism, especially those formulated from the Arusha Declaration, had clear consequences for the overall level of macro- structural divergence and the success of micro-level references to class as a basis for mobilizing political support. First, economic inequality in Tanzania is significantly lower than the levels of inequality in almost every other sub-Saharan African case, as represented by the data presented in tables 9.02 and 9.03. Secondly, perceptions of economic inequality are likewise, considerably

251 lower than the same perceptions found in Kenya, Zambia, and Malawi, as demonstrated in tables 9.04 and 9.05. Finally, as the data presented in table 9.06 clarifies, the tendency for variations in economic conditions to translate into variations in political orientation is substantially lower in Tanzania when compared to Zambia and Kenya, and slightly lower when compared to Malawi.

These macro-structural consequences of African Socialism suggest two things. First, there was a comparatively weaker basis in Tanzania for the rise of spontaneous political protests as a response to feelings of class discontent. Secondly, this also suggests that campaign tactics which incite economic differentials will not have the saliency for mobilizing supporters as similar tactics will in other countries. Yet, Mrema did just that. During the 1995 campaign, he repeatedly evoked the language of haves and have-nots similar to the styles adopted by Matiba in Kenya or Sata in Zambia. Given the fact that real inequalities and perceptions of inequalities are comparative low in Tanzania, and the fact that economic factors do not appear to orient voters’ political preferences, it is no wonder that Mrema’s campaign lost so much steam following his 1995 defeat.

When combined with the previous chapter, this chapter clearly points to the material and ideational consequences of economic policy pursuits following Arusha. What I will attempt to do in the next two chapters are to evaluate the effects that the political pursuits of the party leadership stratum during single-party rule has had on ethnic and religious identity and the prospects for mobilizing political support around these identities.

252

CHAPTER TEN ETHNCITY AS A BASIS FOR CONTENTION

“Indeed, a Deputy must be mandated by the Nation as a whole and have no other concern than to better serve the national interests without allowing himself to be influence by considerations of region, clan, race, or religion.”

Sékou Touré, 1962

“The chief is the chief. He is the eagle who flies high and cannot be touched by the spit of the toad.”

Mobutu Sésé Seko, 1991

[Ochieng] “Hello Matunda! You might not remember me, but we met last week on the matatu.” [Matunda] “Yes! I remember you! We talked about your business plans.” [Ochieng]

“Yes, I’m the one! Maybe we can get together sometime this week and discuss a business proposition. What do you think?” [Matunda] “Hmmm…where are you from anyway?”

[Ochieng] “I’m from Kisumu.” [Matunda] “Ahhhh! My Lou brother, I’m from Maseno. Yes, let’s meet. I just wanted to make sure you weren’t Meru. You know they can’t be trusted.”

While this conversation is fictional, it is an accurate portrait of what the birth pangs of a business partnership might look like in downtown Nairobi. The question of ethnic background is cleared quite quickly, assuming that the Matunda couldn’t tell by appearance or dialect. The willingness to enter into an agreement is largely lubricated by a shared ethnic background and undermined by being from the ‘wrong’ ethnic group. Had Ochieng been a Meru, Matunda would have polity declined a meeting and concocted a story about some urgent thing he needed to attend to.

Had this conversation been carried out in Dar es Salaam between two Tanzanians, the question of ethnicity would surely come up, but probably not as quickly as in Nairobi. Likewise, the dialogue between Matunda and Ochieng in Nairobi, while initially carried out in English or

Kiswahili, would have quickly shifted into Dholuo. Despite having a shared ethnic background, the dialogue between the Tanzanian Matunda and Ochieng in Dar es Salaam might not have shifted as quickly into a local language, instead being carried out in Kiswahili. Furthermore, differences in ethnic background would probably not have taken priority over the possibly of making money through a business deal.

The objective of this chapter is to comparatively examine the relationship between ethnic background and political persuasion. Given the discussion of the efforts at delegitimizing ethnic loyalties under single-party rule, we should not be surprised to find that ethnicity as a determinant of political support in Tanzania to be far lower than Malawi, Zambia and Kenya.

Indeed, the data in this chapter supports this claim. What is perhaps more important is to insert a counterfactual. Had African Socialism aggravated rather than mollified ethnic identity, could we expect the CCM’s dominance to be as clear-cut as it is today? Before moving directly to the data, a brief discussion on ethnicity is in order.

Ethnicity in Theory

Throughout the African Continent, distinctions based on language, beliefs, and practices, have acted as powerful focal points for mobilizing collective action in the political arena. As defined in primordialist terms, clusters of shared practices and beliefs among a particular community have deep, longstanding meaning, often ascribed at birth and in later socialization of community customs and language (Geertz, 1963; Horowitz, 1985). From this perspective, a special problem posed by ethnicity is that, since it is an immutable and non-negotiable social

254 category, it is also inherently rife with conflict in conditions whereby competition over state policy requires bargaining and negotiation.

Yet, the idea that ethnicity is long-standing and immutable is problematic since it simply assumes the existence of ethnic solidarity without ever attempting to explain solidarity. In other words, ethnicity, which is a sociological phenomenon, is placed prior to actions among individuals. The problem here is quite simple. While ethnicity at any given point in time may have causal force, the intensity of that force is defined by an antecedent political process based on interests, which might have little or nothing to do with ethnic concerns (Newbury, 1988). It was the colonial ruler’s concerns about maintaining social control for extracting economic returns which encouraged political interaction that accentuated myths of inter-ethnic difference.

In post independence times, ethnicity has more frequently than not been used by political elites to develop support bases for capturing political office, having little or no real anxieties over things like cultural practices and meanings.

In reality, ethnicity is more appropriately understood as having a strong degree of meaning fluidity and mutability, whose boundaries are divided or consolidated by the an exchange between macro-level structures, including myths, practices, religion, language, class, and wealth, versus micro-level interactions between groups and political elites over the control of state power. Hence, the movement of ethnic boundaries is dependent both upon the saliency of ethnic discourses (i.e. it cannot be invented out of thin air) and elites that perceive individual or group advantages out of organizing and mobilizing ethnic movements. (Newberry, 1988;

Esman, 1994; Rothchild, 1997). Stated differently, while political actors are often responsible for whipping nebulous ethnic meanings into a frenzied and rigidly defined ethnic conflict, the ability to do so depends on the saliency of the monologues these entrepreneurs are trying to sell,

255 a saliency that is defined by knowledge structures which correspond with regional and language differences or differences in wealth, political, or access to education.

Ethnicity in Tanzania

In some respects, urban settings will teach a person more about inter-ethnic relations than staying in the more remote areas of the countryside. During my stay at the YWCA in downtown

Nairobi, I was asked by one of the YWCA clerks if I would be interested in going out with a few friends for some drinks in one the less central parts of the city. Indeed, the clerk was offering a service, considering that, as a tourist, going out alone at night in Nairobi can be quite a risky move. After we negotiated over the number of beers I would by for him, we agreed on a night and a place to meet.

On the night of the outing, the group of us met and journeyed to the Kamukunji area, just off to the western side of the city center, and found a crowded, and somewhat noisy and seedy place with live music to sit, drink, and converse. By the end of the night, after I had learned a few Kimeru words, I realized that, not only were all of my companions Meru, but most of the people in the pub were Meru as well. In all of my times in Dar es Salaam, I never found a single place frequented only by customers from one specific region or with one specific ethnicity.

When I entertainingly pointed out this difference to one of my companions, he simply laughed and said “welcome to Kenya”.

I am undoubtedly not the first tourist to have this experience, nor am I the first person to acknowledge the policy success that African Socialism, broadly defined, had on undermining the saliency of ethnicity in Tanzania. As discussed in chapter six, post-independence development policy in Tanzania impacted the prevalence of ethnicity as an organizing and

256 mobilizing feature of national politics. First, the eradication of customary rule and the curtailed growth of a rural petty bourgeoisie, translated into a lack of elite leadership for the articulation, mobilization, and consolidation of ethnicity. Secondly, the fact that development policy after the

Arusha Declaration was not obviously beneficial only for a select few regions meant that popular resentment toward ethnic exclusion lacked any real meaning or saliency.

These points are echoed in Alfred Sebit Lokuji’s (1995) statement that a single-party leadership committed to arresting “the development of politics centered on ethnic or regional affiliation” is the primary source for the minimal role that ethnic politics plays in post-reform

Tanzania. Professor Mpangala (2000) further adds that, when compared to other countries of the

Region [the] introduction of multiparty politics in Tanzania Mainland gave rise to far less ethnised multiparty politics” (p. 67). This is not to suggest that ethnic identification is a minor part of Tanzanian social life. Okema (1996) makes it abundantly clear that ethnic identification is indeed an important social marker. “Wewe Kabilia gani (What ethnic group are you)?” is a frequently heard question during introductions. Houses and neighborhoods are identified by the tribal backgrounds of their owners and inhabitants (74). However, ethnicity falls short of being elaborated as a site of vertical stratification, and thus a common justification for hostility toward the other. Okema (1996) again provides a nice summary by stating that:

Ethnic sentiments are very much a reality in Tanzania and are becoming more so with time. However, unlike in many countries in Africa they are not allowed to play a major role in national politics, nor have they ever been part of official policy – written or unwritten (p. 74).

Additional research supports these claims. Based on a comparative study of two districts, one in Kenya and one in Tanzania, Miguel (2004) found that ethnic diversity leads to fewer collective action problems in Tanzania relative to Kenya. As the author explains, because cooperation is more difficult where preferences are diverse and social sanctions difficult to

257 enforce, ethnically heterogeneous communities in Kenya fair worse economically than homogeneous ones, largely due to the failures to cooperate in fund-raising activities. In

Tanzania by contrast, the level of diversity has no significant relationship with the level of economic attainment and fund-raising since ethnicity in Tanzania is less likely to produce collective action failures (pp. 328-29). Let us now turn to an empirical discussion of some survey data.

Data on Ethnic Saliency

Evidence of ethnicity as a source for social divisiveness in Tanzania is borne out from a number of REDET studies. In a 2001 opinion poll, respondents were asked about the prospects of using local ethnic vernaculars in election campaigns. An overwhelming 73.8 percent were against the use of vernaculars in election campaigns, although Keera registered a 70 percent support (REDET, 2001, p. 10). In an earlier survey, some 7.0 percent, 7.5 percent, and 7.0 percent out of a sample of nearly 4,900 Tanzanians admitted to belonging to a workers union, religious organization, or a co-operative union respectively. However, only 0.2 percent, 0.1 percent, and 0.2 percent tell of belonging to an ethnic, regional/district, or clan organization respectively. As presented in table 10.01, the same survey also shows that, while 44 percent would approve of their children marrying across religious lines and some 47 percent would feel the same about cross-party marriages, some 73 percent would approve of marriage across ethnic lines (Mushi, Mukandala, Baregu, 2001, pp. 164-71).

Afrobarometer data confirms the REDET findings, and permits a discussion of ethnicity on comparative grounds. For starters, the survey results presented in table 10.02, which tabulate responses to perceptions of ethnic group conditions relative to other ethnic groups, illustrate that

258

Malawians are far more likely to perceive economic inequalities as corresponding with ethnic differences, while Zambians are the least likely. When asked to rate the economic condition of their own ethnic group relative to other ethnic groups, only 8.8 percent of the Tanzanians rated their own group as “much worse” compared to 38.5 percent of the Malawi respondents. In fact,

Tanzanians and Zambians are the least likely to perceive the extremes of “much worse” and

“much less.”

Table 10.1: Opinions of Inter-cleavage Marriage Approval of: Percent Approval Marriage across religion 44% Marriage across parties 47% Marriage across races 64% Marriage across ethnic group 73% Marriage in higher class status 75% Marriage in lower class status 78% Source: Mushi, Mukandala, Baregu, 2001, p. 171

Table 10.02: Own Ethnic Group’s Relative Economic Conditions Kenya Malawi Zambia Tanzania Much Worse 17.5% 38.5% 3.6% 8.8% Worse 27.7% 22.3% 15.3% 35.9% Same 32.1% 23.2% 67.0% 33.7% Better 19.9% 11.5% 13.1% 19.6% Much Better 2.8% 4.4% 1.0% 2.1% n 1238 1152 1154 1199 Source: Afrobarometer, 2005b, c, d, e

A similar trend is shown in table 10.03. Here, Malawians were far more likely to feel ethnic inequalities in political influence, while Zambians were the least likely. Again, Tanzania and

Zambia were the least likely to perceive extremes. When collapsing the categories into “more,”

“same,” and “less” Tanzanians are still significantly less inclined to perceive political inequality when compared to Kenyans and Malawians.

The following two tables are perhaps the most telling and the most important since the nature of the questions get to the heart of the topic of this section. For the data in the first table

259

(10.04), respondents were asked whether or not they experienced unfair treatment based upon ethnic considerations. The experience of unfairness is perhaps more important than the experience of inequality since unfairness tells a bit about the degrees to which normative structures legitimate inequality. Again, Zambians and Tanzanians stand out as the least likely to experience unfair treatment based upon ethnicity. Strikingly, some 54.1 percent of the

Tanzanian respondents reported “never” experiencing unfair treatment based upon ethnicity, compared to 23.6 percent of the Kenyan respondents. On the other hand, while 22.4 percent of the Malawian respondents reported “always” experiencing unfair treatment due to ethnicity, that same experience was reported by only 4.4 percent of the Tanzanian respondents. When taken altogether, out of all four cases, Tanzanians are by far the least likely to experience unfair treatment based on ethnicity.

Table 10.03: Own Ethnic Group’s Relative Political Influence Kenya Malawi Zambia Tanzania Much More 6.4% 24.6% 3.3% 5.0% More 19.5% 11.2% 11.0% 17.2% Same 33.7% 29.3% 61.3% 47.8% Less 27.4% 17.6% 19.0% 25.1% Much Less 12.9% 17.2% 5.4% 5.0% n 1197 1129 1129 1131 Source: Afrobarometer, 2005b, c, d, e

Table 10.04: Own Ethnic Group Treated Unfairly Kenya Malawi Zambia Tanzania Never 23.6% 31.7% 45.1% 54.1% Sometimes 40.5% 23.1% 40.5% 27.1% Often 20.5% 22.8% 10.1% 14.4% Always 15.5% 22.4% 4.2% 4.4% n 1144 1077 1084 1044 Source: Afrobarometer, 2005b, c, d, e

The second table (10.05) genuinely shows the exceptionally low salience that ethnicity plays in Tanzania. When respondents were asked to rate the “two groups they feel most

260 strongly attached to,” an overwhelming 79.6 percent of the Tanzanian respondents felt more attached to national identity over ethnic ones. By contrast, only 24.7 percent of the Kenyans,

23.9 percent of the Malawians, and 19.7 percent of the Zambians felt the same level of attachment to national identity. In all three of these cases, the most frequent response was to feel an equal attachment to ethnicity and national identity.

Table 10.05: Ethnic versus National Identity Kenya Malawi Zambia Tanzania Ethnic ID only 6.4% 18.9% 2.8% 2.9% Ethnic ID more than national 9.4% 4.0% 7.4% 2.9% Equal 44.4% 48.8% 60.4% 6.4% National ID more than ethnic 15.2% 4.4% 9.7% 8.2% National ID only 24.7% 23.9% 19.7% 79.6% n 1228 1077 1197 1221 Source: Afrobarometer, 2005 b, c, d, e

What the data thus far suggests is that Tanzanians are significantly less inclined to define their political and economic experiences through the lens of ethnicity and are far more inclined to feel a centripetal pull of national identity. In what possible ways does ethnicity affect political support?

Ethnicity and Political Contention

Among the politics of Tanzania’s northern neighbor Kenya, evoking ethnic sentiments is commonplace, and a politically viable road toward mobilizing a support base and slamming political opponents. Hence, campaigns quickly devolve into inter-ethnic discourses, sometimes erupting into physical confrontations, injuries, and death. For example, the Lou, without a traditional rite to passage, is commonly understood among Kikuyu as boys unfit to rule

(Nyambura-Mwaura, 2007). In the lead-up to the 2005 Constitutional referendum, Energy

Minister Simeon Nyachae publically and persistently evoked the lack of passage rites among the

261

Lou as a way of attacking Raila Odinga’s credibility as a public official and his opposition to the draft constitution. “Wale hawajatahiri peleka Jandoni” (those not circumcised should be taken for a circumcision ceremony), “Watu wengine hawajavuka daraja” (Some people have not crossed the bridge), and finally “I cannot work with [Raila] as he has not passed through the rites of adulthood (Kenya National Commission on Human Rights, 2006, pp. 30, 46).

At times, ethnic discourses risk inciting aggression and violence. Member of Parliament

Daudi Mwanzia exclaimed that “Wakisimamisha pension ya Moi, then Wakalenjin na Wakamba watavamia State House” (If they stop Moi’s pension, then the Kalenjin and Kamba communities will invade the State House) (Kenya National Commission on Human Rights, 2006, p. 30), while another MP, Alfred Nderitu, preached his view that “They hate Kikuyus because we are hard working. Luos just go fishing and fish is free and thereafter they ask the government for relief maize to make ugali” (Kenya National Commission on Human Rights, 2006, p. 39). These discourses were most recently visible in the aftermath of the 2007 election, where politicians and the local language media fanned ethnic conflict, which resulted in well over a thousand deaths (BBC World Trust Service, 2008).

In Tanzania, where references to ethnicity are evoked, it is often an attempt to paint a political opponent as a conduit for ethnic divisiveness and highlights the dangers of ethnic or

“tribal” politics. Hence, the discourses often look more like ethnicity versus nation rather than ethnic group A versus ethnic group B. Throughout the 1995 election, Nyerere occasionally stepped up to the podium to remind Tanzanians of need to preserve peace and stability, against the threat of sub-national violence similar to that found in neighboring countries (Nyerere’s speech wins, 1995). Again, these reminders carried through the 2000 elections, right up to the most recent elections in 2005. For Jakaya Kikwete, Tanzania’s current President, opposition

262 leaders are “all out to divide this nation along tribal, religious or racial lines” (Mwita, 2005). In a veiled description of the opposition threat, Kikwete commented at a speech at the Jangwani grounds that “people, who sow seeds of discord by preaching tribalism, racism and religious misunderstandings, they better find other countries to live. They have no room here” (Temba,

2005). Similarly, Mkapa stepped up to the plate to give additional substance Kikwete’s construction. In one case, Mkapa referred to the opposition by stating that:

Cote d’Ivoire, and the rest of us in Africa, had thought that this country had overcome the divide-and-rule legacy of colonialism. Little did we know that the specter of that legacy was still lurking under the surface, waiting for a moment to rear its ugly head and cause havoc, suffering, death and discord (as cited in Mkapa: we must, 2005).

A noteworthy example of the difficulties in winning Tanzanian elections based upon inter- ethnic self-other exchanges is found in Mrema’s successful parliamentary bid in the 1996

Temeke by-election, a constituency largely defined as Wadengereko and Muslim. For his part, the CCM’s candidate Abdul Cisco Omari tried repeatedly to frame the contest as one between the Wadengereko versus greater Wachagga dominance in political affairs1. The conclusion of the contest however, saw a Mchagga ‘outsider’ victorious in a constituency barely familiar with the Wachagga2 (National Electoral Commission, 1997; Sued, 2000).

The basis for the failures of the CCM in the Temeke by-election is supported by an opinion poll conducted by ESAURP, which showed that campaign issues that were likely to scare away votes were ones referencing race and religion (Masato, 2005). This opinion polls suggests that

Mrema’s indigenization platform, which, along with anti-corruption, formed a backbone of the

1 It might be recalled that Mrema is Mchagga from the Kilimanjaro region.

2 Indeed, Mrema’s victory in the Temeke by-election can only be described as a landslide. Amid a list of 13 candidates, Mrema scooped nearly 59 percent of the votes, followed by Omari’s 35.5 percent, and 3.6 percent for CUF’s Tambwe R. Hiza (National Election Commission, 1997). 263

1995 presidential contest, may have also had limited appeal, especially when the platform evokes images of racism and divisiveness.

This is not to suggest that opposition parties in Tanzania do not have ethnic roots. Indeed, some of them do. Following the multiparty reforms, party recruitment patterns throughout

Africa tended to depend on the ability of opposition party founders to operate on budgets far more limited than those obtained by incumbents. In many cases, leaders assumed the position of foremost patron of particular ethnic groups, often those from the same home region as the party leaders.

As Mmuya and Chaligha (1994) demonstrate, this is exactly what happened in Tanzania.

For the NCCR-Mageuzi, this meant recruiting from Dar es Salaam, Kagera, Mara, and Mwanza;

CHADEMA, Dar es Salaam, Arusha, Kilimanjaro, Mtwara, Mwanza, and Mbeya; and for CUF,

Kigoma, Lindi, Morogoro, Mtwara, Ruvuma, Singida, Tabora, Unguja, and Pemba. However, ethnicity, whether in terms of a single entity or an inter-ethnic coalition, was simply far too fragmented and lacking in saliency for building a party with viable nation-wide breadth similar to that of the CCM. As a consequence, parties are viewed by most not through the lens of “my ethnic group versus your ethnic group,” but rather as “our nation versus your ethnic group.”

The spatiality of CCM dominance indeed clarifies that opposition support in many cases, appeared unable to extend much beyond the narrow confines of the home regions of respective party leaders.

Depictions of this spatiality are provided in the appendixes C through H, which illustrate the regional breakdown of election results. The map posted in figure 10.01 below will help to geographically orient the following discussion. Votes cast for Augustine Mrema, despite his impressive performance during the 1995 elections, were disproportionately concentrated in the Wachagga region of Kilimanjaro, where he claimed victory in six out of the Region’s nine constituencies (see table 10.06). As the data presented in the aggregate statistics tables in appendix C show, while Mrema’s support was

264 lower than the two other opposition contenders (kurtosis=2.47 for percentage of votes cast), support for Mrema in Kilimanjaro during the 1995 election was nearly 50 percentage points above his level of national support. The data in the aggregate tables of appendix D indicates that support for

Mrema’s party, the NCCR-Mageuzi, was similarly concentrated (kurtosis=2.09 for percentage of votes cast). Kilimanjaro’s support for the NCCR-Mageuzi was over 32 percentage points above the party’s national support level in the concurrently held parliamentary elections. While the roots of the

NCCR-Mageuzi in Kilimanjaro are somewhat weak, Mrema’s roots here are strong. He was born in the

Hai constituency village of Kiraracha and is himself a Mchagga.

Figure 10.01: Regions of Tanzania

Source: Wikicommons, 2008

As an ethnic Msukuma, John Cheyo’s backing barely made it out of a handful of constituencies in Wasukuma regions of Shinyanga and Mwanza. For starters, the range (23.1%)

265 between his most committed region (Shinyanga) and least committed region (Kilimanjaro) is nearly 6 times greater than the total proportion of votes that the candidate collected nationally.

Furthermore, the candidate and the party’s support was the most concentrated out of any of the contending candidates or parties (kurtosis=11.92 for percentage of votes for each candidate and

6.99 for percentage of votes for each party).

Table 10.6: 1995 Presidential Election Constituency Results in Kilimanjaro Constituency CHEYO LIPUMBA MKAPA MREMA Moshi Urban 222 1,130 13,338 30,372 Moshi Rural 308 596 6,381 65,191 Vunjo 139 276 4,027 65,121 Mwanga 240 1,402 20,937 4,915 Rombo 223 304 5,036 66,376 Same West 127 408 9,963 9,769 Same East 162 320 11,349 7,314 Hai 175 1,946 6,968 47,073 Siha 160 259 4,942 19,160 Source: Tanzania Election Monitoring Committee, 1997

A closer inspection reveals that in Shinyanga, Cheyo completely destroyed his opposition in three of the districts, as shown in table 10.07, as well as polling relatively well in several nearby districts in Mwanza. His most notable performance was in the two constituencies of his home town Bariadi. In fact, both regions contributed around 67 percent of all of the nation-wide votes cast for the UDP candidate and 66 percent of all votes cast for the Party’s parliamentary candidates. Both regions registered a combined 14.8 percentage points above the national support level for Cheyo, and 12.8 percentage points above the support for the UDP in the

Parliamentary election. Only two other regions (Iringa and Dodoma) registered support levels higher than Cheyo’s national support level.

The spatial support for CHADEMA’s first presidential contender, Freeman Mbowe, during the 2005 presidential race, only confirms the strong spatial concentration of support for the

266 opposition. On the one hand, his range of support was relatively high, given the small percentage of votes he received nationally. The low kurtosis value of 2.5, as listed on the final page of appendix G, suggests that his distribution of support lacked the amplification of other opposition contenders, such as Mvungi and Mrema, along with Cheyo in the two previous elections. To be sure, a total of 8 out of 26 regions registered vote proportions above the total percentage of voted secured by the candidate, accounting for a combined level of support some

4.3 percentage points higher than the national support level, while constituting 69.8 percent of the votes cast for Mbowe.

Table 10.07: 1995 Presidential Election Constituency Results in Shinyanga Constituency CHEYO LIPUMBA MKAPA MREMA Shinyanga Urban 811 935 21,335 9,771 Solwa 1,156 835 19,531 3,276 Kishapu 2,886 1,713 31,204 6,269 Bariadi East 25,663 2,210 12,457 974 Bariadi West 28,227 2,075 19,155 1,645 Meatu 4,353 824 10,075 3,592 Kisesa 15,537 1,436 6,502 800 Kahama 1,508 2,193 25,247 26,187 Bukombe 2,790 2,180 24,258 15,721 Msalala 2,956 2,611 29,072 11,547 Maswa 15,379 2,265 25,630 5,167 Source: Tanzania Election Monitoring Committee, 1997

On the other hand, the hotbed of Mbowe’s support is largely confined to his home region of

Kilimanjaro, where he scored 20 percent of the votes cast, while CHADEMA’s parliamentary contests polled nearly 22 percent of the regional votes. Not surprisingly, since many of the

Party’s earlier founders were Mchagga from Kilimanjaro, the Region proved to be

CHADEMA’s biggest support base in earlier elections. In 1995, the Party’s parliamentary contenders scored 18 percent of the votes cast, while polling 23 percent in the 2000 parliamentary elections.

267

Finally, Ibrahim Lipumba’s three bids for the presidency are somewhat more unique. In the

2000 Presidential Election, Lipumba emerges as the formidable opposition candidate, securing a breadth (although not volume) of support that resembles Mrema’s 1995 bid. A total of 10 out of

25 regions (Coast, Dar es Salaam, Kagera, Kigoma, Morogoro, Tabora, Tanga, Pemba North,

Pemba South, and Unguja North), scattered throughout Zanzibar and the eastern portions of the

Mainland, along with a section of the northwest, registered levels of support above the total percentage of votes cast for Lipumba. The candidate’s home region is Tabora, which, together with Dar es Salaam, account for his strongest support bases on the Mainland. While eclipsed by

Mrema in 1995, Lipumba also scored big in Tabora during the 2005 election, as did CUF.

What about the associations between ethnicity and political orientation in comparative perspective? To answer this question, I will turn to some of the Afrobarometer data sets previously discussed. Data from Afrobarometer for Tanzania indicates that questions related to ethnic identity and party preference are either statistically insignificant or significant, but not robust. For example, respondents who were less inclined to see their own ethnic group as politically influential were only slightly more inclined to be close to a political party (Pearson

Correlation at .079). Similarly, respondents who perceived higher levels of ethnic unfairness also had small tendencies to place greater trust in opposition (pearson correlation=0.066)3. The most significant and strongest correlation was found where respondents which reported greater levels of trust in people from other ethnic groups also tended to place greater levels of trust in the CCM (PC=0.117).

On a comparative basis, the associations between ethnic identity and political affiliation in

Zambia and Malawi appear not much different than Tanzania. For Zambia, similar comparisons

3 From here on out I will denote pearson correlation as “PC”. 268 using Afrobarometer data produced similarly insignificant and weak associations. Respondents who believed their ethnic lot to be relatively worse were less likely to trust the ruling party

(PC=-0.082). Those who believed that their ethnic group has been treated unfairly by the government were more inclined to trust the opposition (PC=0.079). Respondents who believed that their ethnic group has been treated unfairly by the government were less likely to trust the ruling party (PC=-0.063).

For Malawi, marginally stronger associations were obtained. There was a weak inverse association between respondents who believed their own ethnic group had relatively more political influence and the tendency to feel close to a political party (PC=-0.105). Those who believed that their ethnic group has been treated unfairly by the government were less likely to trust the ruling party (PC=-0.112).

Not surprisingly, the results from responses to the same series of questions from the Kenya survey produced considerably more significant and robust associations than found in any other cases, thus supporting the often cited claim that ethnicity is a primary basis for political mobilization in Kenya. For example, respondents which perceived the political influence of their ethnic group as weak compared to other groups tended to place less trust in the ruling party

(PC=-0.124). Correspondingly, respondents with greater feelings of unfairness based upon ethnicity, tended to place greater trust in opposition parties (PC=0.102), and substantially less trust in the ruling party (PC=-0.235). Respondents who express a lack of trust of people from other ethnic groups are also less inclined to trust the ruling party (PC=0.183), while no significant association could be found in their inclination toward trusting the opposition.

Similarly, respondents who expressed a greater tendency to trust people from their own ethnic

269 group also expressed a greater tendency to trust the ruling party (PC=0.264), while no significant association could be found in their inclination toward trusting the opposition.

Table 10.08: Language according to Political Affiliation CCM CUF CHAD NCCR TLP UDP Language Total Kiswahili 17.1 % 46.0 % 6.7 % 0.0 % 18.2 % 0.0 % 18.2 % Kinyakyusa 3.1 % 0.0 % 0.0 % 0.0 % 0.0 % 0.0 % 2.9 % Kichaga 3.7 % 0.0 % 0.0 % 0.0 % 36.4 % 0.0 % 3.8 % Kihaya 3.9 % 0.0 % 0.0 % 0.0 % 0.0 % 0.0 % 3.6 % Kisukuma 15.5 % 8.0 % 13.3 % 0.0 % 9.1 % 50.0 % 15.2 % Kifipa 3.5 % 0.0 % 0.0 % 0.0 % 9.1 % 0.0 % 3.4 % Kiluguru 2.7 % 4.0 % 0.0 % 0.0 % 9.1 % 0.0 % 2.8 % Kingoni 1.5 % 2.0 % 0.0 % 0.0 % 0.0 % 0.0 % 1.5 % Kihehe 2.7 % 0.0 % 0.0 % 0.0 % 0.0 % 0.0 % 2.5 % Other 46.3 % 40.0 % 80.0 % 100.0 % 18.2 % 50.0 % 46.3 % n 1082 50 15 1 11 2 1161 Source: Afrobarometer, 2005d

Turning back to Tanzania, I decided to test the association between language and party affiliation. Since language contains far more categories than questions about fair treatment or party closeness, I decided to abandon the Pearson technique and instead rely solely on simple cross tabulations. When first breaking the Afrobarometer data down according to language and then party preference, stronger ethnic clusters of party support within the opposition are revealed. As posted in table 10.08 above, probability of selecting a TLP supporter who is also a

Kichagga speaker is 36.4 percent, while the probability of picking a Kichagga speaker out of the entire sample is only 3.8 percent. Similarly, the likelihood of selecting a UDP supporter whose primary language is Kisukuma is 50 percent, while the likelihood of selecting a Kisukuma speaker out of the entire sample is 15.2 percent. Kiswahili speakers are highly overrepresented in CUF at 46 percent compared to 18.2 percent for the entire sample. What does emerge from the table, is that the language profile of the CCM members almost exactly mirrors the language

270 profile of the entire sample, suggesting that the CCM has no clear basis among any particular ethnic group.

Since the Pearson coefficient for language versus ethnicity was only 50 percent (suggesting that the two are only loosely affiliated), I decided to cross tabulate party vote with ethnic identity in order to find more robust associations. However, the results were similar to the results for language. According to the cross tabulated data, some 45.5 percent of the TLP respondents were also Mchagga, while 50 percent of the UDP respondents and 13.3 percent of the CHADEMA respondents were Msukuma. As for CUF, a combined 20 percent were either

Mnyamwezi or Msukuma. With the exception of CHADEMA’s support among the Wasukuma, these support patterns far exceed the ethnic demography of the Afrobarometer sample. Again, the distribution of ethnic support for the CCM almost perfectly matched the ethnic distribution for the sample.

Table 10.09: Political Affiliation according to Language CCM CUF CHAD NCCR‐M TLP UDP n Kiswahili 87.7 % 10.9 % 0.5 % 0.0 % 0.9 % 0.0 % 211 Kinyakyusa 100.0 % 0.0 % 0.0 % 0.0 % 0.0 % 0.0 % 34 Kichaga 90.9 % 0.0 % 0.0 % 0.0 % 9.1 % 0.0 % 44 Kihaya 100.0 % 0.0 % 0.0 % 0.0 % 0.0 % 0.0 % 42 Kisukuma 95.5 % 2.3 % 1.1 % 0.0 % 0.6 % 0.6 % 176 Kifipa 97.4 % 0.0 % 0.0 % 0.0 % 2.6 % 0.0 % 39 Kiluguru 90.6 % 6.3 % 0.0 % 0.0 % 3.1 % 0.0 % 32 Kingoni 94.1 % 5.9 % 0.0 % 0.0 % 0.0 % 0.0 % 17 Kihehe 100.0 % 0.0 % 0.0 % 0.0 % 0.0 % 0.0 % 29 Other 93.3 % 3.7 % 2.2 % 0.2 % 0.4 % 0.2 % 537 Total member of 93.2 % 4.3 % 1.3 % 0.1 % 0.9 % 0.2 % 1161 Source: Afrobarometer, 2005d

When cross tabulating the data to find the party vote profile as a proportion of the linguistic or ethnic sample, we see that nearly all major ethnic and linguistic groups through their weight behind the CCM. As evident from table 10.9 above, there is simply no party that can claim even

271 a modicum of a monopoly over a particular ethnic or linguistic group. The closet one might find is the case of the TLP, where, out of the 44 Kichaga speakers in the sample, on 9.1 percent were members of the TLP.

By contrast, in Kenya, none of the parties can boast membership profiles which close match the sample’s ethnic demography. As indicated below in table 10.10, the largest vote getting party (NARC) clearly enjoys disproportionately large support from the old GEMA alliance between Kikuyu, Kamba, and Meru, which previously fed the tenure of Jomo Kenyatta. When taken together, the linguistic composition of the earlier GEMA alliance constitutes 54.5 percent of the support for NARC, compared to their combined 38.7 of the entire sample. At the same time, the Afrobarometer data also confirms NARC’s failure to win large followings within the

Lou and Kalenjin communities, which historically opposed the GEMA but some of which constituted the most reliable Moi constituents. When taken together, KANU’s support from the smaller groups (Kalenjin, Somali, Other) constituted some 71.1 percent of the overall party support, compared to their combined 40.2 percent of the entire sample.

Table 10.10: Language according to Political Affiliation (Kenya) NARC LDP DP Ford‐K KANU NPK Shiri Other Language Total Kikuyu 28.3 % 1.5 % 68.2 % 0.0 % 14.2 % 40.0 % 0.0 % 14.3 % 20.0 % Luo 4.1 % 43.3 % 0.0 % 11.1 % 2.5 % 0.0 % 0.0 % 0.0 % 11.7 % Luhya 9.6 % 10.0 % 0.0 % 55.6 % 5.4 % 0.0 % 0.0 % 14.3 % 9.4 % Kamba 12.3 % 11.9 % 9.1 % 0.0 % 5.4 % 20.0 % 0.0 % 7.1 % 10.4 % Meru/Emb 13.9 % 0.5 % 22.7 % 5.6 % 1.5 % 0.0 % 0.0 % 0.0 % 8.3 % Kalenjin 4.3 % 11.9 % 0.0 % 5.6 % 27.5 % 0.0 % 0.0 % 0.0 % 10.5 % Somali 2.5 % 4.0 % 0.0 % 0.0 % 11.3 % 20.0 % 0.0 % 0.0 % 4.6 % Other 25.0 % 16.9 % 0.0 % 22.2 % 32.4 % 20.0 % 100.0 % 64.3 % 25.1 % n 512 201 22 18 204 5 4 14 980 Source: Afrobarometer, 2005b

When examining the proportions of ethnic support for particular political parties, we find that, NARC indeed can claim a monopoly over the communities of the old GEMA. Some 74.0

272 percent of the Kikuyu, 87.2 percent of the Meru, and 63.1 percent of the Kamba backed NARC, compared to 52.2 percent NARC support for the sample. The LDP captured 78.6 percent of the

Lou backing, while the party garnered only 20.5 percent of the sample support. Finally, KANU continues to have strong backing among earlier Moi loyalists. The Party captures 55.9 percent of the Kalenjin and 51.1 percent of the Somali support, but only 20.8 percent of the sample support.

Moving on to Malawi, the historical patterns of developmental asymmetries helped consolidate highly fragmented and localized identities into powerful regional voting patterns. In post-independence times, the Chewa, while hailing from the Central Region, was promoted by the MCP and Banda as the ‘official’ culture of Malawi. Chewa emerges as the only national language, while others were banned.

Given this pattern of consolidation and exclusion, one would expect ethnicity to play a major mobilizational role in Malawian politics today. Indeed, recent Afrobarometer data appears to support this claim. As specified in table 10.11, the likelihood of finding a DPP member who is also a Tumbuka speaker is 15.9 percent, while the Tumbuka tongue represents slightly less than 11 percent of the Afrobarometer sample. At the same time, the probability of finding a

DPP supporter who is also a Yao speaker is 5.1 percent, while Yoa make up 12.3 percent of the sample. The respondents which claimed to support the MCP were 85.1 percent Chewa, while the Chewa represented 56.8 percent of the sample. Similarly, 30.9 percent of those which backed the UDF were Yao, which represented only 12.3 percent of the sample.

When examining the proportion of ethnic support for particular political parties, 82.6 percent of the Sena, 64.7 percent of the Tumbuka, and 57.4 percent of the Lomwe respondents are found to favor the DPP, although the party was backed by 44.2 percent of the entire sample. While

273

78.3 percent of the Yao voted for the UDF, the party was backed by no more than 31 percent of the sample. While the plurality of the Chewa backed the DPP (41.5 percent), a disproportionate number also backed the MCP. Whereas 31.1 percent of the Chewa claimed to support the MCP, the party garnered slightly less than 21 percent of the sample support.

Table 10.11: Language according to Political Affiliation (Malawi) AFORD DPP MCP UDF Other Language Total Chewa 0.0 % 53.3 % 8.1 % 47.4 % 34.8 % 57.0 % Yao 0.0 % 5.1 % 2.1 % 30.9 % 0.0 % 12.3 % Tumbuka 100.0 % 15.9 % 5.6 % 2.7 % 26.1 % 10.9 % Ngoni 0.0 % 3.9 % 4.1 % 3.8 % 4.3 % 3.9 % Lomwe 0.0 % 8.4 % 1.0 % 7.2 % 13.0 % 6.5 % Manga'nja 0.0 % 3.1 % 1.5 % 5.8 % 4.3 % 3.6 % Sena 0.0 % 4.6 % 0.5 % 0.7 % 4.3 % 2.5 % Other 0.0 % 5.8 % 0.0 % 1.4 % 13.0 % 3.3 % n 11 415 195 291 23 935 Source: Afrobarometer, 2005c

In Zambia, popular perceptions, along with representation in the media, often defines the character of party clientele networks according to the ethnicity of party leaders. In the Third

Republic, UNIP’s base was repeatedly defined as the Ngoni, Nsenga, and Tumbuka in the

Eastern Province, while the UPND was unambiguously a party of the Tonga in the south. The

MMD, which upset UNIP’s tenure in the 1990 election, assumed the political leadership of the

Bemba, the country’s most numerous ethnic group, while the NP was perceived as a Lozi and

Tonga party (Pozner, 2005, p. 108).

Afrobarometer data unequivocally demonstrates that language in Zambia is a fundamental force for mobilizing political support. As posted in table 10.12, the likelihood of selecting a

MMD member who is also a Nyanja speaker is 12.1 percent, compared to the 8.6 percent chance of selecting a Nyanja speaker out of the entire sample. A similar probability is found among the

Nsenga in relation to the MMD. While the Bemba is strongly represented in the MMD, their

274 representation in the Patriotic Front (PF) and the “other” category is disproportionately high.

The probability of selecting a Bemba speaker out from the distribution of PF supporters is 58.3 percent, compared to the 31.1 percent chance for making the same pick out of the entire sample.

As for the smaller parties, the Forum for Democracy and Development (FDD) and Heritage

Party (HP) membership profiles are clearly over represented among the Chewa, Nyanja, and

“other” category.

The ethnic characteristics become front and center when examining the proportion of ethnic support for particular parties. While supported by 38.7 percent of the respondents, the MMD was backed by 65.3 percent of the Nsenga and 54.3 percent of the Nyanja respondents. The

United Party for National Development (UPND), while backed by 30.6 percent of the respondents, found overwhelming support from the Tonga, at 81.1 percent and the Lozi, at 51.2 percent. The PF was backed by 36.1 percent of the Bemba respondents, while being backed by no more than 19.2 percent of the entire sample.

Table 10.12: Language according to Political Affiliation (Zambia) FDD HP MMD PF UNIP UPND Other Language Total Bemba 33.3 % 25.0 % 32.8 % 58.3 % 21.2 % 11.3 % 60.0 % 31.1 % Nyanja 12.1 % 16.7 % 12.1 % 7.1 % 18.2 % 2.8 % 13.3 % 8.6 % Tonga 9.1 % 8.3 % 5.7 % 0.6 % 0.0 % 39.9 % 0.0 % 15.0 % Lozi 3.0 % 0.0 % 9.2 % 3.2 % 12.1 % 16.9 % 6.7 % 10.1 % Chewa 21.2 % 16.7 % 8.6 % 7.7 % 21.2 % 0.8 % 0.0 % 7.0 % Nsenga 6.1 % 0.0 % 10.2 % 4.5 % 12.1 % 1.6 % 0.0 % 6.0 % Luvale 0.0 % 0.0 % 5.1 % 2.6 % 3.0 % 6.9 % 0.0 % 4.7 % Other 15.2 % 33.3 % 16.2 % 16.0 % 12.1 % 19.8 % 20.0 % 17.4 % n 33 12 314 156 33 248 15 811 Source: Afrobarometer, 2005e

275

Conclusion

Given the historical efforts by TANU to undermine the tribal association and institutions of customary rule, the regime’s overall penchant at avoiding ethnic allusions in national politics, and commitments to ensuring a weak concurrence between ethnicity and class, it is little wonder that and the wealth of data presented in this chapter suggests that ethnic saliency in multiparty

Tanzania lacks the degree of polarization found in neighboring countries. Survey data found only modest tendencies to define differences in economic status, political influence, and fair treatment in ethnic terms. When compared to national identity, the weak saliency of ethnic identity in Tanzania is in stark contrast to the tendencies in Malawi, Kenya, and Zambia to place both national and ethnic identities on equal footing.

It is also clear that ethnic divergence, which offered political elites in neighboring countries almost readymade formulas for launching powerful political challenges to incumbent regimes, offer little in the way of opportunities for mobilizing political support in Tanzania. In Tanzania, references to ethnicity in political campaigns can actually damage one’s chances of winning an election. When ethnicity is referenced, it is less commonly about ethnic group A versus ethnic group B, as found in the other cases, but rather ethnic group versus the nation. In comparative terms, the relationship between ethnicity and language versus political loyalties are significantly weaker in Tanzania as demonstrated in the final series of tables above. In short, whereas the ability to tap ethnic structures in Malawi, Zambia, and especially Kenya are casually connected to regime change, the absence of these structures in Tanzania undermined the prospects of unseating the CCM.

At the same time, it is safe to suggest that, had development patterns in Tanzania resembled those in Kenya, Malawi, or Zambia, multiparty politics in Tanzania would be a significantly

276 different affair altogether. While opposition parties may have been divided by a lack of inter- ethnic trust or personal political ambitions, the scope of the CCM’s election victories would have been significantly less than they are today, as political entrepreneurs would undoubtedly tap into the saliency of ethnic discourses as a basis for mobilizing support. Under such a scenario, the most viable tactic of the CCM would be more directly tied to keeping the opposition divided, while cementing its own ethnic coalition.

277

CHAPTER ELEVEN RELIGION AS A BASIS FOR CONTENTION

“I distrust those people who know so well what God wants them to do because I notice it always coincides with their own desires.”

Susan B. Anthony, 1896

“Those who are trying to make [education] a political issue are an insignificant handful and, thank God, the vast majority of our people are not misled by the hypocrisy of the few who try to cloak political self-seeking under the mantle of pseudo-religious fervor”.

Julius K. Nyerere, 1963

After some 250 pages of text, figures, and charts, we have seen the qualitative and quantitative degrees to which policies pursued under African Socialism, with its emphasis on state economic leadership, political and economic equality, and the constitution of a cohesive sense of nationhood has reproduced a macro-structural environment with comparatively low degrees of macro-structural divergence. The Tanzanian case emerges in stark contrast to what one finds among many other African cases where post reform politics has been characterized by blatant human rights violations or where long-time incumbent parties found themselves out of office following multiparty elections. Chapter eight clearly demonstrated that, thanks to the eradication of rural economic and political power structures following the 1967 Arusha

Declaration and the nationalization of major means of production, the economic basis for building sustainable opposition challenges to the CCM at the onset of multiparty reform were minimal at best. Chapters nine and ten demonstrated that single-party policies in Tanzania directed toward redressing regional development differentials and developing a shared sense of national identity went a long way in reducing the polarization that served as a mobilizational basis for opposition parties in neighboring countries. Hence, in a multiparty environment, the CCM could reproduce its dominance over the legislative and executive functions of the state without resorting to coercive tactics that do not interface well with an international climate stressing the importance of human rights and good governance.

The discussion so far, has managed to steer clear of a full blown exposition of religious fissures as a basis for political affiliation. The goal of this chapter is to turn the helm headlong into an analytic discussion about the relationship between religion and political affiliation. As will be shown in the following pages, some of the previously discussed patterns apply to religious divisions as well, primarily the ability of national identity to stitch together a shared identity where discourses of difference otherwise exist. However, while religion has been relatively subdued during the era of single-party rule, new faith-based discourses of difference are becoming increasingly polarized. This polarization in turn, constitutes a possible threat to the future of CCM dominance.

A Survey of Region in Tanzania

Considering the religious demographics of Tanzania, paired with a history of economic inequalities between Christians and Muslims since the Colonial times, it is striking that religion does not appear to play a big role in shaping political divisions. Approximately 40 percent of the

Country’s population is Christian, of which about 26 percent are Catholics and 14 percent are

Protestants. About 35 percent of the population professes to be Muslim, of which 33 percent are

Sunni, while the remaining 2 percent are Shia.

Due to the expansion of missions and a colonial policy which emphasized proselytization, the growth of Christian funded schools during colonial rule exploded (Mbogoni, 2004). Data provided by the Ministry of Education shows that the number of pupils enrolled in government

279 primary schools in 1956 stood at 10,254 and another 84,062 students were enrolled in primary schools administered under the Native Authorities. During the same year, the number of students enrolled at mission assisted primary schools stood at 235,516 (as cited in Buchert,

1994, pp. 64-5). At the same time, the number of Muslim assisted schools stood at 28, compared to the 1,692 schools funded by the government and Native Authorities (as cited in Mbogoni,

2004, p. 107). Given these vastly different education opportunities, it is no surprise to find a deficit of Muslim representation in civil service positions after independence.

Indeed, educational inequalities which cut along religious lines proved to be a major sticking point in politics at the time of independence. The All Muslim National Union of

Tanganyika (AMNUT), which broke with the TANU Elders Council in 1959, announced their desire to see the British stay on until Muslims gained educational parity with Christians (Njozi,

2003), a position that was widely denounced by Nyerere. Furthermore, despite TANU’s stated commitment to the separation of religion and politics, the state and the Party looked exceedingly tied to the interests of Christians. Nyerere himself was a Catholic and was educated and inspired with the help of a number of Catholic figures and organizations. A Pope delegate was sent to observe the independence celebration. Cardinal Rugambwe gave the Catholic prayer at the Flag

Raising ceremony at the national stadium in 1961 (Mbogoni, 2004). For many prominent

Muslims at the time of independence, it appeared that the absence of the Union Jack meant independence from British political authorities, but not independence from, or parity with

Christian authorities.

However, there were two factors which prevented Muslim grievances from taking on a major national agenda. First was the inability of Muslim elites to tap into lucrative financial resources for building the type of peak organizations which could compete with the Catholic

280

Tanzania Epsicopal Conference (TEC) or the Protestant Christian Council of Tanzania (CCT).

Secondly, those peak associations that did exist were created under the auspices of TANU, most notably the Baraza Kuu La Waislam Tanzania (BAKWATA). The only non-party peak organization that did exist, the East African Muslim Welfare Society (EAMWS), which was chaired by a staunch opponent of TANU named Abdallah Fundikira, was dismantled by the state in 1968 after a number of high level EAMWS members took up positions in the TANU

Central Committee and National Executive Committee and proceeded to form BAKWATA. In short, Muslim leaders were effectively neutralized through a series of moves which accommodated the more mainstream leaders, while sidelining those seen as outside the status quo (Lodhi & Westerlund, 1997).

Inter-faith tensions have again grown more acute, especially since the 1980s, as more fundamentalist hues among Christian and Muslim identities gained ground in public discourses.

Born again preachers and the revivalist movement in 1970s made inroads throughout East

African urban and rural society (van Dijk, 1992) and, politically speaking, evolved into “a force to be reckon with” in the national political arena (Mlahagwa, 1999, p. 299). In Tanzania in particular, the Assemblies of God expanded quickly throughout village society, filling the religious vacuums created by the Ujamaa and Villagization upheavals. At the same time, the

Pentecostal emphasis on faith as “individualistic and exclusive” was a sharp contrast to the ideals of “unity, stability, and harmony” of the Catholic and Protestant denominations, most of which engaged cooperatively with the party-state nexus and mainstream Muslims communities following independence (Ludwig, 1996, p. 221).

In response to the increased militancy of Christian impulses, public open-air preaching, or mihadhara, spread as Muslims of more militant persuasions attempted to construct a similar

281

Islamic revivalism. As the Organization of the Islamic Conference (OIC) grew more active in

Tanzania, new Muslim organizations with sweeping religious and political ambitions, claimed to be the true vanguard of Islam and took on an active effort at discrediting Christianity

(Mbogoni, 2004). Reflecting on these trends, Mbogoni (2004) notes, as revivalist preachers were out to spread and intensify faith in Christ, mihadhara preachers were “out to debunk the theological basis of this revivalism.” (p. 173)

Economic decline and increased cost-accounting measures which accompanied liberalization had the consequence of increasing the need to rely on faith-based schools and health clinics, reducing the ability of the state to exercise political control over educational curriculum (Mbogoni, 2004). At the same time, other Muslim organizations were formed to advance social agendas and spread knowledge of the Koran. Warsha ya Waandishi wa Kiislam

(Islamic Writers’ Workshop), which was founded in 1975 as an appendage of BAKWATA, directed its attention toward promoting Islamic education alongside secular teachings. Baraza la

Uendelazaji Koran Tanzania (Tanzania Quranic Council) (BALUKTA) was founded in 1987 in order to establish and manage Islamic institutions of higher learning and to assist in spreading an overall understanding of Islam.

In contrast to the Tanzania’s mainstream Muslims and Christian organizations, at the onset of multiparty reform larger numbers of religious militants talked more openly in opposition to the religious status quo. According to Njozi (2003), the 1992 Memorandum of Understanding between the CCT, the TEC, and the government, only heightened the perception among some

Muslims of a close relation between Christianity and the Tanzanian State (p. 19). At the same time, while attempts by the CCM to discipline outspoken Muslim officials, such as

282

BAKWATA’s Sheikh Mohamed Ali Al-Bukhri1 and the former Minister of Education Professor

Kighoma Ali Malima2, acted to neutralize religious tensions within state and party corridors, it also instituted a sense of mistrust in what were seen by some as “government Sheikhs”, and displaced conflict away from the public sector to the private one (Njozi, 2003).

In addition to this intra-Muslim demarcation, the tightness between CCT, TEC, and party- state drove a wedge between militant Revivalism and mainstream Christians. This wedge was sharpened by the less than lukewarm support that the dominant Christian establishment gave to multiparty reform and the reluctance of Pastors to give their blessings to opposition politicians during the first multiparty election. Thus, a sort of “common ground” was solidified between

Pentecostalism and opposition parties, giving rise to the likes of the somewhat animated and fiery leader of the Democratic Party (DP), Reverend Christopher Mtikila (Ludwig, 1996, p.

225).

If however, we assume that religious tensions are aggravated by more objective conditions of political and economic exclusion, depravity, and unfairness, then Tanzania might be a less likely place for the spread of inter-faith tensions throughout mainstream society. As some of the data below suggest, religious fissures along economic and political lines remain quite minimal

1 During the early 1980s, under the leadership of Sheikh Mohamed Ali Al-Bukhri, BAKWATA grew more critical of what they called the hegemony of the “Christian lobby” in national politics. The organization increasingly attacked what it saw as enormous educational inequalities between Christians and Muslims and called for a separate Islamic court system for handling marriage and divorce. The government responded by accusing BAKWATA’s leaders of inciting hostiles between Christians and Muslims and threatening the unity of the nation. The government took action against BAKWATA by expelling Sheikh Mohamed from the organization (Lodhi & Westerlund, 1997).

2 In 1987, President Mwinyi appointed Prof. Kighoma Ali Malima to the position of Minister of Education, the first Muslim to occupy that office since independence in 1961. He was eventually forced out due to his charges against the government’s religious bias in education and calls for attention to the plights that Muslims faced in education throughout the country.

283 in Tanzania, and thus probably have low prospects for translating into serious political party fissures.

Saliency of Religion

Some of the data presented in this section suffers from similar problems as the data presented in other chapters: Most of the samples sizes for specific categories are far too small to draw confident conclusions about associations between religion and politics. In order to compensate for small sample sizes, I try to be a bit creative by citing a wide selection of studies which directly and indirectly speak to questions about associations between religious views and political views. At the end of the chapter, I will provide a qualitative summary of the surveys discussed below.

Assuming that greater levels of political awareness and tendencies to discuss politics reflect lower degrees of political apathy and alienation, a REDET (2001) survey suggests that levels of political awareness and discussion, and thus alienation and apathy, between Christians and

Muslims are not so dissimilar. For starters, 35 percent of the Christians sampled in this survey were unable to provide the name of any government ministry, compared to 40 percent of the

Muslims. Similarly, when respondents were asked how many members of the Parliament they could name, 33 percent of the Muslim respondents could not name any MPs while 10 percent could name at least five, compared to 26 percent and 14 percent of the Christians respectively.

As for knowledge of political parties, REDET found that 21 percent of the Christians and 24 percent of the Muslims were unable to name any political party and its respective chairperson.

Furthermore, the similarities between Muslims and Christians in awareness of local politics look a lot like the data above. In both cases, roughly half were unable to name a local councilor.

284

Survey data on the prevalence of discussing politics produces slightly more divergent results. When discounting for uncommitted responses, 37 percent of the Christians reportedly

“never” talked about politics, whereas 46 percent of the Muslim respondents reported the same tendency. As indicated in table 11.01, this difference largely reflects differing tendencies to

“occasionally” discuss, where 41 percent of the Christians versus 34 percent for Muslims, reported occasionally participating in political discussions. Differences in “once a week” and

“daily” discussions are insignificant.

Table 11.01: Discuss Politics According to Religion None Traditional Christian Muslim Every Day 8,6 % 12,5 % 12,1 % 11,3 % Once a Week 8,0 % 4,2 % 9,6 % 9,1 % Occasionally 37,7 % 29,2 % 41,0 % 33,6 % Never Talks 45,7 % 54,2 % 37,3 % 45,9 % n 175 24 2594 1641 Source: REDET, 2001

Other indicators from the same survey show trends similar to those above. Between

Christians and Muslims, there are almost identical expectations of equal treatment by police and perceptions of the benefits and harm of local and national government. Likewise, responses were almost identical when asked about the degree to which they felt that government officials listened to their complaints. In nearly identical percentages (38 for Christians v. 35 for Muslims) respondents from both groups felt that government officials usually listen and take their concerns into account. The widest divergence was found among those that felt ignored: 19 percent of the Muslims and 14 percent of the Christians.

As posted in table 11.02, the same REDET (2001) survey found correspondingly dissimilar levels of education between Christian and Muslim respondents. The single biggest gap was among those with primary education, where 65 percent of the Christians reported primary school as the highest level of educational attainment versus 53 percent for Muslims.

285

At the same time, there is one particularly noteworthy discovery in the REDET data. Nearly

21 percent of the Muslim respondents reported “other” or “didn’t attend” school, compared to

15 percent of the Christian respondents. Hence, the REDET data does indicate that the educational asymmetries, which are so widely talked about among Muslims, are indeed real.

Table 11.02: Education according to Religion Madrassa Adult Primary Secondary Other n None 0.0 % 11.6 % 39.2 % 1.0 % 48.2 % 199 Traditional 0.0 % 12.1 % 30.3 % 3.0 % 54.5 % 33 Christian 0.1 % 4.5 % 64.6 % 16.3 % 14.5 % 2836 Mulsim 4.0 % 4.9 % 53.2 % 17.1 % 20.9 % 1763 Other/DK 0.0 % 0.0 % 37.5 % 0.0 % 62.5 % 8 Total 1.5 % 5.0 % 59.1 % 15.8 % 18.6 % 4839 Source: REDET, 2001, p. 242

Another telling finding from the REDET survey was mentioned in the previous chapter.

When respondents were asked how they would feel if their son or daughter married a person from another religious group, the plurality of respondents expressed displeasure. In fact, out of the four categories presented to respondents (religion, political party, race, and ethnic group), marriages across religious lines seemed to be the most challenging for respondents to stomach.

Some 43.7 percent of the respondents expressed being displeased at the thought of their sons marrying women with different religious backgrounds, compared to the 30.3 percent for political parties, 30.3 percent for races, and 18.7 percent for ethnic groups. The numbers were almost identical for feelings about the prospects of their daughters marrying men from a different religions, parties, races, or ethnic groups.

These feelings of displeasure are even more pronounced on the Isles. As indicated in table

11.03, some 95.3 percent of the Pemba respondents expressed displeasure at the thought of their daughters marrying men from other religious groups, compared to the 81.8 percent for the

Unguja cohorts and 38.5 percent from the Mainland sample. Indeed, it is quite clear that, in

286 relative terms, religion is the social division with the most potential for yielding political divisions, especially in relation to Zanzibar.

Table 11.03: Reaction to Daughter Marrying a Man from another Religion Pleased Displeased Indifferent Other Mainland 15,5 % 38,5 % 38,1 % 7,8 % Unguja 3,3 % 81,8 % 7,3 % 7,6 % Pemba 2,7 % 95,3 % 2,0 % 0,0 % Source: REDET, 2001

Afrobarometer data appears to confirm the REDET findings which point to differences in education across religious groups. When compared to Catholics and Protestants, both Shiite and

Sunni Muslim respondents were substantially more likely to have no formal schooling. All-in- all, 15.1 percent of the Muslim respondents (n=358) reported having no formal education, compared to 7.9 percent of the Christians (n=799). Similarly, 78.3 percent of the Christian respondents reported having attended primary school, compared to 63.7 percent for the

Muslims3.

However, this inequality evaporates of even reverses at higher levels of education. The probability of selecting a Muslim respondent reporting secondary education as the highest level attained is 19.3 percent, compared to 12.4 percent for Christian respondents. The probability of a respondent reporting post-secondary education was almost identical for both groups.

These differences appear even less significant when compared to two other neighboring countries. In Kenya, while 7.3 percent of the Protestant and Catholic respondents (n=926) lacked formal schooling, 34.2 percent of the Sunni and Shiite respondents (n=152) reported having no formal education. At the same time, 36.6 percent of the Christian respondents had some secondary schooling, compared to 19.7 percent of the Muslims cohorts. In terms of post- secondary education, 15.8 percent of the Christians reported education beyond secondary

3 This category includes those that have “some primary school” and those which “completed” primary school. 287 school, versus 8.6 percent of the Muslims. Only at the level of primary education do we see similarities between Christians and Muslims.

Educational differences between Christians and Muslims are even more significant in

Malawi. The probability of a Catholic or a Protestant (n=823) reporting no formal education was 14.7 percent, compared to 36.9 percent for Sunnis or Shiites (n=157), while the probability of a Catholic or a Protestant reporting primary education was 61.1 percent, compared to 51.6 percent for Sunnis or Shiites. Similarly, 22.6 percent of the Protestants and Catholics reported secondary education, versus 11.5 percent of the Sunnis and Shiites. Not one Muslim reported having a post-secondary education, versus 1.6 percent of the Christians.

All-in-all, the REDET and Afrobarometer data suggest that religious differences in

Tanzania do indeed correspond to difference in other socio-economic indicators. While

Christians and Muslims both express similar levels of interest in political affairs, both religious categories are highly skeptical toward the prospects of inter-faith marriage. This indicates a high level of mistrust between Christians and Muslims generally. Additionally, there are serious inequalities in education between Christians and Muslims, especially at the primary school level. This lack of trust, along with differentials in education can quite easily translate into a potential source for ambitious political entrepreneurs to mobilize discontent. Let’s look more specifically at the current relationship between religious affiliation and political orientation.

Religion and Political Contention

At the onset of new party formations in Tanzania, religion appeared to be an important site for shaping political conflict, at least on the margins. Reverend Mtikila emerged as the fiery spokesperson for the political aspirations of those Christians who feared social and moral decay,

288 corruption, and Muslim domination of society. As a consequence, Mtikila, and his small but lively party, has taken the brunt of state censorship efforts. Despite repeated attempts since the opening years of multiparty reform to register the DP as an official opposition party, until 2002 the Registrar of Political Parties has persistently denied DP registration on the grounds that the party had no representation on the Isles. Even earlier, as CUF emerged as a somewhat confusing coalition between James Mapalala’s Mainland Chama Cha Wananchi (CCW), and Shaaban

Mloo’s Zanzibar-based KAMAHURU, the CCM’s attacks against oppositionists took on vicious tones and accusations, as the incumbent felt increasingly insecure about its Zanzibar tenure. The

CUF was roundly accused of “selling out to Islamic Fundamentalists” and Arabs who were exiled following the 1964 Revolution. For some CCM leaders, the agenda of CUF was clear:

“plotting to Islamize and Arabize the Party, Zanzibar, and possibly the United Republic”

(Mmuya & Chaligha, 1994, p. 64).

While the basis for these accusations is unclear, and most likely fictional in almost every respect, what is clear is that accusations such as these resounded in an environment already experiencing rising religious tensions. To complicate matters even more, given the fact that

Muslims made up well over 90 percent of the Zanzibar population, attempts by CUF to mobilize support around the Union question contained obvious risks that questions around the Union would increasingly sound like question about religion. Data indeed shows that political alignments do correspond with religious ones, especially in relation to the Mainland versus

Zanzibar.

Table 11.04 convincingly demonstrates that, at the time of multiparty reforms, bifurcations of political support also coincided with religious differences. While 46.6 percent of the founding members identified themselves as Christian, some 65 percent of the NCCR-Mageuzi’s

289 supporters are Christian. These percentages are even higher among some of the smaller parties, such as the TPP, NLD, and PONA. At the same time, while 42.7 percent of the respondents identified themselves as Muslim, an astounding 93.5 percent of the CUF founding members were Muslim.

Table 11.04: Religious Background of Founding Members NCCR CHAD CUF UMD TPP NLD PONA TYDP Total Christian 65.0 % 43.5 % 3.9 % 9.8 % 83.8 % 93.0 % 87.1 % 9.6 % 46.6 % Muslim 33.0 % 35.0 % 93.5 % 50.0 % 4.1 % 7.0 % 5.2 % 86.8 % 42.7 % Other 2.0 % 21.5 % 2.7 % 40.2 % 12.1 % 0.0 % 7.7 % 3.6 % 10.7 % n 2000 2000 3000 2000 2000 2000 2000 2080 17080 Source: Mmuya and Chaligha, 1994, p. 51

Data from a survey carried out by REDET a few years after the one presented in the table above confirms that religious differentiation is associated loosely with political differentiation.

As depicted in table 11.05, when cross tabulating the raw data from the survey, we see that the opposition is more heavily composed by Muslims. The probability of picking a member of the opposition (n=165) who is also a Muslim (n=97) was 58.8 percent, versus 38.8 percent for the

CCM and 36.5 percent for the sample. Stated differently, if we compared the religion demographics of the party affiliates to the religion demographics of the entire sample, we see that the CCM’s membership profile closely matches the profile of the sample, while the opposition’s profile does not.

Another REDET (2000) poll conducted solely on the Mainland in the lead-up to the 2000 elections clearly shows that this relatively strong association between religious faith and political persuasion carried over into Mainland political divisions. As reported in table 11.06 below, it is clear that Lipumba’s Mainland support tended to be Muslim, while Mkapa and

Cheyo backers were overwhelmingly Christian. Some 76 percent of the respondents who backed

Lipumba were Muslim, compared to 17 percent for Mrema, 29 percent for Cheyo, and 42

290 percent for Mkapa. Again, the religion profile of Mkapa’s supporters closely parallels the sample profile.

Table 11.05: Religious Faith according to Political Affiliation CCM Opposition No Party Other Total None 2.7 % 2.4 % 6.7 % 6.8 % 4.1 % Traditional 0.2 % 0.0 % 1.4 % 2.7 % 0.7 % Christian 58.3 % 38.8 % 62.1 % 48.6 % 58.8 % Muslim 38.8 % 58.8 % 29.8 % 41.9 % 36.4 % n 2915 165 1661 74 4815 Source: Mushi, Mukandala & Baregu, 2001

Table 11.06: Religious Faith according to Candidate Support Mrema Cheyo Lipumba Mkapa Total Christians 83.0 % 71.4 % 23.6 % 57.9 % 54.8 % Muslims 17.0 % 28.6 % 76.4 % 42.1 % 45.2 % n 47 21 106 534 708 Source: Research and Education for Democracy in Tanzania, 2000

However, when cross tabulating the data in order to show political orientation within

Muslim and Christian communities, it becomes clear that, while CUF’s agenda might attract more attention from Muslims versus Christians, the CCM still overwhelmingly attracts support from the majority of each community. As indicated in table 11.07, while 25 percent of the

Muslim respondents backed Lipumba, roughly 70 percent of the Muslim and nearly 80 percent of the Christian respondents still backed Mkapa. Therefore, while CUF may be considered a

‘party of Muslims,’ the oft cited claim by a few renegade CCM candidates that label CUF as a party for Muslims is highly inaccurate.

Data from Afrobarometer, which has been collected from more recent surveys, mimics the patterns found in the REDET data. It is important to note that, according to the

Afrobarometer data, there were no discernable associations between intra-Christian and intra-

Muslim faith differences and political affiliation. This was true for Tanzania as well as the comparative cases. In short, Catholics and Protestants, as well as Sunni and Shiite appear to

291 have highly similar political support patterns. Instead, the associations between religion and political support are far more significant when comparing Christians to Muslims. As demonstrated in table 11.08, the overwhelming majority of Muslims and Christians throw their weight behind the incumbent. The probability of finding a Christian who is also a member of the

CCM is 95 percent, while the same probability for a Muslim is almost 88 percent. At the same time, the probability of finding a Christian who is also a member of CUF is slightly over 1 percent, while the same probability for a Muslim is nearly 12 percent.

Table 11.07: Candidate Support according to Religious Faith Mrema Cheyo Lipumba Mkapa n Christians 10.1 % 3.9 % 6.4 % 79.6 % 388 Muslims 2.5 % 1.9 % 25.3 % 70.3 % 320 Total 6.6 % 3.0 % 15.0 % 75.4 % 708 Source: Research and Education for Democracy in Tanzania, 2000

Table 11.08: Political Affiliation According to Religion Christian Muslim Other None Total CCM 95.1 % 87.8 % 100.0 % 96.2 % 93.2 % CUF 1.4 % 11.6 % 0.0 % 3.8 % 4.3 % CHADEMA 1.7 % 0.6 % 0.0 % 0.0 % 1.3 % NCCR-M 0.1 % 0.0 % 0.0 % 0.0 % 0.1 % TLP 1.4 % 0.0 % 0.0 % 0.0 % 1.0 % UDP 0.3 % 0.0 % 0.0 % 0.0 % 0.2 % n 770 319 13 53 1155 Source: Afrobarometer, 2005d

Cross tabulating to find the religion profile of each party, we find similar trends as those indicated in the aforementioned table 11.06. Below, table 11.09 shows that the probability of finding a CUF member who is also a Muslim is 74 percent, while the same probability of finding a Christian is 22 percent. By contrast, the likelihood of selecting a CHADEMA affiliate who is also a Muslim is 13.3 percent, compared to the 86.7 percent likelihood of finding a

Christian. Not one of the 11 TLP respondents was a Muslim. At the same time, the religion profile for the CCM closely matches the religion distribution for the entire sample.

292

The data presented thus far has one major shortcoming. As cross tabulation narrows the categorical scope, it also renders some of the samples to small to generalize to the larger population. Taken as a whole however, the data aggregates into a clearer and generalizable suggestion: religious differences are not a powerful force for herding the bulk of these communities to two different political directions. Only at the margins, and largely in the context of the Union question, do we find the coincidence of religious and political persuasion. As illustrated by the disproportionate number of Muslims in CUF and the disproportionate number of Christians in what are effectively Mainland opposition parties, the Union question is entangled with the demographic differences between the Isles and the Mainland.

Table 11.09: Religion according to Political Affiliation Christian Muslim Other None n CCM 68.0 % 26.0 % 1.2 % 4.7 % 1076 CUF 22.0 % 74.0 % 0.0 % 4.0 % 50 CHADEMA 86.7 % 13.3 % 0.0 % 0.0 % 15 NCCR-M 100.0 % 0.0 % 0.0 % 0.0 % 1 TLP 100.0 % 0.0 % 0.0 % 0.0 % 11 UDP 100.0 % 0.0 % 0.0 % 0.0 % 2 Total 66.7 % 27.6 % 1.1 % 4.6 % 1155 Source: Afrobarometer, 2005d

When placed into comparative perspective, the association between politics and religion in

Tanzania does not appear exceptionally robust or weak. While the Mwebechai Mosque riots in

1998, and the attacks against pork butcheries in 1993 are indicative of latent religious tensions, in Kenya, arsons against places of worship during election cycles, in some cases resulting in dozens of deaths, has been just as disturbing. A recent, unedited document from the Kenya

National Commission on Human Rights (KNCHR) implicated religious leaders, among others, as helping foment the violence following the 2007 election (Mburu, 2008).

The role of religion in Kenyan politics in the lead-up to reform is in some ways a mirror image of religion’s role in Tanzania. Whereas in Tanzania, religious mainstreamers largely

293 stood behind the status quo, in Kenya, it was mainstream Muslims, Catholics, and Protestant organizations which stood up to oppose Moi’s incumbency and expressed demands for political change. At the same time, the more evangelical groups, most notably the Kenya Assemblies of

God, were among Moi’s most dedicated loyalists.

Afrobarometer data indicates that religious differences in Kenya are associated with political party affiliation to about the same extent found in Tanzania. Table 11.10 below cross tabulates the same type of data for Kenya as presented for Tanzania in tables 11.08 and 11.09 above. For starters, as presented in part one of table 11.10, KANU’s religious profile is disproportionately Muslim, at 38 percent of the 100 Muslim respondents, compared to 18.9 percent of the 850 Christian respondents. By contrast, NARC attracted a far greater proportion of the Christian respondents that Muslim respondents.

Table 11.10: Religion and Party Affiliation in Kenya Part 1: Party according to Religion Part 2: Religion according to Party Chr Mus Oth No Tot Chr Mus Oth No n

NARC 54.5 38.0 30.8 41.2 52.2 90.4 7.4 0.8 1.4 512 NARC LDP 19.9 20.0 46.2 35.3 20.5 84.1 10.0 3.0 3.0 201 LDP DP 2.6 0.0 0.0 0.0 2.2 100.0 0.0 0.0 0.0 22 DP Ford‐K 2.1 0.0 0.0 0.0 1.8 100.0 0.0 0.0 0.0 18 Ford‐K Ford‐P 0.9 0.0 15.4 0.0 1.0 80.0 0.0 20.0 0.0 10 Ford‐P KANU 18.9 38.0 7.7 23.5 20.8 78.9 18.6 0.5 2.0 204 KANU NPK 0.5 1.0 0.0 0.0 0.5 80.0 20.0 0.0 0.0 5 NPK Shirik 0.1 3.0 0.0 0.0 0.4 25.0 75.0 0.0 0.0 4 Shirik Ford‐A 0.2 0.0 0.0 0.0 0.2 100.0 0.0 0.0 0.0 2 Ford‐A SDP 0.1 0.0 0.0 0.0 0.1 100.0 0.0 0.0 0.0 1 SDP Other 0.1 0.0 0.0 0.0 0.1 100.0 0.0 0.0 0.0 1 Other n 850 100 13 17 980 86.7 10.2 1.3 1.7 980 Total All numbers in percent, except column and row n Source: Afrobarometer, 2005b

In part two of table 11.10, it is clear that, out of the 10.2 percent of the Muslim respondents,

75 percent were Shirikisho loyalists, while 20 percent and 18.6 percent backed the NPK and

294

KANU respectively. Furthermore, the proportion of each religion category which supports

NARC is about the same as the proportion of respondents for each religion in the entire sample.

Taken together, the data posted in table 11.10 does not look fundamentally more or less robust than what was found in Tanzania.

Data for Malawi, as presented in table 11.11, looks somewhat similar to the data for Kenya.

The most substantial relationship was found between the UDF and Muslims. Turning first to part one of the table, while 31.2 percent of the respondents claim to support the UDF, the rate of support among the 100 Muslim respondents was robust at 80.3 percent. At the same time, the

MCP, which claimed 20.8 percent of the respondent support, had the weakest performance among Muslim respondents, at 2.9 percent, and a disproportionately large share of support from the “other” and “none” categories, at 38.9 and 40 percent respectively. The DPP’s support, which scored 44.4 percent of the sample support, was disproportionately strongest among

Christians, at 50.3 percent, while disproportionately weak among Muslims, at 16.8 percent.

Table 11.11: Religion and Party Affiliation in Malawi Part 1: Party according to Religion Part 2: Religion according to Party Chr Mus Oth No Tot Chr Mus Oth No n

AFORD 1.2 0.0 3.7 0.0 1.2 81.8 0.0 18.2 0.0 11 AFORD DPP 50.3 16.8 37.0 40.0 44.4 88.2 5.5 4.8 1.4 415 DPP MCP 22.4 2.9 38.9 40.0 20.8 84.0 2.1 10.8 3.1 194 MCP MGD 0.3 0.0 0.0 0.0 0.2 100.0 0.0 0.0 0.0 2 MGD NAD 0.1 0.0 0.0 0.0 0.1 100.0 0.0 0.0 0.0 1 NAD PPM 1.0 0.0 1.9 0.0 0.9 87.5 0.0 12.5 0.0 8 PPM RP 1.0 0.0 0.0 0.0 0.7 100.0 0.0 0.0 0.0 7 RP UDF 23.1 80.3 18.5 20.0 31.2 57.7 37.8 3.4 1.0 291 UDF Other 0.7 0.0 0.0 0.0 0.5 100.0 0.0 0.0 0.0 5 Other n 728 100 100 100 100 77.9 14.7 5.8 1.6 934 Total All numbers in percent, except column and row n Source: Afrobarometer, 2005c

295

These relationships are duplicated in part two of table 11.11. The likelihood of selecting a

Muslim who is also a UDF supporter is 37.8 percent, compared to the 14.7 percent probability of selecting a Muslim out of the entire sample. At the same time, Christian support for the DPP and MCP appears disproportionately strong. The likelihood of selecting a Christian who is also a DPP or MCP supporter is 88.2 and 84 percent respectively, while the probability of selecting a

Christian out of the entire population is 77.9 percent. Similarly, both parties have disproportionately weak support among the Muslim cohorts. Again, when taken together, the data posted in table 11.11 does not look fundamentally more or less robust than what was found in Tanzania.

Conclusion

The qualitative and quantitative data presented in this chapter indicates that polarization is a term that has grown to more accurately reflect religious social tensions in Tanzania. Militant discords among Christians and Muslims have been on the rise for the past several decades, contributing to a widening of intra-Muslim and intra-Christian rifts and sharpening the discourses between Christians and Muslims. At the same time, the coincidence of class-based, especially education and faith-based indicators, should be quite alarming for the incumbent party and might reflect a big shortcoming in the promise of African Socialism. Differentials in access to education, jobs, and wealth will continue to provide real world references for cultivating sentiments of unfairness and injustice that will only accentuate demarcations of faith.

There are however two key qualifications to this increased divergence. First, Afrobarometer and REDET data suggests that, in comparative terms, relationships between religion and politics is not all that robust or unique. The CCM still overwhelmingly manages to boast a membership

296 profile that closely approximates the religion profile of the population. Secondly, the relationship between religion and political affiliation is difficult to separate from more specific tensions on the Isles, and between the Isles and the Mainland. While a high volume of Muslims appear on CUF’s membership roster, this largely reflects the fact that the Party’s agenda is bound to Zanzibar, where the population is almost totally Muslim. Similarly, the high volume of

Christians on the rosters of CHADEMA and TLP reflect the fact that, for all practical purposes, these parties are Mainland organizations hailing from places where Christian populations are disproportionately high.

This latter qualification raises another interesting point about the survey results uncovering educational inequalities across religious lines. Without a deeper analysis of the data, which is beyond the resources available for this study, it is difficult to say the degree to which higher enrolment rates for Christians over Muslims reflects the different style of politics on the

Isles. Indeed, under the ASP, and the CCM, citizens living on Pemba, which are almost all

Muslim, were economically and politically excluded to a degree not found elsewhere in the

United Republic. More generally, from the perspective judging the equalization success of

African Socialism, Education for Self-Reliance was hardly a guiding philosophy for education policy on the Isles. In fact, the citizens of Pemba, which were treated more like subjects during single-party rule, have been repeatedly thought of not as citizens at all, but outsiders with international loyalties rather than national ones. With this statement in mind, let us now turn to the final empirical chapter, which discusses the constitution of national identity as glue that denies opposition parties an ideational basis for mobilizing discontent.

297

CHAPTER TWELVE IDENTITY AND THE CONVERGENCE OF SOCIAL FORCES

“We must tighten our ranks ever closer. We must work as an organic whole, not only to hold our own against the onslaughts of those who desire our downfall, but to further that second revolution on which we embarked when we took the road of independence and nationhood…”

Kwame Nkrumah, 1963

“The ultimate success in the work of building socialism in Tanzania-as elsewhere-depends upon the people of this nation. For any society is only what the people make of it. The benefits to the people of a socialist society will depend upon their contributions to it-their work, their co-operation for the common good, and their acceptance of each other as equals and brothers.”

Julius K. Nyerere, 1968

Always jarry from a combination of uneven roads, frenzied traffic patterns, and the remainder of what was once perfectly rolled wheel-bearings, the half-hour taxi ride from Julius

Nyerere International Airport to the Dar es Salaam city center never fails to be filled with lively, sometimes circumspect conversation with the taxi driver. A warm “karibu” is routinely followed by a sentence or two honoring Tanzania’s peaceful history, along with a few trailing words reassuring that Tanzania’s past is also its present and future. Probing deeper into conversation, and a quick refresher in Kiswahili, reveals a powerful sense of pride in Tanzania’s truly remarkable unity, to the point where the periodic clashes between CUF and the security forces over political situation in Zanzibar fades from betraying this sense of pride. Indeed, for many international onlookers, such pride is often seen as misplaced, and a borderline reckless ignorance about the sometimes deadly realities of the Union between the Mainland and the Isles.

However, when looking comparatively, as the myths and realities behind national pride often do, there is a large degree of justification for the drivers’ senses of community achievement in preserving peace and unity throughout most parts of the country. In fact, I go as far as to suggest that, compared to nearly any other 945 thousand square kilometer plot of

African terrain, or among some other cluster of 38 million people inhabiting the Continent, in

Tanzania we can talk more solidly of a common corporate identity which spans nearly every corner of the country, including some sections of Zanzibar. It is the nature of this national consciousness that is perhaps Tanzania’s biggest post-independence success story, and, coincidentally, a major component for the CCM’s continued incumbency today.

The causal relationship between national consciousness and CCM incumbency can be defined through two venues. First, national consciousness weakens the potency of sectional discourses, which might otherwise serve as a catalyst for political mobilization. The nature of the relationship between polarization and political contention has already been repeated time and again, and does not need repeating here. Suffice it to say that by integrating competing and complimentary local identities, and constructing a cultural narrative which differentiates the community from the external world (March & Olsen, 1995, p. 53), national consciousness matures to form a common thread which stitches together society’s otherwise disconnected social forces, and, in the terminology of this piece, weakens polarization. Secondly, and most relevant for the remainder of this chapter, the relationship between national unity and incumbency is an explicit extension of the ideational space monopolized by the CCM: as the institutional impetus and guardian of national unity against the constant barrage of internal and external threats to nationhood. Let us fill in some background before specifying the theory and data behind this empirical connection.

The goal of this chapter is to demonstrate that two important macro-structural and micro- level features of the Tanzanian polity make the CCM’s tenure a genuinely exceptional case.

299

First, the comparatively high level of social coherence, or low levels of macro-structural divergence, is derived from a comparatively powerful shared national identity. Secondly, this shared identity operates as a macro-level causal force for undermining sub-national loyalties, and a micro-level causal force for feeding the success of CCM political campaigns. In order to demonstrate these claims, I start off by reviewing some of the policy components of African

Socialism, as discussed in chapter six, and argue that these policies are indispensible casual mechanism for explaining Tanzania’s powerful shared sense of nationhood. I then move on to show how fear of internal and external threats to nationhood conditioned the initial skepticism toward multiparty reform and the popular embracement of the CCM as the nation’s guardian.

Finally, I move on to a discussion of how this fear operates as perpetual capital for feeding the

CCM’s success during election campaigns.

The Exceptional Nation

Given the history of colonial partition, along with the ‘tribal’ concoctions which followed, and the subsequent diversity of historical narratives, stereotypes, and languages which were encompassed by the post-independence African state, the ideal of some abstract correspondence between a common historical narrative and state which embodied that narrative seems hopelessly Eurocentric. Of course, throughout the plethora of independence struggles across the

Continent, there were shared perspectives over the immorality of colonial rule, which helped these movements cohere. National and “Afro-centric” consciousness was defined from a counter hegemonic position, opposed to the Eurocentric representations of the “exotic” and “primitive”

African other. Yet, as a force of counter hegemony, African history and potentialities were defined in terms of colonial yardsticks of moral and material progress (Olaniyan, 1995). Local

300 oral traditions, often interpreted by Europeans as mere folklore or legend, were upheld by

African writers and political leaders as having a powerful historical role in collective consciousness formation, and in some cases, as exemplars of national consciousness (Lüsebrink,

2002).

During the times of independence struggles, forging a ‘homeland’ for a community based upon common language, religion, or historical and cultural interpretation (Skidmore, 1993, p.

254), was seen by many of these western educated political leaders as not only vital for independence and national survival, but as the normal mode of organizing modern political entities. Typically, networks of personal loyalty and kinship was transposed to a territorially bound collective through geological and geographical metaphors, such as “soil” and

“homeland”, and biology, such as ‘brotherhood’ and ‘family hood’ between citizens (Tambiah,

1985, p. 4; Trinh, 1989, p. 94; Malkki, 1992, p. 26). The state was painted with a paternalistic edifice, a defender of the territorial integrity of the nation and moral well-being of its citizens from foreigners seeking to create disorder and disunity with an eye for continuing the exploitive practices of the past. Where nationalism was successful, it legitimated a seemingly natural and universal dominance of the state while supplanting segmented loyalties and eliminating opposition groups (Kilson, 1963; Eriksen, 1991; Bačová, 1998).

However, in the typical case, nationalism failed to supplant segmentation. Instead, the state and the party dominating its offices, succumbed to the persistent need to stitch together fragile coalitions through a delicate mixture of tactical exclusion and inclusion. What emerged was a

“party-state” which, while espousing the virtues of African unity and the desire for nationalism, was in reality a highly fragmented regime perpetually on the brink of collapse and military takeover (Zolberg, 1966).

301

The failures of nation-building in Tanzania were far more limited, as the ingredients for the diffusion of a national consciousness were far more complete. If lower degrees of macro- structural divergence are indicative of greater probabilities in forging shared cultural narratives, then the relative weaknesses of class and ethnic differentiation and resource inequalities outlined a development trajectory lendable to forging a national consciousness. To be sure, forging a national consciousness was always central to Tanzania’s development efforts. Whereas institutions for social control in neighboring countries often were delegated to local authorities, which sometimes had their own interpretation of past and present political conflict, in Tanzania, the consolidation of social control was single-mindedly focused on the formation of a national consciousness. The dismantling of the native authority system and the failures of petty bourgeoisie development outside the reach of the party-state nexus, translated into a rural society whereby the capacity and incentive for rural elites to consolidate and mobilize localized discourses were far weaker than what was found in neighboring countries.

Polarization was additionally undermined by a collection of other policies. In chapter six, I described education and language policy, and again suggested that both were directed toward defining and consolidating national consciousness. The use of Kiswahili was universalized, selection for secondary school enrolment based on a quota system, and regional uniformity was established in the distribution of education and other social services (Okema, 1996). Askew

(2002) suggests that the spread of Kiswahili, and the fact that it was disassociated with any particular ethnic group, along with the Party’s efforts to construct it as the “language of revolution” and “unity,” without a doubt “enabled and facilitated the spread of national consciousness” (p. 182). Okema (1996) shows that Kiswahili helped “diffuse” social tensions by making it easier for exchanges of humor between social cleavages (p. 59). Miguel (2004)

302 demonstrates that language policy in Tanzania is one of the principle bases for the weakening of ethnic saliency, the development of a national consciousness, and greater success in cross cultural cooperation in Tanzania (p. 335).

The bonds between individuals and nation were infused with additional adhesive through

Education for Self-Reliance, which was more stridently dedicated to education for commitment to social change, including commitments to the aspirations of the nation over the personal ambitions of the individual. Civic education aroused suspicion of sub-nationalism, imperialism, and racism, along with the promotion of a sense of community responsibility. The goal, according to Cooksey, Court, and Makau (1994), was to undermine “separateness” based on regional origin or social status, in favor of a “deeper national identity of the socialist Tanzanian”

(p. 203). As a consequence, Tanzanian students have a strong sense of national identity, especially when compared to similarly situated students in Kenya, Zambia, or Malawi (Court &

Kinyanjui, 1980, p. 69, Miguel, 2004, p. 336).

Similarly, aside from raising the literacy rate among the adult population, adult education was also geared toward mobilizing the masses and instilling a “consciousness of Ujamaa”

(Okoko, 1987, p. 69). One-year national service obligations were also implemented at end of secondary school. National service was tailored toward bringing young adults out of their home areas by moving them to different regions of the country, and bringing them into contact with peers from different ethnic backgrounds. This process of integration was generally encouraged at every stage of education beyond primary school. Mwakikagile (2004) describes the school that he attended as “fully integrated,” where African students “lived in the same hostels with

Asians and Arab students” and were taught by African, Asian, and European teachers (p. 110).

303

The assertion of national imagery and the promotion of cross-regional contact went beyond education and language policy. Symbols, such as the national motto, “Uhuru na Umoja”

(Freedom and Unity) and the Uhuru torch were promoted by the party as representations of

Tanzanian nationalism, as well as symbols of TANU. Perhaps the most telling event is the

Uhuru Torch Race, which, until 1992, was organized once a year by the TANU Youth League, later renamed Umoja wa Vijana1. In the race, the Uhuru Torch was carried throughout every region in Tanzania, including those on Zanzibar. To have the torch raced through your village is a big deal indeed, typically met with festivities and celebrations which elicited participation from nearly every villager. Additionally, villagers participate in carrying the torch to the next village and are encouraged to contribute fuel for the torch. As Okema (1996) points out, the

Uhuru Torch Race was a main event for involving “citizens as active participants” in the life of the nation (pp. 53-4).

There is ample evidence showing that these aforementioned development policies and rituals did produce a comparatively strong sense of nationhood in Tanzania. Much of this evidence was discussed in chapter ten’s coverage of ethnicity. The smoking gun is the Afrobarometer data presented in table 10.05, which is derived from a question which asks respondents to rate whether they felt closer to an ethnic identity or a national identity. Amazingly, 79.6 percent of the Tanzanian cohorts cited feeling close to “national identity only,” compared to 24.7 percent for Kenyans, 23.9 percent for Malawians, and 19.7 percent for Zambians. The plurality of the

Kenyan and Malawian respondents, and the majority of the Zambian respondents, felt equally close to national and ethnic identities.

1 After multiparty change in 1992, the responsibility for the torch race was moved from the CCM to the National Service.

304

Tanzania’s major exception is Zanzibar, where the ability to construct a constellation of policies which would bring the Isles into a broader sense of national consciousness was constrained by the highly polarized macro-structural environment overlaid by the 1964

Revolution. The competitiveness born out of this polarization is described with the election results presented in chapter five. Out from the parity of the election results, and the destructive potential inherent to the self-other discourses on the Isles, Karume’s approach to governance emerged as far more autocratic when compared to the rule on the Mainland. The oppressiveness of the Revolutionary Council did little to moderate pre-independence macro-structural polarization. Pemba in particular was the target of systematic oppression along with political and economic exclusion. The treatment received by the Island’s population is cited as a primary reason why Pemba today is such a strong opponent of the CCM (Kweka, 1997, p. 146). Sheriff

(1994) picks up this point by claiming that the Revolution “suppressed the other parties to establish a one-party state by force, thus depriving nearly half the [Zanzibar] people of their democratic rights”. He goes on to claim that the “socio-political division in the country has consequently festered under the surface of imposed unanimity” and constitutes a major source of political conflict today (p. 133). However, with the blocking of political expression through party politics, the channels for discontent became the Mosques, which heavily populate the Isles and the coastal areas of the Mainland. Consequently, as presented in the previous chapter, Union matters have assumed religious overtones, while religious matters have, at times, assumed anti-

Union positions, and thus, positions easily rendered as threatening to national order and tranquility (Sheriff, 1994, p. 146).

With the Zanzibar caveat noted, the strength of national identity in Tanzania acted as the thread which pulled together cleavage differentiation, thus neutralizing the ideational base for

305 mobilizing contention and opposition. Aside from growing discontent against corruption and authoritarianism, at the onset of international pressures for political reform, there were few alternative ideational domains by which incumbency could be rejected. Given social cleavage weaknesses, political elites found little incentive to stake out political careers as an opponent of the CCM. Thus, not only did a national cohesiveness deny the ability of party outsiders an ideational basis for mobilizing discontent, it also provided a sense of intra-party cohesiveness resistant to the fragmentary tendencies found in cases like KANU and UNIP.

Furthermore, the legitimacy of the regime itself rested on a principle of standing above sectarian conflict, specifically ones revolving around ethnicity. Miguel (2004) provides a nice summary:

Taken together, the various components of the Tanzanian reforms – the promotion of Swahili as a national language, political and civic education in schools, the dismantling of tribal authorities, and the relatively equal regional distribution of resources – contributed to the growing strength of a coherent and popular national identity that binds Tanzanians together across ethnic lines (p. 338).

Indeed, when compared to the weak social base for opposition formation, the social basis for the

CCM appeared almost impossible to infiltrate and divide. This was especially true when national unity is painted against the background of destruction in neighboring countries. For most Tanzanians, it is the CCM which prevents this chaos from exploding in the United

Republic.

Vanguard of National Unity

Following the unruffled push for multiparty reform, two new events presented the CCM with additional opportunities for projecting the unifying capacity of the Party as justification for continued incumbency. The first, and most immediate, was intense conflict and genocide in a

306 number of Tanzania’s closest neighbors. While sectarian conflict and civil war were, since the times of independence, constant reminders of the potential dangers of political fragmentation, the genocide in Rwanda in 1994, which resulted in the death of some 800 thousand Rwandans and the flood of an additional nearly 2 million refugees into Zaire and the northeast portions of

Tanzania, only heightened the apprehension around the potential instability and insecurity ensuing from conflict between sub-national identities. An equally important lesson was the ethnic violence brought on in the lead-up to the 1992 multiparty election in the otherwise peaceful and stable Kenya, which left some 1,500 dead and hundreds of thousands internally displaced (Africa Watch, 1993). Violence in Burundi in 1988, and a civil war following the

1993 assassination of President Melchior Ndadaye, along with the escalation of violence in

Zaire following the fall of Mobutu, served as additional warnings of the lurking danger of sub- nationalism.

Secondly, the intensification of fears against terrorism following the 1998 bombings of the

American Embassies in Dar es Salaam and Nairobi, bolstered apprehension around the threats of foreign intrigue and infiltration, and the necessity of preserving regime competence and commitment against a world of increasing hostility and insecurity. These fears only increased after the offensive against the United State in September, 2001. All-in-all, the position repeatedly articulated by the CCM was one of a seemingly perpetual sense in insecurity. As folks like Ramadhani Omar Mapuri, the CCM publicity Secretary, would time and again emphasize, subnational interests, rising criminality and subversive international agendas were, through chaos and bloodshed, bent on destroying the very foundations of Tanzanian society, and could only be combated by the CCM.

307

Popular fears toward sub-national and foreign threats, for better or for worse, functioned as the basis for early popular skepticism toward multiparty reform and an asset readily accessed by the CCM in the post-reform context. The subjective fact of this knowledge structure posits the

CCM as the creator and protector of nationhood, composed of elements protecting against the dangers of divisiveness and the meddling of foreign intervention. Alfred Sebit Lokuji (1995), of the Department of Political Science and Public Administration at the University of Dar es

Salaam, points to the touch of irony in relation to ethnicity and reform. For Lokuji, despite the fact that Tanzanians show minimal tendencies toward ethnic favoritism, “the ‘fear of violent conflict’ seems to be the thread that holds together the initial phobia for multipartyism.”

Lokuji’s perspective is confirmed by data from a number of surveys conducted around the time of political reform. When describing the results of a 1990 survey, Baregu and Mushi

(1994) claim that 69 percent of the electorate preferred single-party elections, while 17 percent desired multiple parties. The remaining cohorts were undecided. Of the more than two-thirds which preferred single-party rule, 89 percent cited “unity and peace” as the most important justification. Again, out of those favoring single-party elections, some 96 percent preferred the

CCM, as opposed to those wanting a different party to preside over single-party elections (p.

116).

Various data presented in the findings of the 1991 Nyalali Commission, which was charged with the task of collecting opinions and concerns from all corners of the country over the issue of political pluralism, also supports Lokuji’s statement. First, as captured in tables 12.01 and

12.02 below, there was an overwhelmingly strong support throughout Tanzania for the continuation of single-party rule, although Zanzibaris were far more accepting of the prospects of political pluralism when compared to the mainland cohorts. All-in-all, nearly 80 percent of

308 the Mainland respondents prefered to continue single-party rule, while slightly over 55 percent of the Zanzibar respondents held the same view.

Table 12.01: Continuation of the Single Party System Yes No No Response No Committal Mainland 79.7 % 19.0 % 0.7 % 0.6 % Zanzibar 55.4 % 43.0 % 0.2 % 0.5 % Total 77.2 % 21.5 % 0.7 % 0.6 % Source: United Republic of Tanzania, 1992, p. 69

Table 12.02: Adoption of a Multiparty System Yes No No Response No Committal Mainland 18.8 % 76.3 % 4.3 % 0.6 % Zanzibar 42.7 % 55.6 % 1.3 % 0.5 % Total 21.5 % 74.1 % 3.9 % 0.6 % Source: United Republic of Tanzania, 1992, p. 70

Additional data provided in table 12.03 shows that, 56.6 percent of those who supported change rejected immediate changes in favor of a more gradual approach. Moreover, the reasons given for supporting the continuation of single-party rule included the belief that the “single- party system brings national unity and peace,” enhances “national defense and security,” and undermines “racial and religious discrimination” (United Republic of Tanzania, 1992, p. 77).

This basis for this support is consistent with the reasons given for rejecting multiparty reform, including the fear that multiparty competition will “increase divisive tendencies in the Union,” bring “racism, discrimination and anarchy” which will “divide the people into irreconcilable groups”, the view that supporters of political pluralism are “power hungry” and greedy, and the perception that political pluralism is imposed from forces outside possible due to a

“neocolonial” agenda (United Republic of Tanzania, 1992, p. 78). Similarly, among those responses reported in table 12.01 which supported a gradual shift toward pluralism, most were concerned about the possible “chaos” brought about by abrupt change (p. 88).

Another survey conducted by Gero Erdmann (1994) a few years after the Nyalali

Commission Report indicated a generally more welcoming attitude toward political change. At

309 first glance at least, it would appear that, either people’s views were changing or the findings of the Nyalali Commission were flawed2. A total of 48.5 percent felt that the decision to go multiparty was good, while 43.5 percent felt that the decision was not a good move. However, the 51.1 percent of the rural sample, which at the time, made up some 80 percent of the

Tanzanian population, thought that the 1992 reforms were not such a good idea, compared to the

43.1 percent that was more receptive. Additionally, the general fears toward political change were undeniably consistent with the findings of the Nyalali Commission, suggesting that their findings had at least some credibility. As the following two tables illustrate, there was a general skepticism toward possible benefits of multiparty change, and, more critically, a widespread suspicion that pluralism would lead to disunity. Table 12.04 shows that, by far, the majority believed that the development of the country and of the lives of Tanzanians would not improve because of multiparty change. Table 12.05 clearly reports an overwhelmingly widespread belief that multiparty change would undermine the basis for national unity and peace.

Table 12.03: Should Changes be Immediate? Yes No Do Not Know Mainland 26.1 % 56.6 % 8.3 % Zanzibar 55.7 % 35.1 % 9.2 % Source: United Republic of Tanzania, 1992, p. 87

Table 12.04: Benefits of the Multiparty System The Multiparty System Helps Improve Agree Totally Disagree Don't Know the Development of the Country 40.1 % 47.9 % 11.5 % Life of the Peasants 33.6 % 49.8 % 15.7 % Life of the Townsman 33.2 % 47.0 % 18.9 % Life of all People 21.2 % 59.9 % 18.0 % Source: Erdmann, 1994, p. 9

2 I state this simply because of the controversy surrounding the possible bias of the Presidential Commission on Single Party or Multiparty System. It was widely believed that, because the Commission was selected by the CCM, and largely composed of CCM loyalists, the findings of the Commission were flawed. There are two reasons why I am suspicious of such criticism. First, the results on the Commission’s report closely followed the results of the aforementioned 1990 survey. Secondly, the recommendations of the Committee, which at times was highly critical of a wide number of authoritarian legal measures, ultimately embraced the need for reform. 310

Table 12.05: Possible Effects of Multiparty System The Multiparty System Rural Urban Total Is a threat to national unity 68.4 % 54.1 % 61.8 % Will strengthen tribalism & factionalism 67.1 % 51.9 % 60.1 % Is hopeless 62.5 % 45.5 % 54.7 % Will strengthen democracy 29.1 % 42.0 % 35.0 % Source: Erdmann, 1994, p. 10

This apprehension and suspicion probably accounts for most of the overwhelming stability of the CCM’s popular support, irrespective of the Party’s list of election candidates, just prior to

Mrema’s defection that changed the fortunes of the CCM in the polls. This support was clearly revealed in the Nyalali Commission Report, which found that, among those supporting the continuation of single-party rule, almost all preferred the CCM over single-party rule by another party. This position also emerged in the 1994 Erdmann survey, where nearly 85 percent expressed their preference for the CCM (see table 12.06). Judging from this survey, it appears that, while Tanzanians were rapidly thawing to the idea of multiparty politics, they were not thawing to idea of rule by any other party other than the CCM. Given the apprehension surrounding multiparty reform presented in the tables above, one credible interpretation is that the CCM was perceived by those sampled as the most capable organization for guarding against the possible ills from political reform.

Table 12.06: Party Preference Rural Urban Total CCM 88.8% 80.4% 84.9% CHADEMA 2.8% 2.7% 2.7% DP 0.3% 3.3% 1.6% NCCR‐M 0.4% 0.5% 0.5% TADEA 0.6% 0.3% 0.4% UMD/CUF 0.4% Don’t Know 1.5% 2.5% 1.9% None 5.3% 9.9% 7.4% Source: Erdmann, 1994, p. 11

The CCM’s position on Zanzibar is far more tenuous. Historically, the ASP and the CCM have been perceived by roughly half of the island’s population as an institution for the

311 oppressive expression of Mainland interests, over the interests of Zanzibar. Indeed, this position has a strong element of truth. The ASP itself was given life by TANU, and drew support from

Mainlanders who migrated to the Isles to labor on clove plantations. This perception was given full life when the National Executive Committee selected an Unguja old guard named Idris

Abdul Waki, over Pemba’s reformist Seif Sharif as the CCM candidate for the Zanzibar

Presidency. Seif Sharif’s 1988 expulsion from the CCM only precipitated this view (Shariff,

1994).

The motivations behind the Union are widely believed to have been a way for the ASP to hedge against internal threats, a point which continues to be a source of contention for those opposed to the CCM. Furthermore, issues of the legality of the Union, and the corresponding expanded list of “Union matters” at the expense of Zanzibar autonomy, incited heated, sometimes violent political disagreements. Under single-party rule, and within the CCM specifically, raising these questions however, can have major political consequences, as illustrated by the forced resignation of Aboud Jumbe from the Zanzibar presidency after articulating the oft-cited legal argument for a three-tiered federal union form, rather than a two- tiered union (Kweka, 1997).

Popular rejection of the CCM’s rule over the affairs of the Isles is clearly discernable from the aforementioned surveys. Tables 7.8 through 7.10 shows quite convincingly that support for single-party rule was far less entrenched on the Isles. The results of a 1990 survey confirm

Zanzibar’s disproportionately strong position against single-party rule. Whereas 69 percent of

Tanzanians sampled supported single-party elections, only 40 percent of the Zanzibar contribution to the sample held the same view. Support for one-party rule on Zanzibar was lowest among younger cohorts, those engaged in trade or studies, and those that rented property.

312

Not surprisingly, these youth and traders were some of the most outspoken critics against the

CCM and the most committed supporters for political change (Shariff, 1994).

‘Branding’ the CCM

Historically speaking, the role of persuasion, some might say manipulation, as a method of gaining consensus and compliance has been a comparatively strong feature in the formulation and implementation of development policy in Tanzania, especially in relation to Mainland politics. The importance of persuasion, and the corresponding development of knowledge structures, which, in turn, shape subsequent political interactions, cannot be understated. To be sure, there is an overwhelming amount of political party literature citing the importance of ideologies for long-term tenure. Du Toit (1999) for example claims that political regimes possessing “compelling ideologies are likely to secure long-term incumbency with which to try to restructure state and society, and to secure hegemony” (215). Schubert (2004) claims that part of the electoral success of the KMT in Taiwan is derived from the party’s ties to the ideals of

Taiwanese nationalism. In Mexico, it is difficult to discuss the longevity of PRI’s electoral success without referring to the Party’s ability to continue to remake peasant revolutionary imagery during times of elections (Story, 1986; Collier & Collier, 1991; Langston, 2002).

For the sake of simplicity, I distinguish between two contexts in which persuasion might take place. The first context is best characterized as highly agential, where the definitions of pressing issues and needs, societal enemies and allies, are all up for grabs. These turbulent times of “reconstruction” politics3 as it might be called, begs for getting a critical mass of people to

3 I borrow the term “reconstruction” politics from Stephen Skowronek’s (1997) application of “reconstruction politics” to the “politics that presidents make” in the United States. For the author, who wants to explain structural and agential bases for presidential success, the politics of reconstruction is marked by a looming crisis, such as an economic depression, which calls into question an existing set of policy commitments. Status quo 313 reinterpret history, to see themselves and the world around them in a different light, to perceive possibilities for change that were previously allusive to the common eye. For better and for worse, these are the discourses for the making of agrarian revolts, bourgeoisie revolutions, national independence movements, and totalitarian rule.

More relevant for this chapter is the politics whereby institutions and macro-structures have greater causal force. In the aftermath of reconstruction, new institutions emerge with vested interests in ensuring stability and continuity. The emergent political elite render their continued survival by routinely “articulating”4 their central role in maintaining the seemingly “natural” and “just” order of social affairs. In the process of articulation, the relationship between the political elite and the state is not unlike the symbiosis between dominant classes and the state as conceptualized by Gramsci and Poulantzas. For neo-Marxists in general, the “successful mobilization and reproduction of the active consent of dominated groups by the ruling class” is a product of a sustained ideology articulated through the megaphonic institutions of the state and civil society, which legitimate exploitive class relations, whether from a commercial, petty, or bureaucratic bourgeoisie (Jessop, 1982, p. 148).

It would be a far reach to argue that the CCM today, and even TANU under Ujamaa, maintained a consistently compelling ideology. While Ujamaa was perhaps the most clearly

policies or institutional commitments are “vulnerable” since they lack popular support as a means for dealing with a particular crisis. At the same time, the sitting president is “opposed” to the party that earlier established these vulnerable commitments. Taken together, presidents such as Franklin D. Roosevelt, who came into office amid an economic depression and a series of laissez-faire institutional commitments, had a popular mandate and elite support to change them, yielding a relatively agential environment conducive to political change.

4 Again, the term “articulate” borrows from Skowronek’s (1997) framework for conceptualizing American presidential success. The “politics of articulation,” in contrast to reconstruction, is a relatively uneventful leadership situation where the existing commitments are “resilient” and the current president is affiliated with the party that established them. Presidents during these times make promises to maintain the ‘good work of the past’ and demonstrate the regime’s ‘vitality in the changing times.’ President Lyndon B. Johnsons’s expansion of the welfare state was a further articulation of the highly resilient set of policy commitments laid down by Roosevelt’s New Deal.

314 specified variant of African Socialism, many peasants simply treated concepts such as class exploitation and encroaching institutions of the state with skepticism and suspicion. For a day in the life of a peasant, the policy aftermath of Ujamaa appeared as alienating and threatening as the prospects of encroaching capitalism (Hyden, 1980). However, Ujamaa meant more than just socialism. Following the TANU’s success at reconstructing political order during the struggle for independence, the framework of Ujamaa emerged as the scaffolding of national consciousness: the definitions of insider, outsider, and the respective threats to nationhood posed by an assortment of internal and external interests. Therefore, it is perhaps more relevant here to talk about compelling ideas as ones with “semantic density” (Eriksen, 1991, 265), or concepts

“fixed” in everyday language where meanings require no definition (Bačová, 1998, 30). During political battles, these concepts can be expressed in a few key words or phrases that conjure up a socially salient image, which is used in turn to convey some positive values to the conjurer or negative values to the conjurer’s opponent.

In fact, the use of ideas with semantic density may be far more valuable than coherent ideologies for regime survival during the onset and aftermath of economic and political reform pressures. Whereas systematic ideologies may grow rigid and inflexible, making major liberal reforms difficult to carry out without causing major party factions, basic concepts with strong fixity, may still be loose enough to accommodate a wide array of shifting interests within the regime and within the larger national community.

Election campaigns are surely the times where dense terminologies will be used by political contestants to stake out political claims and terrains in order to defend or attack the status quo.

Incumbents will regurgitate justifications for their continued occupancy of state offices, while opponents will try to debunk the status quo by showing that the incumbent either failed to

315 adhere to its own ideals, or that the ideals themselves are flawed. For the incumbent, successful campaigning will largely depend on the regime’s ability to link itself with the preservation of something of salient importance, and the reformation of those areas which need reforming.

A useful way for conceptualizing the use of rhetoric in political campaigns is found in a study by Jacobs and Shapiro (2000) on the use of polling data and focus groups in political campaigns in the United States. In what is described as “crafted talk,” the authors argue that, rather than “pander” to public opinion, as defined by polling data, politicians use polling and focus group research to isolate the most ‘alluring’ words, symbols, and arguments and deploy these findings in campaigns designed to move public opinion in line with desired policy. In other words, existing opportunity structures (alluring words, symbols, and arguments which define the macro-structural context) act as political capital for political elites attempting to ‘sell’ particular policies to the public.

While the work of Jacob and Shapiro concerns politics in the United States, the global use of well paid campaign consultants from industrial democracies, and advisors trained in the most sophisticated marketing tactics, warrants the assumption that “crafted talk” is applicable to election campaigns throughout much of the world. I argue that the CCM has become one of the most successful campaigning machines in Africa. Not only does the CCM excel in raising an enormous pot of financial resources when compared to the array of opposition parties unable to find financing beyond the personal fortunes of their respective leaders, it also appears that much of the polling data that has been referenced in this work, has been used by CCM strategists to

“craft” campaign messages and slogans in order define the stakes of political competition, and the nature of the actors involved. Plainly stated, the monologues crafted and channeled through

316 the communication networks of the incumbent regime5 during much of the campaigning, references many of those fears borne out in the aforementioned surveys. As for the reproduction of incumbency, the CCM’s surviving hegemony is a product of 1) the incumbent’s ability to identify a salient sense of national consciousness and unity, 2) the CCM’s continuous references to the prospects of disunity as the most pressing political issue, and 3) the continued articulation of the CCM as the seemingly natural, if not proven, defender of the nation against those international and sub-national threats which are allied with the opposition. Let me now turn to an assessment of the multiparty campaigns in order to show how the campaign rhetoric reflects these knowledge structures, and how this in turn, solidifies electoral support.

5 The principle mode of communication today is through private and state media outlet. At the dawn of political reform, there were five state or party owned newspapers and one radio station. As of 2005, there were 32 radio stations, 15 television stations, 18 daily newspapers, and some 80 regular publications (Tanzania Election Monitoring Committee, 2006). By-in-large, the media in Tanzania has operated to the benefit of the CCM. In the 2000 elections, the CCM received some 377 hours of radio and television attention, while the next largest attention getter (CUF) received a mere 131 hours (Media Council of Tanzania, 2001, 4). Also, in the 2000 election, a daily news commentary on Radio Tanzania Dar es Salaam (RTD) entitled “Maneno Hayo” was a popular outlet for voicing approval of the party in power (Media Council of Tanzania, 2001, 7). Aside from the CCM owned mouth piece, Radio Uhuru, the government owned RTD and Sauti ya Tanzania Zanzibar (STZ) showed a strong biased toward the ruling party (Media Council of Tanzania, 2001). In what was perhaps the most extreme tale from the 2000 election, Abood Television (ATV), which was owned by a CCM parliamentary aspirant in Morogoro, outright refused to give coverage to the opposition (Media Council of Tanzania, 2001, p.4, Tanzania Election Monitoring Committee, 2006, p. 136). In the 2005 elections, Radio Uhuru, presenters were notorious for cutting into song play with slogans like “CCM Dume” (CCM the Bull) and “Kidumu Chama Cha Mapinduzi” (Long Live the CCM) (Tanzania Election Media Monitoring-2005, 2006, p. 28).As for the private outlets specifically, they have failed to encourage political discussion and information. While the RTD made strong efforts in 2005 toward proving educational information to voters through programs such as Majira, Mbiu za Uchaguzi, Haki Ni Yako, and Mazungumzo Baada ya Habari, it would seem that the private media took the lead in violating the established Code of Conduct, largely through a lack of professionalism through un-sources, concocted stories, splashing sensationalized headlines (Tanzania Election Media Monitoring-2005, 2006, p. 15). In Tanzania Leo, a story headlined “Mkapa ‘aiomba’ mizimu ya Nyerere, Karume” (Mkapa ‘appeals’ to Nyerere’s, Karume’s spirits) simply reported on Mkapa praying at Nyerere’s grave in Butiama, without ever reporting on how a Roman Catholic would rationalize praying to a spirit other than Christ (Media Council of Tanzania, 2001, p. 19).

317

Colonial Stooges and Arab Conspirators

While Nyerere was steadfast in his opposition toward engaging in racially charged rhetoric as a partition for ‘insider’ and ‘outsider,’ the colonial concocted system of tribes, and by extension ethnic formations, were treated as an institutional expression of foreign interests.

Hence, between the time of independence and the dismantling of the native authority system, the chiefs were commonly constructed as ‘colonial collaborators’ (Lindemann & Putzel, 2008).

Additionally, the elite civil servants were constructed by the party leadership stratum as bourgeoisie “colonial stooges,” a perspective which laid some of the foundations of the Arusha

Declaration generally, and the leadership code in particular (Lonsdale, 1968, p. 342).

In Tanzania today, such attacks are most vehemently directed toward the opposition, where the incumbent routinely links the motivations behind opposition to the CCM as derived from some underhanded foreign agenda from “colonialists” or, more recently, “terrorists.” On one occasion, Mkapa links the instability in Cote d’Ivoire, the “divide-and-rule legacy of colonialism,” and “havoc, suffering, death and discord” (Mkapa: we must, 2005) to the distinct possibility of a similar connection between the Tanzania oppositionists and underhanded foreign agendas bent on bringing a similar plight to the United Republic. Hard evidence of these discourses first arises out of the 1993 and 1994 local elections, where a number of prominent incumbents openly claimed that opposition parties were “stooges of foreign forces” committed to “destabilizing” the country (Research and Education for Democracy in Tanzania, 1997, p.

20).

During election campaigns on the Isles, it is often subtly uttered by CCM officials that if

CUF wins on the Isles, the Arabs who fled the island after the 1964 revolution will return

(Mihangwa, 2005). For example, when addressing a rally in Zanzibar, Mkapa talked about

318

CUF’s rejection of the 1964 Revolution, and the resort by CUF, through its connections to interests in the Middle East, to the violence and bloodshed as a means for usurping state power.

He moves on to state that “We will continue to defend our national integrity and independence.

No one should interfere with our independence” (Kasumuni, 2005b). In other words, the contest between CUF and the CCM is defined in terms of the contest played out in 1964, where the spirit of TANU and the ASP within the CCM will once again deliver a victory against foreign impositions. Most recently, the incumbent leaders have turned to fears of terrorism as an asset for tying the opposition, particularly CUF, to threats from foreign interests. Thus, in the wake of the post-2000 election violence in Zanzibar, CCM leaders dubbed CUF as a “terrorist movement” (Opposition rejects, 2001).

The Child and the Petty Squabbler

This sense of chaos is not simply about the agenda of foreigners located in other countries, but about the composition of the opposition as alien to what it means to be a capable leader. The notion that opposition politicians lack the capacity to lead is most visibly articulated in the use of the term ‘child’ as a metaphor for the opposition. Here, the incumbent attempts to relate the young age of the opposition to the “childish” behavior of its leaders as a contrast to the value placed on age, and in some sense, gerontocracy as the necessity of good leadership. During the

1995 election, Nyerere repeatedly referred to the opposition as politically immature “fools,” despite the fact that his son Makongoro was a leading member of the NCCR-Maguezi (Lorch,

1995). During the 2005 election, Uhuru newspaper “systematically portrayed” opposition candidates as simply “unfit” to rule the country (Tanzania Election Media Monitoring, 2006, p.

15). During the same election, the media consistently deployed the term “politically mature” to

319 describe the CCM while “immature” to describe the opposition (Masato, 2004). In one case,

Uhuru published a blatantly biased photo of a TLP councillorship candidate addressing a rally, where the picture was taken in way to show the candidate addressing only the children attendees. The article described the candidate as ‘addressing kids on a street corner’ (Tanzania

Election Media Monitoring, 2006, p. 23).

Members of the incumbent party never miss the opportunity to paint opposition figures as a bunch of petty child-like squabblers, arguing over personal property. The reality of inter and intra-party conflict within the opposition certainly makes for easy references. Inter-party disputes over nominations, such as Cheyo’s outspoken rejection of Mrema’s nomination for the

1995 presidential race, or the DP and SAU objections against Chadema’s Vice Presidential nominees for the 2005 election (National Electoral Commission, 2006), are often made to appear as childish turf wars. The internal splits which threatened the nomination of an NCCR-M candidate in the 1997 Makete by-election, or the UDP candidate in the 1996 Temeke by-election

(National Electoral Commission, 1997), of the splits and expulsions revolving around Mrema’s nomination on the TLP ticket for the 2005 presidential election (Tanzania Election Monitoring

Committee, 2006), and the ensuring court battles between Mrema and Jaffu over the right to lead the TLP (Kahoho, 2005), are often drafted as products of self-interested, immature, and opportunistic opposition leadership. In reference to Mrema’s declining support, a newspaper carried an article in the lead-up to the 2005 election with the headline “Moshi wazidi kurejea”

(Moshi still pouring back to CCM). Articles such as this depict the opposition as self-interest opportunists seeking to be aligned to the most electorally viable party, which fits within the campaign schema driven by the CCM (Tanzania Election Monitoring Committee, 2006, p. 137).

320

The Crook, the Thug, and the Devil

The worst case outcome of an opposition victory is found not in petty squabbling, but in the rapid escalation of violence that accompanies competition between sub-national actors with little respect for law and order of the nation. Again, these representations are not uncommon throughout Africa. Raila Odinga referred to opposition to the Bomas plan as “wakora” (thugs or crooks) (Kenya National Commission on Human Rights, 2006, p. 54). In Zambia Michael Sata and Frederick Chiluba (who backed Sata’s Presidential bid) were routinely called “crooks” or

“vultures” assisted by “thugs” armed to bring down Mwanawasa tenure (Malupenga, 2006).

In Tanzania, the CUF has bore the brunt of what is probably little more than concocted or misrepresented stories of conspiracies and ill-intentioned plans to stir-up ‘trouble’ during elections. Officials repeatedly warned citizens of the intentions of the opposition, principally

CUF, to disrupt elections through mischief. Karume warned Zanzibaris that the opposition was planning to “put on CCM uniforms and bring about confusion” during polling day (Karume to improve, 2005). On the Mainland, in less than two weeks before the scheduled date for the 2005 election, Kikwete linked a series of armed robberies to the disposition of opposition leaders, and stated that “we’ll have to fight until we win, because we can’t allow the robbers to rule our country” (Kihaule, 2005). While touring the Korogwe District, Salma Kikwete claimed that opposition was planning to smear women with chicken’s blood on the day of election

(Mwanakatwe, 2005). Without the disclosure of any sources, Mzalendo passed a story claiming

“vikao vya usiku vya CUF vyapanga mikakati ya vurugu siku ya uchaguzi” (CUF was holding secret meetings on the Isles which conspired to cause chaos on Election Day) (Media Council of

Tanzania, 2001, p. 14). Without disclosing any evidence, during a presidential campaign rally in

Dar es Salaam, the City’s CCM regional chairman warned CUF members against alleged plans

321 to “disrupt the general elections.” His warnings were followed by what has come to be a ritualistic CCM practice of parading on stage a vocal and critical opposition defector. In this case, former CUF MP, Frank Magoba, informed the crowed that “he was aware” that CUF youths “have been given special training on what to do during the polling day.” He went on to generalize by saying that “CUF has a lot of problems and if we are not aware the opposition party will lead us to chaos and disharmony” (Nyanje, 2005b). Uhuru carried a similar story, claiming that CUF supporters were intending to scare voters by smearing themselves with goat’s blood (Media Council of Tanzania, 2001, p. 14). A Sunday News reporter went to the point of branding the Mkapa critics as “the devil, the architect of misinformation, who knows, through his many agencies, to fabricate lies and create mistrust, bigotry, hatred, envy and misunderstanding” (Bahati, 2005).

Machetes and Murder

These illustrations of the opposition go far beyond accusations of plans to create election- day chaos. The Tanzania Election Monitoring Committee (2001) reported that, in their open air performances, a CCM campaign theatre troupe frequently “ridiculed opposition leaders, depicting them as blood thirsty, incompetent and power hungry bullies while painting a godly and angelic picture of the ruling party’s policies and leaders” (pp. 80-1). In the 2005 election,

Karume repeatedly pointed to CUF as the perpetrator of violence and divisiveness. The opposition he proclaimed, especially CUF “are people who are always grumbling about something and are keen on advocating for acts of violence” (Karume decries, 2005). One paper carried an unfounded headline claiming “CUF yaanza zoezi la kugawa mapanga” (CUF has started handing out machetes) (Tanzania Election Monitoring Committee, 2006, p. 137). In

322

Ilala, Kikwete was especially harsh when stating that “what we are witnessing is that at the end of their campaign rallies, groups of youths storm the streets beating people and looting their property”. He went on by asking the crowd “even before victory they are beating us, so what will happen if they win?” The crowd responded by saying “They are going to kill us” (Nyanje,

2005c). Prior to describing the party’s Development Program for Economic Growth and Poverty

Alleviation at a rally in Soni Village in Bumbuli constituency, Dr. Shein urged people not to vote for those that advocate social division and instead, vote for those dedicated to maintaining peace and stability. In what can only be described as fear mongering, Dr. Shein targeted the opposition and stated “Do not vote for them. They preach hate politics. While people want development, they talk about machetes” (Nyamungumi, 2005a).

The Genocidal Agent

The image of the machete carries with it painful and fearful images of genocide in neighboring countries. However, the rhetoric goes far beyond the mention of machetes. While the CCM itself was largely responsible for initiating multiparty reforms, Party spokespersons maintain a consistently crafted critique against the application of political pluralism to African conditions, a critique that has been carried throughout much of Africa from the early days of independence on up to today. Both theoretically and practically, the design flaw in multiparty politics is understood as a problem of arriving at a national consensus from the sum of competing societal interests, where the nature and interests of the political contestants are conceptualized as challenges to the very core of the nation, national consciousness, and thus emerging as a threat to national security. As the crafted talk goes, such divisiveness can only lead to instability and violence, as evidence in neighboring countries illustrates. Thus, during the

323

2005 election, when reflecting on the aftermath of election violence in the 2004 civic elections,

Mkapa cites that political pluralism is a “dangerous agent” leading to the “moral decay” of the nation (Nyanje, 2005).

But the rhetoric goes beyond citations of moral decay to illustrations of the actual deadly outcome of multiparty politics generally, and an opposition victory specifically. During a campaign speech to introduce Karume as the CCM’s hopeful for the Zanzibar presidency,

Amour stated that “We shall never allow [Zanzibar] to sink into the abyss of violence, whether they [CUF] like it or not.” He goes on to state that, rather than promoting development,

“Multiparty politics has instead introduced a scenario where some individuals concentrate on using abusive, vulgar language and threatening violence…” The article containing these quotes was carried on the front page of the August 25th edition of the African, with huge headlines entitled “Z’bar will not turn into another Rwanda” (Z’bar will not turn, 2000). Norwegian election monitors noted that, during the 2005 campaigns, the CCM generally represented the opposition as “synonymous with disorder and possibility of war” (Mørck, 2006, p. 17). An article in the Express pointed to a clear tendency of the CCM to paint the opposition as parties

“of law violators and hooligans who will turn Tanzania into another Rwanda when given power” (Mwanjisi, 1995). In the 1995 election, numerous reports cited that the CCM used videos of the Rwandan genocide to illustrate the possible outcome of an opposition victory

(Sundet, 1996). For reasons unclear, ITV televised images of the 1994 Rwanda genocide during the 1995 campaign period. A similar screening was done by Abood Television (ATV) in the lead-up to the 2000 election, whose owner claimed that his station was a “no go” area for opposition candidates (TEMCO, 2006, p. 136). In a review of the 2000 election, Iversen (2000)

324 comments that spokespersons for the CCM repeatedly highlighted the atrocities committed in neighboring countries as analogous to what would happen if the opposition wins the election.

CCM Guardianship

The words of the previous pages have pointed to a construction of knowledge which defines the opposition as a threat to the sanctity of the nation and the safety of its people. By contrast, with a proven track record of fighting colonialism and ethnic loyalties, while constructing a sense of national unity, the CCM is offered up to the electorate as the most effective safeguard against violent sub-national and international threats to the nation. Within the macro-structural ideational space of the Tanzanian polity in other words, the CCM occupies the position as the pillar of peaceful national unity and is seemingly inseparable from national symbolism. During election cycles, the CCM employs this ideational capital to its advantage.

Again, this theme is persistently evoked during political campaigns. Following the nomination of Kikwete in 2005, Mkapa repeatedly stressed that the main campaign symbolism for the party should be one of national unity, peace, and tranquility (Kasumuni, 2005a). Mkapa’s recommendations were strongly adhered to by Kikwete, who placed “national unity, peace and tranquility” at the center of his campaign platform (Tanzania Election Monitoring Committee,

2006, p. 66). Out of the top five priorities outlined in his election manifesto, which included economic growth, agricultural development, education, and quality health services, maintaining peace and tranquility topped the list. By contrast, peace and tranquility were never mentioned in the election manifestos announced by the other presidential aspirants, which instead placed fighting corruption and eradicating poverty as first priorities.

325

Generally speaking, these were the platforms presented to the public over the past three elections. In 2000, the leader of the Uhuru Touch Race Jordan Rugimbana, directed voters to vote for the party which stands for peace and development (Media Council of Tanzania, 2001, p.

28). Strengthening the performance of the party was articulated by Mangula as a necessary

“pillar of national unity, the Union and our social-economic development” (CCM moves to enforce, 2005). Shein echoed these same sentiments at a Segerea rally in Illala, where he stated that the CCM was in the best position to manage the country’s political pluralism. For Shein,

“The fourth phase government will ensure that peace prevails and we continue to have tranquility and solidarity, as you all know, we cannot achieve development goals without peace”

(as cited in Nyamungumi, 2005b). At his inaugurating speech in December, 2005, Kikwete placed the maintenance of unity, peace, and tranquility at the very top of agenda list of things to be achieved under his new leadership. During the speech, this incoming President raised the sense of urgency around his leadership by citing that “there is a real possibility that someone might one day undermine our national unity on the pretext of political freedom,” and citing that those “parties associated with violence” which also lost the election is clear evidence that

“Tanzanians treasure” national unity. He moved on to explicitly state that the central task of the fourth phase is to “ensure peace, stability, and unity” as a necessary conduit for economic development and poverty alleviation (Kikwete, 2005).

In short, national consciousness is comparatively strong throughout Tanzania, as is the pride in a history of peace and tranquility and ‘chaos and disorder.’ Against these popular conceptualizations, the ideational superstructure is complemented when the position of the CCM is elevated to that of a paternalistic protector of national order from internal and external threats, including threats from the opposition. What’s left for the remainder of this chapter is an

326 assessment of the degree to which these knowledge structures comport with more recent data on public opinion and perceptions.

Caricature Diffusion

Empirical data sufficient for establishing links between campaign rhetoric and voting patterns are difficult to find and assess. What is needed is broad survey research, something well beyond the capacity of this dissertation. The most available data useful for the pursuit of these links are derived from a number of opinion and polling surveys, statements from experts working closely with elections, interview data, and personal impression during my various trips to Tanzania.

The first necessary piece of data is to understand the breadth of media penetration, by far the largest carrier of political monologues. In terms of newspaper readership and radio listenership, in can be assumed that many of the monologues were heard quite a lot, if not on a daily basis in the months prior to elections. This is especially true for the radio, which is the sole source of information for many living in rural areas. As of 1994, some 49% followed public affair regularly over the radio, while 23 percent read newspapers regularly, and less than 3 percent regularly followed public affairs through the television (Mushi, Mukandala & Baregu, 2001, p.

44-47).

Survey data on the attention paid to political campaigns suggests that, while most pay little attention to campaigns, the bulk of the population probably pays close enough attention to digest the rhetoric propagated by the CCM. As table 12.07 suggests, over half of the population is exposed to campaigning, the largest portion of which is undoubtedly CCM monologues (Mushi,

Mukandala & Baregu, 2001, p. 158).

327

Finally, campaign symbols and slogans can spread like wild fire when politic ideas are exchanged in everyday discussions. Again, while survey data suggests that nearly 40 percent of

Tanzanians never discuss politics, the data presented in table 12.07 shows that over half of the sample at least occasionally discussed politics (Mushi, Mukandala & Baregu, 2001, p. 140). The frequencies presented here probably work to the advantage of the CCM in two ways. First, the frequency within the range between daily and occasion discussions is probably enough for the diffusion of campaign slogans throughout much of the rural population. Secondly, in order for opposition parties to sufficiently challenge the three-decades of socialization under single-party rule, political exchanges will need to be routine parts of day-to-day conversation. This is especially true since, given the discussion in the previous chapter; opposition parties simply lack a clear social base for organizing dissent. However, in light of the high frequency of occasional and no discussion, political exchanges are simply not high enough to sustain opposition messages.

Table 12.07: Measures of Campaign Penetration Follow Public Affairs Discussion of Attention to on Radio Public Affairs National Campaigns Nearly every 42.8% 10.6% Much attention 27.3% day Once a week 6.9% 8.5% Little attention 27.1% Occasionally 31.8% 34.8% No attention 39.6% Never 15.1% 37.2% Other 1.5% Other 0.9% 1.5% Don't know, no answer 4.6% Not known, no 2.4% 6.6% answer Source: Mushi, Mukandala & Baregu, 2001, pp. 46, 140, 158

Data pulled from several surveys indicate that campaign monologues have had a tremendous impact on the social constructions shared by many Tanzanians. In explaining the popular distaste for multiparty politics, and opposition parties in general, Mushi, Mukandala, and Baregu

(2001) reflected in a 1994 survey and conclude that one of the principle causes for the

328 overwhelming skepticism toward opposition parties and political pluralism was the ruling party’s “campaign of disinformation about the implications of multipartyism,” which consistently proclaimed “siasa za vurugu, fujo” – the opposition parties would create chaos (p.

85). Mhina (2001) also points out that “the fact that the majority of Tanzanians are still suspicious of the opposition parties…is mainly a result of three decades of one party conditioning” and has “recently been fuelled by CCM propaganda that the opposition parties stand for chaos, invoking in the process the specter of genocide observed in neighboring

Rwanda and Burundi” (p. 213).

More importantly, this survey asked a few key questions, such as ‘how would you describe those that vote for your party versus those that vote for opposition parties,’ or ‘how would you describe leaders of your party versus those of opposition parties.’ The answers to those questions clearly suggest that the incumbent’s campaign to discredit multiparty politics, and opposition parties, had an impact. Some 48 percent of the respondents used the term “patriotic” to describe those that vote for their own party, versus the 23 percent who used the term to describe those that vote for other parties other than their own. A total of 45 percent applied the same term to the leaders of their own party, while only 21 percent labeled the leaders of other parties as patriotic (Mushi, Mukandala & Baregu, 2001, p. 184).

“Self-serving” was used by 14 percent of the respondents against those who vote for other parties, but only 1 percent used it to describe those that vote for their own party. The term also described the leadership of their own party for 16 percent of the respondents, versus leadership of other parties for 2 percent (Mushi, Mukandala & Baregu, 2001, p. 184). Other terms, which are tabulated in table 12.08 include “traitor,” “imperialist stooge” and “wise.”

329

Furthermore, this survey asked for inputs on the role played by opposition parties and the

CCM in facilitating the transition to multiparty politics. In line with election results which would soon follow, 49 percent of the respondents believed that the CCM was playing a positive role as a facilitator, while only 18.4 percent held the same view for the opposition parties. Of those with favorable views of the CCM, most cited the party’s spirit of cooperation, ability to maintain peace, while avoiding the pitfalls of chaos (Mhina, 2001, pp. 208-209).

Table 12.08: Views of Voters and Leaders People who People who Leaders of Leaders of vote for your vote for your party other parties party other parties Patriotic 48.4% 23.0% 45.0% 21.1% Self‐serving 1.0% 14.3% 1.6% 16.0% Wise people 16.2% 10.0% 19.1% 10.6% Traitors 0.0% 3.8% 0.1% 4.0% Ignorant 0.0% 5.1% 0.0% 3.9% Imperialist stooges 0.2% 6.7% 0.2% 6.9% Other 14.8% 14.2% 14.6% 14.0% Don't know, no answer 19.4% 22.9% 19.4% 23.5% Source: Mushi, Mukandala & Baregu, 2001, 184

A 1999 survey conducted by Research and Education for Democracy in Tanzania (REDET) again appears to find a relationship between the opinions of government performance, opposition performance, and the campaign agenda set by the incumbent party. Among other things, the survey found that 70 percent had a favorable view of the way in which the political leadership of the CCM performs its job, which included managing the political transition and maintaining peace and unity. By contrast, the opposition as a whole was rated most unfavorably.

Some 46 percent approved of the way in which the opposition parties acted as watchdogs and maintained civil peace and unity (Killian, 1999).

These results were again echoed in a 2000 REDET survey, where the belief that the candidate preferred would strengthen peace and harmony was the top specific policy

330 justification6 (Research and Education for Democracy in Tanzania, 2000, p. 4). Similarly, a

2001 REDET survey found that, out of 27 issues considered to be of most interests, 41.5% mentioned “peace and harmony during elections,” the highest ranking of all 277 (Research and

Education for Democracy in Tanzania, 2001, p. 17).

Survey data from Afrobarometer indicates that, while there has been a notable degree of

‘warming-up’ to multiparty rule among Tanzanians, there is still an overwhelmingly powerful tendency to see multiparty rule as a threat to unity. For starters, table 12.09 below indicates that more Tanzanians still prefer one-party rule to multiparty rule. Among the 1,278 Tanzanian respondents, 26.1 percent approved strongly of one-party rule, while another 19.6 percent generally approved. In total, 42.6 percent either disapproved or strongly disapproved of one- party rule, while 45.6 percent either approved or strongly approved.

Table 12.09: Approval of One-Party Rule Tanzania Kenya Zambia Malawi Strongly Disapprove 23.5% 45.9% 50.7% 50.9% Disapprove 19.1% 29.0% 35.2% 4.8% Neither Approve/Disapprove 6.9% 3.1% 4.3% 1.4% Approve 19.6% 12.8% 7.2% 7.5% Strongly Approve 26.1% 6.7% 1.4% 32.6% Don't Know 4.8% 2.5% 1.3% 2.8% n 1278 1278 1200 1200 Source: Afrobarometer, 2005b, c, d, e

Table 12.09 also shows that, when compared to Kenya, Zambia, and Malawi, Tanzanians are far more inclined to register approvals of single-party rule. Only in Malawi, where 32.6 percent of the respondents strongly approved of one-party rule, do we see single-party approval

6 The respondents were asked why they prefer certain candidates over others. The rankings were reported by REDET as follows: personal leadership qualities (23%), belonging to the same party (22%), candidate’s last achievement in leadership (20%), would strengthen peace and harmony (14.7%), candidate’s party policies (11.7%) (Research and Education for Democracy in Tanzania, 2000, p. 4).

7 The top five mention issues were as follows: peace and harmony during elections (41.5%), don’t know (9.0%), good election system (7.9%), good election management (7.4%), respondent’s candidate or party won (5.7%) (Research and Education for Democracy in Tanzania, 2001, p. 17). 331 ratings that exceed those found in Tanzania. Still, in aggregate terms, only 40.1 percent of the

Malawians, 19.5 percent of the Kenyans, and 8.6 percent of the Zambians expressed approval or strong approval for single-party rule, compared to 45.6 percent of the Tanzanians. At the same time, 75.0 percent of the Kenyans, 85.8 percent of the Zambians, and 55.7 percent of the

Malawians disapprove or strongly disapprove of one-party rule, compared to 42.6 percent of the

Tanzanians.

Table 12.10 provides some additional, albeit conflicting elaboration to the data presented in table 12.09. When asked whether political parties are divisive and therefore many parties are not needed, some 30.5 percent of the Tanzanians responded with strong agreement, while 12.7 generally agreed. However, more respondents agreed that many parties were needed to ensure real choice during elections, somewhat contradicting to the findings in table 12.09 above, and certainly a change from the surveys carried out in the early 1990s.

Table 12.10: Political Parties are Divisive vs. Many Parties Needed A=political parties are divisive and therefore many not needed, B=many parties need to ensure real choice Tanzania Kenya Zambia Malawi Agree very strongly with A 30.5 11.8 13.2 27.2 Agree with A 12.7 12.8 18.7 3.0 Agree with B 13.9 32.2 32.3 5.5 Agree very strongly with B 37.3 38.2 31.0 62.8 Agree with neither 2.1 2.9 4.1 0.8 Don't know 3.5 1.9 0.8 0.8 n 1304 1278 1200 1200 Source: Afrobarometer, 2005b, c, d, e

Still, in comparative terms, Tanzanians are clearly more skeptical toward multiparty arrangements. Whereas 24.6 percent of the Kenyans, 31.8 percent of the Zambians, and 30.2 percent of the Malawians either generally or strongly agreed that parties are divisive and therefore many are not needed, 43.2 percent of the Tanzanians responded the same way. At the same time, while 70.4 percent of the Kenyans, 63.3 percent of the Zambians, and 68.3 percent of

332 the Malawians believed that many parties are needed to ensure genuine choice, only 51.2 percent of the Tanzanians responded the same way. Therefore, while Tanzanians do appear less skeptical today than in the past about the divisive nature of multiparty politics, in comparative terms, Tanzanians still have strong reservations against multiparty politics.

Conclusion

Whereas chapters eight through eleven may best be characterized as explaining why CCM dominance has been so durable and opposition so fragmented and lacking in broad support, this chapter might be best understood as addressing how the CCM deploys its advantages in multiparty election campaigns. Given the ensuing conflict in neighboring countries throughout the early 1990s, one of the premier concerns in the lead-up to political reforms and the first multiparty elections in Tanzania was the preservation of that which Tanzanians feel most proud of: national unity, peace, and tranquility. It appeared that throughout Tanzanian society, single- party rule and the CCM were most frequently understood as the best safeguard against national disintegration. Hence, the dominant campaign themes have since revolved around issues of national security.

In the subsequent years, the CCM has managed to tactfully craft its campaign in a way that keeps national security as the defacto focal point of election campaigns. While talk of corruption and poverty are of considerable importance, campaign imagery which ties genocide and terrorism to both sub-national and foreign interests, the same interests which back an inexperienced array of opposition parties, are bound to undermine any chance of significant opposition inroads into national politics.

333

CHAPTER THIRTEEN TANZANIA IN COMPARATIVE PERSPECTIVE

Multiparty reform was the fatal blow for the tenure of most long-standing incumbents throughout Africa. Benin’s 1990 reforms, followed by the 1991 election, swept away President

Kérékou and the sixteen year rule of the Parti de la Révolution Populaire du Bénin (PRPB). In

Zambia, Kenneth Kaunda and the United National Independence Party (UNIP) suffered a similar fate following the 1991 reforms. Parallel developments occurred in Guinea Bissau, Cape

Verde, and Malawi. By 2002 in Kenya, the opposition parties that managed to hand Moi and

KANU victories in the 1992 and 1997 multiparty elections due to their personalized division popularized through ethnic appeals, managed to finally come together to unseat the incumbent party.

By contrast, multiparty reform in Tanzania appears to have resulted in little more than some tactical changes on the part of the CCM, but has done little to encourage an upwelling of thoroughly organized opposition parties with ties to sizable popular bases. As of the 2005 multiparty election, the CCM managed to collect more votes than the party garnered in the 1995 and 2000 elections, and remains one of the few cases in Africa where party dominance is paired with a comparatively favorable record on civil liberties, political rights and stability.

Argument in Review

From the outset of this study, I observed that long-time incumbents presiding over single- party African states were faced with a dilemma when international and domestic demands converged over the need to reform existing single-party rule to allow for multiparty elections.

On the one hand, the costs imposed by the suspension of aid and loans due to failures to reform would be borne by existing patron-client relations. As a class, the bureaucratic bourgeoisie’s ability to accumulate is highly dependent on the state’s location as a mediator between international and domestic financial transactions. The suspension of aid would certainly have been a blow to the cohesion within this class and the legitimacy of the ruling party as an attractive election machine for reproducing class hegemony. At the same time, the state as a transaction mediator permitted clientele networks to expand beyond the state to capture loyalties from other social forces. In short, the suspension of aid would mean that clientele networks would run short of capital and threaten the social cohesion which guaranteed incumbency. This was a particularly acute problem in conditions where high levels of macro-structural divergence already jeopardized clientelistic control and presented entrepreneurial politicians with a viable opportunity to exit the ruling party and press for political change.

On the other hand, the risks to continued incumbency by allowing political entrepreneurs to openly establish opposition parties, to cultivate support bases, and to contest elections would be a direct challenge to prospects for future incumbency. This was an acute problem for regimes already facing intense domestic reform pressures. Regimes which chose the former path – i.e. block reforms altogether – ultimately raised domestic demands for reform. To be sure, Moi attempted to go down this route, but was faced with even stiffer opposition for doing so. For regimes which choose the latter, most were superficial reformers of sorts, which permitted multiparty elections but hedging against defeat through widespread coercion, intimidation, and election manipulation. In the majority of cases, reforming regimes either barely survived a first multiparty election, or were unseated altogether.

Compared to the scenarios above, the case of Tanzania is clearly an outlier. While international pressures for reform were present, domestic pressures for dismantling the one-

335 party state remained halting. Despite the limited pressures, and the fact that survey data indicated that the majority of the Tanzanians appeared to approve continued single-party rule by the CCM, with Nyerere’s blessing the National Executive Committee decided to go multiparty anyway. While reform was certainly inspired by international pressures, there is no doubt that much of the inspiration behind reform stemmed from a concern that, if reform was not forthcoming, opposition elements would draw on international events and discourses opposed to

“non-democratic” rule as a rallying cry for building political support (Mmuya & Chaligha,

1992; Sundet, 1996). This problem would have been greatly compounded had the international community suspended aid as penalty for a lack of reform. As Nyerere so courageously remarked as early as 1990, “When you see your neighbor being shaved, wet your head to avoid a dry shave…” (as cited in Baregu, 1994, p 222). In short, the CCM reformed without domestic pressures for change. In what some have described as a “guided” or “managed” transition, the ruling elite in Tanzania effectively denied the opportunity for nascent opposition leaders the chance to mobilize against single-party rule (Kiondo, 1995; Whitehead, 2000).

The assumption that the CCM’s central leadership “pre-empted” the opposition however, cannot really deal with the question as to why the domestic demands for multiparty reform and opposition to the CCM was so comparatively weak in the first place. The question then logically shifts to one that asks why mobilization against the incumbent and the rules which sustained it was so weak prior to reforms. Do these causes have any bearing on the paltry performance of opposition parties after reform? If so, are there reasons to believe that these causal forces are in a state of flux which could undermine the future of CCM hegemony?

The short answer to these questions is that, as an outgrowth of the necessity to mobilize support for colonial legislative elections and ultimately independence, TANU’s formative years

336 never sowed the seeds of discord to the degree found in neighboring countries by being associated with any particular ethnic group. Following independence, policies of universalism, egalitarianism, and state-led development had the consequence of undermining the growth of substantial levels of structural divergence by fostering a strong sense of national identity, while weakening the growth of social forces with the economic and ideational basis for mobilizing popular and sustained opposition challenges. In short, decades of African Socialism denied the ability of political entrepreneurs to mobilize resources around identities with the intensity for demanding political reform and opposing the CCM once reform was in place.

More specifically, as elaborated in chapter five’s description of early multiparty elections, the peripheral nature of the colonial economy in Tanganyika generated the conditions whereby social cleavage structures in throughout the colony at the time of independence in the early

1960s were far too fragmented and amicable to act as a basis for mobilizing major challenges to single-party consolidation. Stated differently, ideational polarization and the the lack of highly capitalized class structures presented political elites with no real viable opportunities for launching their own party organization out of the hope of collecting seats in Legislative Council or sitting at the helm of the independence ship. The polity on the Mainland was not defined by the interests of well endowed petty bourgeoisie leaders nor highly polarized sub-national identities which could have come together for challenging the occasion of TANU consolidation by what were largely leftist urban party ideologues and civil servants with tacit backing from the

Britain and other western powers. This structural climate had two effects. First, political party consolidation did not involve a vertical integration of ethnic identities in relation to access to state power. Secondly, the role of conservative rural interests took a back seat to those of the

337 rising party ideologues and eventually the bureaucratic bourgeoisie. Both of these effects radically departed from the typical African case.

Uninhibited by the conservative force of a powerful petty bourgeoisie popularized by ethnic narratives, by 1967, Ujamaa emerged as the overarching ideological framework for political and economic development. While failing to curb the economy’s problems of dependence, and opening the door for the growth of a technocratic bureaucratic bourgeoisie, Ujamaa yielded policies that asserted state control over the commanding heights of the economy, fostered a strong sense of universal national identity, and reduced inegalitarian political and economic trends, especially those inequalities which would have cultivated ethnic sentiment.

By the middle years of the 1980s, Tanzania, like nearly every other African country, was steeped in an economic crisis. This crisis had two effects. First, it prompted a dramatic change in the ideological programs pushed by international donors and lenders, which increasingly favored economic liberalism. Secondly, it delegitimated the project of the party ideologues, who were increasingly on the sidelines as the new policy environment favored the rise of the more reform-minded technocratic elements of the bureaucratic bourgeoisie.

As the 1980s gave way to the 1990s, international reform pressures shifted toward the political realm. What made Tanzania exceptional was that, rather than being faced with rising reform pressures from within, the mainstay for reform came from international actors. Why was this the case? The answer to this question is straight forward. For starters, the policies pursued during single-party rule subdued capital accumulation outside the state, making it difficult for political entrepreneurs to mobilize resources for political party building. However, even with the ability to mobilize resources, the popular basis for opposition support would have been significantly. The basis for popular contention was undermined by Ujamaa’s success at fostering

338 a real sense of economic equality across regional and ethnic groups through universal education and access to healthcare, and imposing a national consciousness which cut across nearly every section of society. In short, while single-party rule could no longer be defended against the political reform pressures originating largely from international actors, the CCM, now dominated by a bureaucratic bourgeoisie less preoccupied by adherence to ideology and more worried about class reproduction, could engage in the reform process at a relatively low risk to its own hegemony.

The legacies of economic ownership by the state, weak ethnic allegiances, and, by African standards, a strong sense of national identity, each established during the days of single-party rule, continues to constitute enormous assets for the CCM during its bid to stave off election contenders. With the ability to exert influence on the allocation of resources, and the Party’s image as founder and guardian of a nation surrounded by disorder, chaos, and, opportunism opposition parties simply cannot find the material and moral space for contesting the CCM’s well-crafted political campaigns. Hence, not only did African socialism provide the macro- structural causative forces for weakening the domestic push for multiparty reform, but also act as a macro-structural environment which renders political competition at the micro-level highly advantageous for the CCM.

At this point, it is important to repeat a warning that has been stated at several points in this piece, namely the material and ideational variables cannot be treated as discrete from one another, but as mutually reinforcing conditions. For example, to ask what would have happened if we hold things like political and economic equality constant (lower probability of discontent), but give political entrepreneurs access to lucrative financial resources for building party organization, is a question that is disjointed from reality. Increasing the level of resources

339 available to political entrepreneurs would likely have impacts on the overall level of inequality as well, and thus, according to the model, increasing the probability of popular discontent. In short, the policies of Ujamaa, and their structural consequences must be treated holistically.

The CCM in Comparative Perspective

Throughout this piece, Malawi, Kenya, and Zambia were routinely discussed as contrasts to the case of Tanzania. While each case differs significantly from one another, the common threads that run between them include the following: First, the consolidation of single-party regimes took place amid societies with powerful petty bourgeoisie classes able to draw upon ethnic narratives as a basis for mobilization. Hence, in order to stitch together clientelistic control centered around the executive, ethnic calculus constituted a basis for ordering access to the state. Secondly governance following initial consolidation nurtured the development of tensions and contradictions which made opposition mobilization a relatively easy affair. The growth of class-based sentiments among labor and commercial classes are most notable in

Zambia and Malawi. All three cases were characterized by the reproduction of discord through ethnic favoritism, uneven economic development across regions, and coercion as the backbone for maintaining law and order. This was especially true in the years leading up to reform, as opposition elements were able to draw from an environment rife with discord against single- party rule and lucrative nodes of ideational and material resources which could be readily mobilized against the ruling party.

No other cross country case exhibits the level of contrast better than Kenya. The organization of political parties in competition over colonial legislative council seats sowed the seeds of discord that would bring havoc to KANU’s incumbency maintenance project. Ethnic

340 tensions at onset of independence became the principle concern of Kenyatta after independence.

Throughout the Kenyatta years dominance was ensured by a mixture of coercion and ethnic balancing through patronage, which, among other things, accentuated wealth differentials between ethnic groups.

Following the death of Jomo Kenyatta in 1978, Moi ascended to the Presidency and proceeded to reverse previous patterns of inclusion and exclusion. Whereas Kenyatta’s policies disproportionately benefited a class of Kikuyu petty capitalists, throughout the 1980s Moi began to dismantle the economic and political standing of the Kikuyu in favor of the smaller and poorer groups, most notably the Kalenjin. One by one, Kikuyu politicians, including Kenneth

Matiba and Mwai Kibaki, found themselves harassed, coerced, and eventually shut out of the corridors of political power.

However, these regime outsiders were not without resources for challenging Moi, resources which were endowed with additional leverage once international pressures for reform mounted.

The language of ethnicity provided a readymade ideational infrastructure for articulating demands for multiparty change, grievances against KANU, and a common moral thread for conditioning political party loyalties in various popular arenas. What’s more, the cultivation of lucrative petty and commercial bourgeoisie sections meant that political entrepreneurs would find themselves with the material wherewithal for constructing sustainable opposition organizations. In short, the legacies of single-party rule in Kenya are found in the production of a macro-structural environment divergent enough to fertilize the growth of domestic demands for multiparty change and the consolidation of a highly competitive multiparty environment following multiparty reform.

341

A similar process unfolded in Malawi, albeit one where the ruling party, the Malawi

People’s Congress (MCP), was far more monolithic. Here, as the number of party’s branch offices and members expanded rapidly during the late 1950s, the initial organizational developments of the MCP looked somewhat like TANU’s during the same time period. But, the similarities really stop there. For starters, the ethnic milieu in Malawi at the time was far more consolidated than that found on the Tanzanian Mainland. Decades of selected missionary education and colonial modes of domination assisted in the growth and regional consolidation of

Tumbuka, Chewa, and Yao identities (Vail & White, 1989; Kaspin, 1995). Secondly, rather than seeking to transform those structural and institutional implications of colonial rule, Banda instead simply focused on reforming them. As the MCP under Banda’s leadership radically centralized social control, the Chiefs were retained as a mechanism for mobilizing rural support for Banda’s policies (Kaspin, 1995; Kaunda, 1998). At the same time, Banda, strove to solidify the popularity of his rule in the Central Region among the numerically dominant Chewa. To this end, the capital was moved to Lilongwe from Zomba in the south, the Chichewa language promoted as the Malawian national tongue, and Chewa iconography promoted as the symbol of the nation (Kaspin, 1995).

Perhaps the most significant was the development of elite cohesion through a sort of “quasi- state capitalism” through politically supported large-scale agricultural enterprises (Thomas,

1975, p. 33), which ultimately wedded the petty bourgeoisie to the state. Through preferential bank credit and low taxes, the Banda’s regime encouraged well connected political leaders to rapidly expand estate farms, while, through high taxes, simultaneously discouraged smallholder production and encouraged the growth of a sizable low-paid labor force (Kaunda, 1998).

342

At the onset of international pressures for political reform, the configuration of social forces in Malawi had the wherewithal for building and sustaining opposition to single-party rule. The salience of ethnicity left enough ideational capital for mobilizing political support along regional lines, while economic capital formation in the South gave the United Democratic Front (UDF) an enormous financial advantage. Banda responded to rising contestation by jailing opposition leaders, most notably Chakufwa Chihana, which in turn became a rallying point for broadening popular opposition against Banda and the MCP (van Donge, 1995).

As in Kenya, the legacies of single-party rule in Malawi are represented by a macro- structural environment divergent enough to sow the seeds of domestic demands for multiparty change and highly competitive multiparty politics following reform. Unlike Kenya, where

KANU managed to win the first two multiparty elections thanks to a divided opposition, the

MPC and Banda in Malawi lost the first multiparty election to the UDF and Bakili Muluzi.

Out of all three contrasting cases thus far, the complications of single-party consolidation, the inability to fend off reform demands, and the loss of tenure in the first post-reform multiparty election, were most acute in Zambia. Similar to Nyerere in Tanzania, Kaunda’s

Humanism contained ideals of egalitarianism, universalism, and aspirations of nation-building which attempted to move beyond the trappings of local allegiances. Unlike Nyerere however, at the time of independence Kaunda was far more constrained by Zambia’s more divergent social milieu. He faced a stiff resistance from a collection of social forces, some of which existed in the lead-up to independence and some which were incidentally empowered in the years afterwards.

From the very get-go, Kaunda was challenged by rival independence leader Harry

Nkumbula and the African National Congress (ANC), who strove to build the popular base of

343 the ANC through appeals to Tonga identity as a counterweight to the perceived Bemba affiliations of UNIP (Sichone, 1996). Regional rivalries persisted after independence, which emanated primarily from the ANC, UP, and UPP, each of which appeared as an ethnically divided petty bourgeoisie challenge to party and state consolidation. Even after the emergence of single-party rule in 1972, challenges to consolidation persisted, as local UNIP leaders governed exclusively in order to fend off potentially more popular rivals who were once members of the opposition. Hence, local-level party development suffered as local party leaders attempted to restrict access to the policy-making arena and clientelism from contending notables

(Gertzel, 1984).

Organized labor, most prominently the Mine Workers Union of Zambia, was an important ally for UNIP during independence and shortly thereafter. By the 1970s, this support begins to weaken, as shown by UNIP’s loss of votes in the Copperbelt during the 1973 election. In response, UNIP leaders tried to sideline labor. But by 1980, the Zambia Congress of Trade

Unions (ZCTU) was able to demonstrate its ability to organize mass strikes and protests against the state of the economy, increased wealth gaps, and mismanagement in the hands of UNIP leaders

Finally, business interests grew more powerful during the 1970s, as most of the people who successfully contested elections had backgrounds in business (Momba, 2003). By 1975, the business community had the capacity to resist elements of Kaunda’s ‘Watershed Speech’ to the

National Council1, which outlined his move toward Humanism, Party Supremacy, and the termination of private enterprise by 1980. Had Kaunda been able to present his speech to a social milieu similar to that which received Nyerere’s Arusha Declaration, UNIP may have been

1 The National Council met twice a year, and consisted of all regional party officials, district governors, Central Committee members, the ZCTU president, members of mass organizations, senior civil servants, and members of parliament (Momba, 2003).

344 able to reproduce macro-structural conditions similar to those produced in Tanzania. However,

Kaunda was far less fortunate. As suggested by Gertzel (1984), because of his inability to rely on local elites, which had their own ambitions and agendas, Kaunda, like Nyerere, had to rely on centrally appointed officials to circumvent local elites. However, unlike Tanzania, the center consisted of the “petty bourgeois and business class elements” which did not necessarily see

Humanism in a favorable light and were primarily concerned with “upward mobility and access to office” (p. 105).

In Tanzania, Nyerere and TANU leaders had the advantage of governing over an environment where the tug and pull of diverse social forces was far weaker than commitments among leaders to national unity and African Socialism. The situation for Kaunda was different.

Given the pull of such a diverse array of identities and interests within UNIP, and the fact that none of the participants really bought into the vagueness of Humanism, Kaunda’s ability to gain compliance rested solely on patronage and the selective use of coercion. Kaunda’s dominance held so long as patronage flowed (Sichone & Simutanyi, 1996). But, by the end of the 1980s, cohesion within UNIP rapidly disintegrated due to the continued deterioration of the economy, decline in the flows of patronage, and the rise of food riots and strikes (Momba, 2003). Political entrepreneurs found themselves in the midst of a near perfect environment for challenging single-party rule and a Kaunda dominated UNIP. The momentum of mass protests, a readymade political institution in the ZCTU, and a business community grown tired of economic mismanagement, became the mixture for effecting political reform, fracturing UNIP, and ousting Kaunda from office in the first multiparty election (Momba, 2003).

The post-reform multiparty environment in Zambia remained highly competitive, with contestation revolving around tensions between those social forces that single-party rule was

345 unable to subdue. Eight political parties defined the Zambian political landscape between 1991 and 19962, many of which were built up following defections from the MMD following the first multiparty election. While a few were one-man shows, a few others were grounded in the rhetoric of ethnicity and regionalism (Simutanyi, 2005), most notably the National Party3.

Election results clearly indicate that the polity has grown even more competitive since the election of 1996, especially between the MMD and UPND.

So far, the comparative case summaries have included only contrasting cases, whereby multiparty reforms resulted in the demise of incumbency. But, what about other cases where opposition parties across several multiparty elections failed to unseat ruling parties? Stated differently, how do levels of macro-structural divergence impact regime tenure in other dominant-party systems?

In order to explore this question, I will now turn my attention to a pair of other noteworthy cases of party dominance in Africa, one of which has already been widely discussed. The first is a case drawn from the contrast been political developments on Tanzania Mainland versus

Zanzibar. Given the major election defeats of the CCM on Pemba across the three most recent multiparty elections, the Isle’s history defined by overtly racialized class conflict, and the comparatively high level of brutality as a cornerstone of dominance following independence,

Zanzibar appears to be the exception to nearly everything that has been said in this piece about

Tanzania. This has required the periodic assertion of adjectives like “Mainland” Tanzania, or a resort to the proper-noun “Tanganyika” as an attempt to show that those patterns which have

2 These eight included the MMD and UNIP, as well as the Agenda for Zambia (AZ), Movement for Democratic Process (MDP), National Party (NP), National Lima Party (NLP), United Democratic Party (UDP), and Zambia Democratic Congress (ZDC) (Simutanyi, 2005, p. 12).

3 The National Party, similar to the United Party which was outlawed in 1968, draws support from the Western Province and the Lozi

346 been attributed to Tanzania generally, are actually reserved to only one, albeit the largest portion of Tanzania.

Still, Zanzibar is of interest not only as a disclaimer to general statements, but as a supportive illustration about some general claims put forth throughout my work. First, explaining the degrees of success and failure of party dominance calls for a historical investigation of the exchanges between social forces, party consolidation, and government policy, as the causal components for understanding the scale of party dominance. Secondly, an important route to prolonged party dominance depends on a form of party consolidation devoid of polarization and discord, and the continued reproduction of conditions which provide for a sort of social convergence around the legitimacy of the party as an integral component of the political system. In contrast to dominance on the Mainland, single-party dominance on Zanzibar has included none of this, and, as a result, the CCM’s position on the Isles is highly tenuous today.

Indeed, the ASP’s route to power was fundamentally different from TANU’s path. Whereas

TANU in Tanganyika grew as an economic and political network between normally amicable constellations of social forces, pre-independence political parties on the Isles were outgrowths of exceedingly abrasive social divisions largely between African plantation workers, Arab plantation owners, and Shirazi yeoman farmers mostly confined to Pemba. As political competition tightened in the lead-up to independence, Arab landlords began evicting African tenants and destroying their crops, while Africans boycotted Arab-owned businesses. Afro-

Shirazi Party supporters were increasingly being dismissed from jobs in the dockyards and as household servants (Maliyamkono & Kanyongolo, 2003). Rhetorically hostile and “dirty” election campaigns and razor-thin election victories, as identified in the final portion of chapter

347 five, characterized the exchanges between the ASP, the Zanzibar Nationalist Party (ZNP), and the Zanzibar and Pemba Peoples Party (ZPPP) during the pre-independence elections (Ayang,

1970; Mmuya & Chaligha, 1994; Othman, 2006). As one observer comments in the lead-up to independence:

Irreconcilable differences between them [the political parties] have led to a legacy of bitterness and intransigence, and there is a great danger that force and violence, endemic during the era of nationalism, may be institutionalized as the sole mechanism for the resolution if conflict in the post independence period (as cited in Maliyamkono & Kanyongolo, 2003, p. 141).

Such a statement could hardly have been applied to politics in Tanganyika

If the political exchanges prior to independence sowed the seeds of discord into an emerging party system, patterns of conflict resolution after independence only helped to transpose the conflict to the present-day time (Maliyamkono & Kanyongolo, 2003; Othman,

1990, 2006). For starters, the first independence government, one composed in December of

1963 by an Arab/Shirazi coalition, lasted barely a month. Encouraged by TANU and other foreign powers, Karume and ASP came to power following a reign of “terror” and a violent and bloody revolutionary overthrow of the first independence government of Abdulrahman

Mohamed Babu. As one survivor commented on the situation:

They rose and murdered all the Arabs they could catch, slitting their throats by night as they slept. Some of us escaped to dhows but my parents were chased along the beach by a crowd – though my mother was African. They ran into the sea to swim to the boats, but some of the crown followed, caught hold of their hair and drowned them (as cited in Clayton, 1981, p. 78).

Shortly following the Revolution, the Revolutionary Council, under pressure to neutralize the more radical elements of the Revolution, “hurriedly signed” the Articles of the Union (Shivji,

2006, p. 89).

348

Despite the fact that government officials claim to have attempted to redress the social divisions on the Isles, it is relatively clear that governance under the ASP, and CCM following the 1977 merger between the ASP and TANU, had as much to do with vindication as with alleviating conflict (Mapuri, 1996; Maliyamkono & Kanyongolo, 2003). For starters, governance under the ASP, especially under Karume, relied heavily on silencing dissent through the use of the police and military. Indeed, the ASP’s style of governance contained elements of paranoia, which prescribed whatever means necessary to ensure that the Revolution which overthrew the Arab aristocracy would dominate for years to come. From vantage point of ASP leaders, nothing akin to Nyerere’s inclusive and rhetorical style was permissible in an environment plague with such a violent history of class and racial conflict.

Secondly, Pembans were sidelined economically and politically following the Revolution.

During the era of single-party rule no Pemban could reasonably aspire to a leadership position within the CCM or the government. This discrimination continues today. Similarly, Pemba was excluded from infrastructure and industrial development, and generally suffered from the lack of social services (Kweka, 1997). At the same time, clove exports, the bulk of which comes from

Pemba, constituted a wealth of income for government revenues. In short, the state was collecting revenues from Pemba, but Pemba was getting nothing in return.

Finally, while TANU had enjoyed close relations with the ASP even before independence, this relationship would grow tighter in the years after independence, culminating in the merger between the two parties to form the CCM. The motivation behind the cooperation and eventual merger was driven by a desire to weaken dissent by effective diluting some of the renegade elements within the ASP and to clean up some of the problems with corruption on the Isles.

However, the cost of the merger, and the loss of Zanzibar autonomy specifically, only served to

349 expand the scope of conflict from one that was primarily isolated to the Isles, to one that had profound implications for national and Mainland politics.

Suppressed conflict on the Isles took on a new life following the 1992 adoption of a multiparty constitution. As Othman noted back in 1995 in reference to the emerging political situation on the Isles, “I had hoped that 30 years after the Revolution, a certain cohesion had developed, but I have been proved wrong; we have gone back 30 years as far as ethnic tensions are concerned” (as cited in Maliyamkono & Kanyongolo, 2003, p. 143). As if to mock

Othman’s remarks, the exchanges between CUF and the CCM today resemble those seen between the ASP and ZNP, with the added legalistic and religious components. Furthermore, the level of multiparty competition on the Isles is far more intense that what is found on the

Mainland. In fact, it is likely that without the CCM’s periodic resort to coercion, intimidation, and election manipulation on Zanzibar, the CCM’s tenure on the Isles would be a thing of the past.

In summary, the Zanzibar case highlights three important facts relevant to the argument of this work. First, levels of macro-structural divergence are indeed positively related to levels of political contestation at the onset of multiparty reform, as seen by the enormous electoral support for CUF among certain sections of Zanzibar society. Secondly, policies during single- party eras that attempt to politically and economically exclude sections of the population only produces and reproduces social polarization. As seen in the exchanges between CUF and the

CCM, the consequence of polarization can be quite violent and deadly. Finally, in order to continue to dominate where divergence is high, regimes will be required to continue a policy of exclusion and repression. There are a host of other African countries, namely Cameroon and

Togo, which offer up similar comparative statements to those presented by Zanzibar.

350

That said, one other case deserves some attention here. To the south of the copper rich country of Zambia, in the landlocked and sparsely populated country of Botswana, the

Botswana Democratic Party (BDP) has governed continuously since independence in 1966. The

BDP has done so while having a parliamentary multiparty democracy with a respectable track record on observance of civil and political liberties and rights. Therefore, the BDP offers up a potentially useful comparison to the CCM’s dominance in Tanzania by virtue that presently both have comparatively limited levels of coercion and election rigging, while winning decisive and sustained election victories under multiparty elections.

If the Botswana case tells us anything, it is that there is indeed no single recipe for party dominance. I say this simply because the formative processes giving rise to dominance by the

BDP on one hand and TANU on the other, along with the development pursuits of each regime following independence, could hardly have been more different. Each case does however, elicit one common them: during the terminal years of colonial rule each country was among the poorest and least developed in the world, and thus generally escaped the patterns of divisiveness sown by heated tensions between rural conservative forces versus the more ambitious urban ones.

To be sure, social dislocations from colonial rule were comparatively minimal in Botswana.

The British established Bechuanaland Protectorate as a mere defensive barrier for repelling

German and South African invasions, and thus had limited economic ambitions in the

Protectorate. In what Du Toit (1999) referred to as “benign neglect,” colonialism resulted in a

“minimal dislocation of authority structures” and minimal degrees of social polarization (1999, p. 196). For Picard (1987) the term “structural underdevelopment” rather than benign neglect defined Bechuanaland’s integration into the colonial system: as a peripheral component of a

351 regional capitalist economy, serving largely as a labor reserve for mines and farms in the vastly more developed South Africa. Applying either concept, benign neglect or structural underdevelopment, the Protectorate was of certain peripheral, although not unimportant, concern to the British colonial calculus, similar to the placement of Tanganyika. For example, in the early 1960s, Bechuanaland could claim to have no more than 170 primary schools, 3 junior secondary and 2 senior secondary, 1 teacher training college. In 1965, only one year prior to independence, there were barely 16 students in Bechuanaland prepared for college-level training

(Samatar, 1999, p. 64). These figures are strikingly similar to those obtained in Tanganyika during the same period.

A second commonality, no doubt causally derived from their respective peripheral statuses under British colonialism, was relative ethnic passivity and low level of social polarization generally (Du Toit, 1999). While the Tswana constitutes a majority, historically speaking, the

Tswana majority maintained a peaceful coexistence with smaller groups. In pre-colonial times, the territory soon to become Bechuanaland, was composed of relatively hierarchical and centralized morafes (nations), each composed by a network of villages, wards (Kgotla), families, and households bound together by bonds of kinship, religion, and clientele exchanges (Samatar,

1999). In colonial times, the morafe were more or less integrated into a colonial administration focused on repelling invasions from outsiders and extracting taxes for financing its excursion.

The British had little interest in transforming the economy of Bechuanaland and therefore, never engaged in the sort of land confiscation found in Kenya or mining operations found in Zambia which would have resulted in a pool of proletariat farm workers, tenet farmers, and landless peasants. Hence, while pre-colonial and colonial histories between Tanganyika and

Bechuanaland were fundamentally different in some respects, at the onset of independence,

352 social forces in both societies were relatively amicable. In turn, the emerging nationalist struggles were not as “antagonistic” as elsewhere in Africa (Molutsi, 1991, p. 5), establishing a climate favorable to weakly contested single-party consolidation.

Similarities however, stop there. Whereas chiefs in Tanganyika were in a weak position for spearheading challenges to TANU’s consolidation, similarly conservative elements in

Bechuanaland were powerful mediators between the peasants and the colonial state, and considered themselves to be the rightful heirs to the outgoing British officials (Molutsi, 1991).

Chiefs were, to be sure, a powerful social force with the potential for blocking BDP policy, or even launching a political party grounded in a coalition of ethnic identities. The question now becomes: why was there not a significant challenge to the BDP‘s consolidation at the time of independence? First, whereas TANU was led primarily by urban elements with little personal grounding in rural Tanganyika, the BDP was led by Seretse Khama, the heir to the chieftainship of the colony’s largest ethnic group (Colcough & McCarthy, 1980), the Bamangwato. Seretse himself represented the younger, less conservative elements of Ngwato society, and, in 1949, actually defeated a more conservative chief in a popular vote in the Kgotla (Samatar, 1999).

This meant that Seretse was the defacto leader for about 25 percent of people living in

Bechuanaland Protectorate and, from the perspective of the outgoing British, a moderate leader worthy of support. Given the history of amicable social relations in the colony, an atmosphere of trust and conciliation prevailed, making it easier for chiefs to see the BDP and Seretse in less threatening and more trusting terms.

Secondly, and perhaps the most critical, was the timing of the introduction of popular elections. In Kenya and Tanzania, popularly elected candidates to colonial legislative councils began in the late 1950s, the logic of which called for the formation of political parties (Maxon,

353

1994; MacDonald, 1966). In Botswana, a non-elected Protectorate Legislative Council was formed in 1960, but no major national election contestation took place until the National

Assembly election of 1965. Given the absence of popular elections prior to independence, there was no real incentive for political elites, most notably the chiefs, to ambitiously expand their political base in an attempt to collect votes. Had the onset of elections taken place earlier, as in the four other cases, competition in Botswana at independence would certainly have contained more polarizing references to ethnicity as political entrepreneurs attempted to solidify popular backings.

Despite the decisive victory for the BDP in 1965, the institutional position of the chiefs was a force seen by Seretse as essential for ensuring rural social control after independence.

Therefore, the chiefs were placated rather than eliminated, a move which in turn continued to guarantee that opposition parties would be weak. As Proctor noted in 1968, “each modern democratic institution [in Botswana] which was brought into existence to replace a chiefly function incorporated the chief as an ex-officio member or even chairman” (p. 66). Most notably, chiefs were given paid advisory positions in a newly constituted House of Chiefs, which, among others, was composed of eight chiefs from Botswana’s eight principle ethnic groups (Picard, 1987).

What emerged from this relatively smooth road to independence was a parliamentary multiparty democracy dominated by a somewhat conservative BDP with mutually beneficial, yet non-politicized ties elites in the countryside, no doubt contributing to the BDP’s ability to govern in a style that receives favorable evaluations on respect for civil liberties and political rights. Recall that, according to Freedom House data reported in figures 4.01 through 4.04, and

354

U.S. State Department data reported in tables 4.05 and 4.06, Botswana ranks as among the least likely countries in the developing world to commit politically related human rights violations.

Furthermore, the petty-bourgeoisie coalition behind the BDP oriented its development policies toward a pro-western, capitalist development model which proved to be highly successful for sustaining remarkably levels of macro-economic growth. Directed by the Ministry of Finance and Development (MFDP), akin to Japan’s MITI, the backbone of the economy included cattle export, copper and nickel mining, and after the mid-1970s, diamond exports.

However, while macro-economic development was successful, it nevertheless resulted in wide income asymmetries. Tables 9.02 and 9.03 previously illustrated the comparative degree of inequality, where Botswana ranks as among the most unequal, whereas Tanzania ranks as among the most equal.

Despite the growth of enormous wealth differentials, from the first National Assembly election in 1965 to the most recent 2004 election, the BDP has governed continuously, winning wide election margins along the way and successfully resolving three leadership successions4.

If there is an optimal model for reconciling international demands for multiparty elections with an interest in maintaining hegemony, the BDP is probably the closest approximation. The longevity of the BDP’s tenure has been ensured through considerable cohesion among the

Country’s elite, paired with a relatively non-mobilized, if not passive population. Let’s look at each in turn.

While having ties into different morafes and kgotla, the “coalition of wealthy, well- educated, cattle-owning elites” which emerged as the reigning economic class at independence shared similar interests and visions toward economic development. Because the BDP placated

4 The percentage of votes received by the BDP between 1979 and 2004 are as follows: 1979 – 75.4%, 1984 – 68.0%, 1989 – 64.8%, 1994 – 53.1%, 1999 – 57.2%, 2004 – 51.7% (African Elections Database, 2007a).

355 elites, rather than excluding them, the chiefs, along with administrators and bourgeoisie classes, had a vested interest in co-operating to defend market oriented development strategies from the growth of populist demands (Samatar, 1999). Rapid economic growth during the 1970s and

1980s5, and stable growth throughout the 1990s6 ensured that this inter-elite equilibrium was not disturbed.

This inter-elite bargain, which, as cited earlier, entailed a mutual recognition on the need to prevent the growth of populism, ensured that Botswana’s population today remains relatively passive. As identified earlier, since popular elections were introduced late in the protectorate, the tendencies for political mobilization in competition over representation were comparatively absent. Therefore, the notion of ‘mass-based’ parties never developed nor was the ruling party a mass-based national liberation movement to the degree found elsewhere (Samatar, 1999). Given the fact that elites could agree on the need for economic growth, and that economic growth actually persisted, there was simply no incentive for politicians to upset this equilibrium by mobilizing a popular base7. Additionally, the accommodation of the chiefs by the BDP political

5 According to data from the 2007 IMF World Economic Outlook, between 1980 and 1989, Botswana’s economy more than doubled, whereas Kenya’s grew at 16 percent, Malawi’s at 23 percent, Zambia’s at 3 percent, while Tanzania’s shrank some 5 percent. During the same time frame, Botswana’s economy only shrank two periods (1982 and 1985), followed by double digit growth from 1986 through 1989. By contrast, Kenya’s economy shrank three consecutive years (1981 through 1983) plus two others (1985 and 1989), Malawi’s shrank four periods (1982, 1984, 1985, and 1987), Zambia’s shrank five consecutive years (1982 through 1986), while Tanzania’s contracted four periods (1984, 1985, 1987, and 1989).

6 According to data from the 2007 IMF World Economic Outlook, comparatively strong growth persisted throughout the 1990s, albeit far more modestly. Between 1990 and 1999, Botswana’s economy grew at some 22 percent, compared to 6 percent for Kenya, and 3 percent for Malawi. Remarkably, Tanzania’s economy more than doubled during the 1990s (largely between 1995 and 1998) while Zambia’s shrank some 16 percent. GDP per capita growth for Botswana between 1990 and 1999 soared at 66.2 percent compared to 41.5 percent for Malawi, 19.4 percent for Tanzania, 13.5 percent for Kenya, and -13.4 percent for Zambia.

7 This reality is partly captured in voter turnout data. Election turnout rates for Botswana are as follows: 1969 – 54.7%, 1974 – 31.2%, 1979 – 55.2%, 1984 – 77.6%, 1989 – 68.2%, 1994 – 76.6%, 1999 – 77.1%, and 2004 – 76.2%. All figures are the number voters as a percentage of the registered population (International Institute for Democracy and Electoral Assistance, 2007).

356 leaders effectively ensured that the regime had access to mechanisms for ensuring local social control, which so many other African countries lacked.

The development path set out by the BDP has not however, unfolded absent of dialectical developments which continue to whittle away BDP votes today. To be sure, the elections since

1989 have grown more competitive, with the percentages of votes gradually shifting away from the BDP toward a divided array of opposition parties, the net benefactor being the Botswana

National Front (BNF). In the 2004 election, the BDP scored less than 51.7 percent of the votes, the poorest the party has ever performed in an election.

BDP rivalry is not a new phenomenon. At the time of independence, the main BDP contender was the urban-based, somewhat leftist Botswana Peoples Party (BPP), which represented the ambitious drive to eradicate chiefly authority, and the noble objections to

Botswana’s close economic ties to South Africa. Not surprisingly, Tanzania’s counterpart to the

BPP was none other than TANU. However, considering the relative power of the chiefs in the

Bechuanaland Protectorate, the consolidation of power by these urban intellectuals would not see the level of success had by TANU, which, quite simply faced no serious contestation from similarly conservative social forces.

Today however, this urban-rural divide is becoming more distinct largely because rapid economic growth also delivered the growth of working class and younger middle-class professionals (Molomo, 1991; Sokhulu, 2004). The realization of working and professional classes as a social force was clearly evident in the groundbreaking election of 1989, where the

BNF8 secured 26.9 percent of the votes, and the election of 1994, where the party increased its vote share to 37.7 percent. As of 1998, the position of the BNF was weakened due to a factional

8 As an observer member of the Socialist International, the BNF represents a political force with leftist ambitions, championing the causes of the Country’s poor.

357 split, resulting in the formation of the Botswana Congress Party (BCP). If however, one combines the BNF and BCP9 vote totals since the split, the increasing weakness of the BDP is once again revealed. In the 1999 election, the combined vote total was 37.9 percent, while that number increased to 42.7 percent in the 2004 election.

At the same time, the weakening of BDP performance has contributed to the growth of a tense electoral environment. For certain, blatant intimidation has not been entirely absent from

Botswana’s elections, especially when competition is stiff and the hegemony of the BDP is questioned. Opposition speakers are at times, "heckled" by authorities, while voters are told of dire consequences for voting for the opposition (Picard, 1987, p. 153). More recent attempts have been made to strengthen the state’s media coverage of the BDP elections when the

Minister of Information and Broadcasting directed the state media to only cover the BDP candidates for the 2004 election. In response to a serious dispute over election rolls during the

1999 elections, a state of emergency was declared for the first time in Botswana (Sokhulu,

2004). In 2004, President Mogae threatened to dissolve the parliament if it did not support his choice for presidential succession (Sokhulu, 2004).

To summarize, while the CCM and the BDP have managed to dominate their respective polities even under multiparty rules, the historical casual chains spanning from the time of independence to contemporary elections are quite distinct for each case and distinct from most other processes throughout the African Continent. Both cases share one common thread: as peripherally incorporated into British colonial rule, social dislocations from economic development and political upheavals were relatively minimal in each case, thus minimizing the rise of social tensions around questions over control of the post-independence state. From this

9 The ideological ambitions of the BCP are fundamentally the same as those of the BNF. The factional split with the BNF that gave rise to the formation of the BNC was more to do with personality rivalries rather than policy differences.

358 point however, patterns are highly divergent. In Tanzania, the overall structural weakness of rural social forces meant that the hegemony of a largely urban-based bureaucratic bourgeoisie and leftist ideologues was not difficult to establish and consolidate. But, as Hyden’s seminal

1980 piece argues, TANU’s weak links to rural power brokers also meant that extracting local compliance in development policies was notably difficult. By contrast, in Botswana, chiefs were well positioned to challenge the dominance of largely rural, but progressive petty bourgeoisie at independence, but lacked the interest to do so thanks to the ability to build trust and cooperation around specific sets of goals.

The paths taken to maintain conditions for hegemony were equally divergent. In Tanzania, opposition parties were outright proscribed for nearly thirty-years, by which the ambitions of

Ujamaa, while often coercive and alienative, also mobilized the population around the concept of familyhood, nationalized all major economic points with the potential for capitalization, and leveled economic and social benefits across major sections of society. At the onset of multiparty change, the only social force to truly consolidate its position since independence was the party- state nexus. The possibility of Mainland social cleavages acting as a significant casual force was simply far too remote due to the lack of capital and polarization. By contrast, in Botswana, the reproduction of hegemony never relied on a formal proscription of opposition parties. Instead, the BDP was built out of an alliance of vested interests that tied the elites of nearly every major social force to the goal of capitalist development and sustained by the delivery of development and a de-mobilize population.

359

Generalizing to Non-African Cases

Gazing beyond the unique political histories found on the African Continent, one finds an ample number of cases of party dominance, along with an ample body of research to go along with each case. One particularly important question needs attention before this comparative study can be rightfully brought to a close. What elements from the case of party dominance in

Tanzania can be generalized to other cases outside the African Continent? In an attempt to tackle this question, I will look at a simplified sketch of each variable drawn from the Tanzanian case, and apply this sketch to other cases of party dominance.

As the analysis has held thus far, a regime’s formative era is a likely source for the most important causal variables for establishing an environment conducive to prolonged party dominance over politics. For starters, the rise of party dominance generally takes place amid some sort of epic struggle, which can only be fought through some sort of epic cohesion

(Simkins, 1999). The definition of the stakes of the battle, that is to say self-rule vs. rule by outsiders, freedom vs. enslavement, progress vs. regress, or order vs. chaos, are ideas with the power to move millions of people into political action against enemies defined with a similar sense of urgency, i.e. thieves, exploiters, reactionaries, or murderers. As Scott (1995) work suggests, political climates having been defined as a life and death struggle against hostile enemies elicits positive impacts on the longevity a political organization, even when the initial struggle which gave rise to the organization has long since past (p. 53).

An obvious example where a political party emerged out of an epic struggle is the case of the Partido Revolucionario Institucional (PRI) in Mexico. Against the backdrop of violent struggles between elite cliques, including military generals, agrarian and labor groups, and local political bosses, the PRI was born as a remedy to social turmoil. This somewhat Hobbesian state

360 of affairs, which ensued during the 1920s as the fallout of the Mexican Civil War, was finally subdued in the 1930s through a grand coalition led by General Calles, and consolidated with a popular peasant base as a left-leaning political party under Lázaro Cárdenas (McDonald & Ruhl,

1989). In Taiwan, a similar sense of urgency over the prospects of an invasion by Maoist mainland forces was the principle impetus for the rise of the Kuomintang (KMT) in 1949 as a formal political party. Reified as a quazi-Leninist party of sorts, the cohesion within the KMT was maintained out of the dire need for build a new nation and to protect against these hostile

Chinese forces on the Mainland (Chu, 2001).

More common still, epic struggles include the prospects of ending colonial rule, as experienced by so many countries following the decades after the Second World War, or more recently, by countries released from domination after the fall of the Soviet Union. The People’s

Action Party (PAP) in Singapore emerged to contest colonial elections and dominated the political scene in the lead-up to independence from Britain in 1963. In Malaysia, the United

Malays National Organization (UMNO) was formed in 1946 as a Malay opposition to the

British backed Malayan Union and economic dominance of the Chinese. Later, UMNO formed an alliance with the Malayan Chinese Association (MCA) and the Malayan Indian Congress

(MIC) to contest colonial elections against Malayan Communist Party (MCP) and the Pan-

Malayan Islamic Party (PAS) and eventually emerge as the main player of Independence from

Britain (Esman, 1987).

The presence of an epic struggle is by itself an unlikely condition for ensuring party dominance simply because societies are typically composed by an array of social forces which may differ over what the epic struggle is about and the policies which the emergent regime should embody. As the Tanzania and Botswana cases illustrates, diversity alone is not a

361 problem, so long as relations between society’s component parts have been historically amicable enough to remain together under a broad front or coalition. The crux of the casual argument is that epic movements that seeking to include societies most powerful social forces stand a much better chance of emerging in a position conducive party dominance than those movements which end up advancing the causes of particular sections of society while isolating and excluding challenging social forces.

For certain, as described above, the emergence of the broad-based coalition undergirding the PRI in Mexico fits this description. Likewise, in the multiracial, multilingual, and multi- religious Singapore, PAP, which initially took on a strong leftist agenda, eventually broadened its base among elites after adopting a more pragmatic approach to democratic socialism (Chang,

1968; Mutalib, 2000).

The situation in Malaysia looked somewhat more complicated, given the fact that during

British colonial rule in Malaysia, the urban Chinese acted as the petty and commercial bourgeoisie intermediaries between British capital and Malay peasants (Esman, 1987, p. 402).

At the time of independence, the elite factions of the Chinese and Malay communities were each in a numerical or economic position to make credible threats when trying to piece together a workable coalition. The Malay constituted a majority, which could be readily mobilized as a powerful popular political force by UMNO if need be. At the same time, the blackmail potential of the Chinese was based in the fact that they controlled most of the Country’s wealth. However, deals were struck at the time of independence without having to mobilize popular bases or divert economic resources, which would have certainly polarized the political landscape. The British backed bargain of 1957, as a sort of elite pact between the more affluent urban Chinese and the largely poor but majority Malay, ensured a privileged political position for the Malay in

362 exchange for a free-enterprise system which ensured the continuation of Chinese economic power (Giliomee & Simkins, 1999, p. 13).

Just as the presence of an epic moment does not determine the longevity of party dominance, neither does the breadth of the movement’s base. Unity under conditions of an epic moment is a different process, with different institutions and adversaries than unity under stable government, where the actual participants are forced to confront one another over day-to-day governance issues.

In the broadest possible sense, there are two variables relevant for to reproduction of party- dominance once the initial rise to power has passed. The first is obviously a prolonged period of time. As Pempel (1990) specifies:

The longer the party is in power, the greater the opportunity it has to use state resources to shape and reshape its following. In addition, the longer a party remains in power the more compelling the pressures for social groups, even those initially hostile to the party, to accommodate to its seemingly unshakable control (p. 7).

The second variable, one most evident in the Tanzania and Botswana cases, is the ability and willingness of the regime govern in a way that pragmatically manages society’s existing patterns of divergence. More specifically, the likelihood of prolonged party dominance is increased 1) where states are empowered with the ability to exert considerable influence on the allocation of resources, 2) where resources are allocated in a way that does not accentuate social divisions and attempts to alleviate existing tensions, and 3) where major political battles are continually defined through the lens of the epic struggle which brought the regime to power in the first place. Again, taking a cue from Pempel (1990), a party which is able to satisfy all three will reinforce and strengthen its own power base, producing a “virtuous cycle” of party dominance without the need to constantly resort to the coercive appendage of the state (p. 15).

363

In Singapore, PAP has to some extent, satisfied all three. The Party had governed for nearly

40 years, largely through a form of “inclusionary corporatism” which provided the regime with the network to society’s most powerful social forces as a channel for exerting social control and for responding to societal demands (Mutalib, 2000, p. 316). The cohesiveness of this arrangement was reproduced through an economic agenda focused on phenomenally high growth and social equity (Chang, 1968; Asher, 1994; Mutalib, 2000). Through his “dynamic and charismatic leadership” style, Prime Minister Lee Kuan Yew was able to build a common thread among diversity by creating a strong sense of national identity through “cultural, political and economic transformation” (Chang, 1968, p. 761).

In Taiwan, the KMT was able to fundamentally stabilize a social setting that could have otherwise been highly divisive, especially between immigrants from Mainland China and those indigenous Taiwanese which resided on the island prior to the late 1940s. As described by Yun- han Chu (2001), the KMT maintained a “stable political order through an elaborate ideology akin to socialism, a cohesive and highly penetrating apparatus organized along Leninist democratic centralist lines, and a powerful and pervasive but less visible security apparatus…” coupled with an “inclusive” recruitment policy and a “complex web of party-run or party- invested enterprises” (p. 269).

As for the longevity of PRI’s dominance in Mexico, a coalition initially stitched together through patronage and broader attempts to address issues of social justice (McDonald & Ruhl,

1989), the some seventy-year period of near uncontested hegemony was based on broad

“corporatist consensus” among labor, peasants, military, and professionals, where corporate sectors maintained control over members, while state exerted control over sector elites

(Giliomee & Simkins, 1999, p. 10). Elite cohesion was buttressed by the relatively inclusive

364 sexennial presidential system, whereby roughly 80 percent of the administrative posts, including the presidency, changes hands every six years. Thus, each clientele network (camarilla) remained loyal to the PRI, anticipating to one day be sitting in Los Pinos (McDonald & Ruhl,

1989; Weldon, 1997; Langston, 2002).

The PRI was also widely, although not universally popular due to the party’s persistence in defining political battles as part of the Revolution and itself as the mother of national order. The epic character of PRI’s rule has been the primary basis for maintaining a loyal following among the masses, especially the peasants (Story, 1986; Collier & Collier, 1991; Langston, 2002), who, at times, even appropriated revolutionary discourses to attack state failures (Aitken, 1996).

Indeed, Lázaro Cárdenas, whose presidency delivered on a highly popular land reform program during the 1930s, is still somewhat of a popular hero among peasants (Klesner, 1997).

However, a number of contradictions surfaced that would ultimately weaken the PRI’s electoral performance, First, the regime never really addressed income distribution, which tended to benefit corporatist insiders at the expense of outsiders, leading to widening income gaps and isolated but potentially destabilizing outbreaks of discontent (Giliomee & Simkins,

1999, p. 32). By the late 1980s, following a decade of economic crisis which was met with fiscal austerity (Pastor & Wise, 2005), the sizable left-wing Corriente Democrático (CD) faction of

PRI, broke away to form the Frente Democrático Nacional (FDN) after PRI selected a técnico named Carlos Salinas as the Party’s presidential candidate for 1988. The era of shaky PRI dominance set in as Salinas won by a mere 51 percent of the votes (McDonald & Ruhl, 1989;

Langston, 2002), while the electoral landscape underwent an increasingly intense level of competition between the Partido Acción Nacional (PAN) and PRI in the more developed

365 northern states, and the Partido de la Revolución Democrática (PRD) and PRI in the less developed southern states (Klesner, 2005).

Secondly, the growth of the commercial sector, partially as a result of the economic reforms of the 1980s, also meant that PRIs monopolistic control through existing corporatist channels was increasingly compromised. Hence, rising commercial interests eventually became the lynchpin for the slow ascendance of PAN throughout the more urbanized northern sections of the country (Story, 1986; McDonald & Ruhl, 1989; Klesner, 1997). Finally, in the 2000 election, PAN’s Vincente Fox won with some 42.5 percent of the vote compared to 36.1 percent for PRI’s candidate Francisco Labastida.

Despite its election losses over the past two decades, PRI remains a major electoral force, especially in state and local elections where the party still maintains a strong machine like presence. Even in the 2006 presidential election, PRI secured well over 20 percent of the vote.

This is an interesting comparison to the dismal performance of UNIP and KANU in the years following their first election defeats, suggesting that there is some merit to Giliomee and

Simkins’s (1999) assertion that not only does inclusiveness increase prospects of prolonged dominance under multiparty elections, but also helps ensure that the dominant party is still a relevant political actor once regime turnover does occur (p. 11). The restoration of the KMT in

Taiwan following the 2008 elections also supports this claim.

In Malaysia, the stability brought on through the aforementioned inter-racial bargain that was struck at independence would not last largely because the Malay masses were excluded from the economic benefits of the deal. Finally, in the lead-up to the 1969 election, the economic subordination of the Malay, paired with politically opportunistic rhetoric from political parties, translated into widespread racial violence (Esman, 1987) that left some 196

366 people dead, 149 more wounded, and around 6,000 Kuala Lumpur residents homeless (Hwang,

2003). UMNO’s response to the crisis was shore up cohesion among elites, and to stem the growth of Malay unrest which could have easily spelled doom for UMNO leadership. The government’s response was to first construct the Barisan Nasional (BN), as a broad-based coalition between Malay, Chinese, and Indian parties, along with some twelve others following the race riots of 1969. The BN was a somewhat successful attempt to shore up community elites by recommitting its pledge to allow inclusive access to the policy arena (Balasubramaniam,

2006).

Secondly, through a program called New Economic Policy (NEP), UMNO leaders pushed to “eliminate the identification of race with economic function” by enforcing a rapid increase in the Malay share of national wealth over a twenty-year period (Esman, 1987, p. 403). As a consequence, economic and social policy since 1969 has greatly favored the Malay, including giving Malay preferential treatment in public employment, college admission, as well as a privileging the status of Malay language, leading to the growth of a business class whose success is linked to UMNO’s hegemonic position within the BN (Case, 2004).

Finally, throughout the 1970s and 1980s, the BN coalition was highly successful in positioning itself as the engine for national unity, harmony, stability, and economic growth

(Mauzy, 1983) and as an “ideal vehicle” for achieving development amid interests based on divisive ethnic considerations (Balasubramaniam, 2006, p. 24).

However, the policies that flowed out of the 1969 racial tensions have also proven to be some of the most pressing challenges facing the BN generally, and UMNO specifically. Most critically has been an increased marginalization rural Malay, who have not benefited from NEP to the same degree as those living in more urban localities. States like Selangor, Negri Sembilan,

367

Johore, and the island of Penang have disproportionately benefited from foreign investment, infrastructure development, and high wage jobs, while Kedah, Kelantan, Trengganu, Perlis,

Pahang, and Sabah have been left behind (Balasubramaniam, 2006). This urban/rural bifurcation coincides with an increased level of polarization, accompanied by stronger support for the

Islamic Party of Malaysia (PAS) among the older, more traditional Malay left behind in the rural areas. This intra-Malay cleavage also coincides with a rift within the Chinese community and the stronger support for Democratic Action Party (DAP) among those Chinese who accuse

MCA as selling out to UMNO (Balasubramaniam, 2006).

What these comparative cases manage to illustrate is the not so surprising statement that there are indeed, multiple routes to the emergence of party dominance. In some cases, regimes come to power amid highly divided societies on the brink of the breakdown of internal social order, as the case of Mexico and Malaysia clearly illustrate. In other cases, parties come to power where social dislocations and animosities are comparatively low, making it easier for the emergent regime to consolidate its initial gains. Clearly, the Botswana and Tanzania are examples of the latter scenario. The rise of the KMT in Taiwan and PAP in Singapore appear to occupy mid-points between these two scenarios.

Other differences emerge at the conclusion of the initial consolidation phase. In some cases, quazi capitalist development paths are pursued as a method for producing positive sum gains for placating the interests of potentially divergent factional forces. This was undoubtedly the illustrated in each case, save for Tanzania. As the case of Tanzania goes, a utopian socialist development path was pursued which effectively undermined the growth of social forces with the material capacity or ideational space for forming an oppositionist challenging the CCM. In fact, I would argue that this utopian option was never really available to any of the other cases,

368 despite the tendencies of some of these regimes tout a leftist ideology. Quite clearly, ambitious utopian socialist development trajectories would be seen as a threat from social forces with bourgeois ambitions. Therefore, where well endowed bourgeoisie classes persist during the regime consolidation, any attempt at implementing utopian socialist ambitions would undoubtedly result in violent confrontations. Subduing these social forces would require the escalated use of the state’s coercive arm, which, as already demonstrated time and again, is contradictory to the prospects of party dominance after the onset of multiparty reforms.

However, each case of prolonged party dominance discussed in this chapter contains two common themes. In each case, the potentially centrifugal pull from elites is not arrested through repeated attempts at eradicating them, but from a long-term willingness of the regime to bring society’s most powerful political and economic players into the policy-making arena. Long-term dominance requires leaders from all social forces to recognize the value in engaging in accommodating practices and sharing in the benefits of economic development and political representation. Secondly, the most favorable route for achieving electoral victories in an international climate demanding respect for political rights and civil liberties is to define the stakes of the political contest as issues of national security or advances in revolutionary ambitions. Only at the ideational stage can a regime truly hope to captivate the imagination of its audience, to elicit a level of fear not toward the regime, but toward a polity without it. Only by crafting the regime as the pillar of that which the nation holds dear can the regime truly hope to generate the popular solidarity to defend its tenure. In the words of March and Olsen (1995), solidarity such as this provides the basis “for mobilizing citizens to defend the political system and a frame within which internal conflicts can be resolved” (p. 51).

369

CHAPTER FOURTEEN CHALLENGES AND IMPLICATIONS

“Everything was forever, until it was no more.” Alexei Yurchak, 2005

The sheer breadth and depth of the Chama cha Mapinduzi’s political dominance has been remarkably resilient over the past three multiparty elections. In fact, the incumbent’s position looks far more certain following the 2005 election than it did during the 1995 election. Looking ahead, the outcome of the 2010 election is not likely to betray the patterns of the past. Jakaya

Mrisho Kikwete, who will certainly vie for a second presidential term, will likely face only token rivalry within the CCM. Likewise, barring a defection of a high profile CCM leader, no politician outside the party looks to be in a position to seriously contest his candidacy. Freeman

Mbowe, while engaging in a respectably professional campaign in 2005, is not likely to see his base expand much beyond the northern portions of the country. Since the late 1990s, Augustine

Mrema has been perpetually involved in seemingly endless petty court battles over party property and leadership disputes. As a consequence, he has been widely discredited as a political opportunist and has virtually no hope of making a political comeback strong enough to be seen as a serious threat. Finally, Ibrahim Lipumba, CUF’s presidential aspirant, while capturing 11.7 percent of the votes in the 2005 election, is hopelessly bound to the politics of the Isles. In fact,

CUF’s real battle lies not with the Union Presidency, but the Zanzibar Presidency.

The landscape in 2015 might however, look fundamentally different. For starters, the politics of the incumbent machine will be steeped in the fueds of an inevitable internal power struggle between those that vie to replace Kikwete. Furthermore, changes in patterns of

370 accumulation domestically, and power relations internationally, will undoubtedly impact the

CCM’s survival into the next decade. The goal of this final chapter is two-fold. First, I will discuss some of those domestic and international factors which could both enhance and undermine the CCM’s dominance in the years to come. Secondly, I will discuss some broader practical and theoretical implications of my work.

New Challenges for the CCM

Aside from the Union question, which has been a consistently bitter theme across each multiparty election, there are three general trends which stand to intensify macro-structural divergence: 1) Visible increases in wealth inequality, 2) the slow but persistent growth of commercial and urban professional classes whose wealth accumulation is less clearly tied to the regime when compared to that of the bureaucratic bourgeoisie, and 3) marked tendencies toward religious and racial appeals are some of the most notable sign posts for the growth challenging of material and ideational structures. The consequences that these changes might have on CCM dominance in the next decade should not be taken lightly. Increased divergence will translate into greater potentiality in finding the salient messages, slogans, and rationale, along with the financial backing for formulating dissenting discourses and organizing parties with the capacity to sustain that dissent. Let us look at each sign post in turn.

Since the trade liberalization reforms of 1982, and IMF Economic Recovery Programs of

1986, 1989, and 1992, which demanded greater fiscal discipline, currency devaluations, and cost accounting measures in the provision of social services, the CCM’s socialist claims looked increasingly disingenuous. Similarly, the National Investment Act of 1990, which established the Investment Promotion Center as the organizational motor for attracting private foreign direct

371 investment, and the Banking and Financial Institutions Act of 1991, which gave the green light for the establishment of private banks, increasingly ensured that private sector development would be the primary engine for economic growth.

At the same time, internal CCM reforms would, on one hand, undermine the fundamental basis for bureaucratic bourgeoisie class reproduction, these reforms would, on the other hand, also provide a venue for state and party elites to draw on resources from an increasingly privatized economy. Prior to 1992, the party relied on most of its funding from government subventions. The reality of multiparty reforms required the CCM to sever this tie. In a brief moment, through a constitutional amendment, the financial benefits accrued by the CCM’s relationship with the state, were wiped away. The employment of thousands of party workers was terminated, while many regional and district workers were going without pay (Kapinga,

1995).

As the party-state nexus was loosening, another landmark move effectively meant that the

Leadership Code would be repealed. The 1992 Zanzibar Declaration was, in reality, the formal death blow to the CCM’s ability to rightfully lay claim to the socialist mantra. Now, party and state officials were free to engage in capitalist behavior by buying equities, owning multiple rental properties, or participating in business ventures. This effectively meant that the bureaucratic bourgeoisie looked increasingly like a wealthy commercial class, with an enormously large pool of personal wealth and amenities. Similarly, the Zanzibar Declaration blew the door open for the CCM to draw resources from this emerging commercial class as a compensation for the elimination of the state subvention (Okema, 1996).

What do these developments mean? For starters, from the standpoint of the organizational adaptation, it is an example of the vitality and flexibility of the ruling party to be adapted to fit

372 changing structural circumstances. Where multiparty reforms were an inevitable threat to the financial capacity of the organization, through promises of the selective enforcement of tax codes and licenses, ambitious party leaders managed to successfully reposition the CCM’s network to connect to a rising class of wealth business elites.

There is however, from the standpoint of incumbency, and social justice generally, there is a downside. Economic reforms and increased private sector capitalization also translates into the growth of a narrow class of wealthy individuals, including public officials, to a degree never tolerated during the days of single-party rule. In the period between 1999, my first trip to

Tanzania, and 2006, my most recent trip, the number of extravagant suburban houses have clearly increased in tandem with the number of poor people living in and around certain sections of Dar es Salaam. While increased wealth inequalities is, by itself insufficient as a causal basis for the growth of sustained political contention, when paired with political entrepreneurs with the wherewithal for framing dissent and disseminating claims, rising wealth inequalities can be a hotbed for the simultaneous rise of contention. Signs of this discontent are clearly visible from the multitude of newspapers sprinkled with letters and editorials commenting on the extravagant lifestyles, multiple mansions and rental properties, and amenities enjoyed by some of the higher order public officials and business elites.

In the past, as stressed time and time again in the foregoing chapters, party ideologues went to great lengths to denounce wealth accumulation as a symptom of selfishness, greed and unjust inequality, practices having no place in Tanzanian society. While many Tanzanians today regard

Ujamaa as a failure, these same Tanzanians are not necessarily embracing the modality of capitalism, nor those economic tendencies that come with it. The ideals of Nyerere still have

373 powerful resonance when defined loosely, especially among peasants and middle to older aged citizens.

During a friendly chat back in 2005 with Professor Rwekaza Mukandala, now Dean of the

University of Dar es Salaam, we discussed a number of the prominent campaign themes surrounding Tanzania’s third multiparty election. One theme which appeared to be a constant point of reference, at least within the CCM’s internal vetting process, was each candidate’s strident attempt to stake out a unique position as the rightful bearer to Nyerere’s ideals. During the general election, Kikwete repeatedly attempted to imbue his campaign with the symbols of

Nyerere, as a man of the people. In a frank statement, Mukandala confidently said “party factions revolve around claims over Nyerere. The candidate which successfully lays claim to be an embodiment of Nyerere’s ideals will win the nomination and possibly the election.” Given the reality of growing wealth inequalities, especially in a context where the overwhelming majority of the citizens are poor, the viability of this campaign tactic for CCM insiders may be drawing to a close.

The fissures of discontent over wealth inequality were clearly evident as early as the first multiparty election, where Mrema managed to pose a respectable challenge to the CCM. While for many, Mrema appeared as a boisterous and somewhat brash populist, his sometimes careful attention toward criticizing the CCM on one hand, while, at times, paying tribute to the aspirations of Nyerere, was a mixture with a real potential for upsetting the significance of the

CCM party brand. This was especially true in reference to Mrema’s anti-corruption platform, which sought to depose the credibility of the CCM by citing how the bureaucratic bourgeoisie were failing to live up to Nyerere’s ideals by getting rich while the rest of the country only grew

374 poorer. For Mrema, the bureaucratic bourgeoisie was the culprit and thus the single biggest obstacle to eradicating the country’s three enemies: poverty, ignorance, and disease1.

In the end, Mrema’s efforts failed to develop into a sustained opposition challenge. The salient campaign imagery of the time was national unity, peace, and tranquility amid political change, to which the CCM could credibly lay claim to. To a large extent, this imagery dominated the 2000 and 2005 elections as well. However, given rising inequalities and the visible ability of politicians to amass enormous pools of wealth, any politician that successfully manages to position him or herself as the rightful heir to the ideals of Nyerere, while discrediting the CCM as trading in Nyerere for the ability to get rich, is certain to pose a formidable challenge to the ruling party’s tenure in the future. In the lead-up to the 2005 election, the CCM Central Committee was well aware of this possibility and therefore threw their weight behind Kikwete’s more populist candidacy over some older party stalwarts. As

Mukandala pointed out in our aforementioned discussion, had Kikwete not secured the Party’s nomination, he would have likely bolted from the CCM to stake out a position as an opposition candidate. Given his confidence, his repeated messages of unity and order, and his popularly perceived embodiment of all those ideals that Nyerere stood for, Kikwete’s campaign contained the perfect mix of ideational symbolism and rhetoric for winning the presidency and bringing the era of the uninterrupted CCM dominance to a close.

Mrema’s 1995 failure to win the presidency was, without a doubt, a product of a macro- structural environment that offered limited opportunities for mobilizing economic and ideational resources for competing against the CCM campaign machine and brand. However, this too is

1 In might be recalled that the “poverty, ignorance and disease” were declared at the time of independence to be the nation’s three enemies. The war was repeatedly used as a basis for accentuating platforms during single-party elections and cultivating support for Ujamaa. Even today, politicians frequently attempt to garner support by citing their commitments to eradicating these three enemies.

375 changing. Liberalization, having resulted in growing income inequalities, has also spawned the growth of a commercial class with the capacity shaping the competitiveness of the political landscape. While the political orientations of these business elites have tended to steer in the direction of the CCM, undoubtedly shaped by lucrative clientele networks which return favorable tax rates and licensing treatment, they are at the same time a class whose mode of accumulation is radically different than, and often time as odds with, the bureaucratic bourgeoisie. An electoral alliance between the urban and rural poor and rising commercial class, say between parties like the pro-business CHADEMA and the more populist TLP, cemented by a leader with popular clout among the sizable young to middle-aged population, would indeed constitute a formidable opponent for the CCM.

In fact, such an alliance would perhaps describe the scenario where Kikwete defects from the ruling party. In contrast the campaign practices of the past, where the CCM National

Executive Committee effectively managed campaign resources, the 2005 election cycle was unique in the sense that Kikwete ran his own parallel campaign virtually independent of the

CCM’s campaign. While Kikwete’s candidacy benefited from an influx of CCM financial capital and was able to draw upon the symbolic value of the CCM brand, Kikwete simultaneously ran his own campaign out from an entire floor of the Luxurious Millennium

Tower Hotel. He had his own source of funds, which undoubtedly drew from lucrative connections to a rising class of business owners. He had his own staff of highly trained campaign advisors who helped him design a campaign with a winning message. In 1995, a man named Augustine Mrema had a similar degree of popular appeal as Kikwete did in 2005. When

Mrema defected however, the ability to mobilize resources for staffing his campaign with technocratic experts and occupying an entire floor of a luxurious hotel was completely out of the

376 question. Today however, the equation looks different. Had Kikwete defected, not on would he have been able to credibly mesh his campaign imagery with salient ideational structures, he would have also been able to lay claim to lucrative financial support for giving his campaign the level of organizational sustenance that Mrema never had.

The CCM’s vulnerabilities do not end here. Questions over the structure and arrangements of the Union, (i.e. two or three-tiered design and nature and degree of Zanzibar autonomy) are legal challenges fundamental to the Union Constitution, and thus have the potential to open the door for a full blown Constitutional debate. For the CCM, this could also mean the introduction of other issues seen as more central to CCM hegemony, such as prodigious executive authority in law-making and appointing and the partiality of the courts. While serious constitutional reform might not be enough for lowering election certainty in the short-run, over the long-term, it might weaken the ability of the CCM to utilize state machinery as a tool for weakening opposition parties.

At the same time, by blocking debate over the Union question in the legal arena, as the government and party leaders continue to do, the supporters of reform take on a sometimes maximalist stance, i.e. quietly calling for the dissolution of the Union itself. This development is compounded by the fact that, because tensions over the Union coincide with demographic differences, and because political entrepreneurs persistently attempt to find a mobilizing discourse, religious and racial appeals are increasingly found defining the parameters of political conflict. While this trend might not be a threat to CCM incumbency, it is a possible threat to social stability.

Despite these rising vulnerabilities to the CCM’s future tenure, over the past decade or so several international structural conditions have shifted, raising the possibility that otherwise

377 legitimate opposition action may once again be proscribed. First, at the discursive level, the use of references to “terrorism” since the attacks on the U.S. in 2001 has become dangerously ambiguous, allowing for a broader range of contention to qualify as terrorism. In tandem with the definitional ambiguity, international institutions have constructed a host of enforcement measures to ensure that countries pay closer attention to combating terrorism. The United

Nations Counter-Terrorism Committee (CTC) was established in 2001 to ensure that member states took appropriate measures to combat terrorism, while the CTC Executive Directorate was created to oversee technical assistance to member states. While the Organization for African

Unity (OAU) passed the Convention on the Prevention and Combating of Terrorism in 1999, since 2001 the OAU’s successor, the African Union (AU), has recommitted itself to combating terrorism (Whitaker, 2007).

Dubbing opposition leaders as terrorists and then applying the appropriate legal measures to proscribe their political activities is nothing new in Africa and elsewhere. It would seem that, in the hasty rush to respond to the terrorist attacks against the U.S., policy-makers in Washington and elsewhere forgot about the tendencies of terrorism to be used as a pretext for combating legitimate dissent or policy-makers simply ignored prior concerns about good governance and respect for human rights altogether. Fighting terrorism has, for decades, been widely used as legal and moral venues for repressing not only groups that exert violence on civilian populations for political gain, but also to repress otherwise legitimate opposition. The difference now is that

Western countries, primarily the United States, are somewhat less inclined to see the targeting of legitimate oppositionists as a violation of civil liberties and political rights, most notably in places like Indonesia, Malaysia, Uzbekistan, and Djibouti (Carothers, 2004). Countries that cooperate in broader anti-terrorism efforts are likely to get a pass on domestic repression, and

378 even benefit by counter-terrorism training, surveillance, and weaponry (Yom & Al-Momani,

2008). In reality, so long as oppositionists are successfully branded as terrorists, from the perspective of international anti-terrorist efforts, there is nothing to over look anyway.

Indeed, despite sizable protests from the country’s Muslim community and NGOs,

Tanzanian politicians have shown relatively clear ambitions toward complying with these international anti-terrorism pressures. The government has enhanced the legal framework for combating terrorism, most notably The Prevention of Terrorism Act of 2002 and the Prevention of Money Laundering Act of 2006. The government established a Counter Terrorism Unit and plans to establish a Counter Terrorism Center in the near future (Kagwanja, 2006; Whitaker,

2007).

If anti-terrorism measures are used by the government to stifle opposition, it will undoubtedly be directed toward CUF and the politics of the Isles. To be sure, CUF has already cited the poor definition of terrorism as an attempt by the CCM to “monitor, detain, and threaten its opponents” (Ehrenkranz, 2006). At demonstrated in chapter twelve, during the contentious times of elections, CUF leaders are routinely accused of having ties with terrorist organizations in the Middle-East and are typically depicted as murderous villains bent on bringing chaos and disorder to the country.

A second change with the potential for shoring-up the CCM’s capacity to regenerate tenure is the increasing connectedness between the United Republic and China. As highlighted by several high-level exchanges in the early 1990s, China resurgence of interest in Africa has been shaped by the country’s rapid economic growth and a correspondingly more complex energy, trade, and geopolitical concerns (Muekalia, 2004). This resurgence has culminated in escalating volumes of merchandise trade, development assistance, foreign direct investment, and contract

379 labor agreements between Tanzania and China (Wang, 2007). At the same time, the rising spatial reach of the Chinese economy relative to that of long-standing European and American economies offers up potentially new opportunities for regimes to act more authoritarian ways.

For starters, in aid and development relations with the developing world, China has shown a greater inclination to partner with regimes without exerting pressures for political reform and observance of human rights. To do so would clearly contradict the political realities in China’s own domestic sphere. As pointed out by Keenan (2008), “China explicitly disclaims any desire to change the international policies or practices of its trading partners and recipients of its aid or investments” (p. 3). China’s principle concern reflects an understanding of the world of international relations through the lens of realpolitik rather than ambitions to spread institutional reforms informed by an ideology. For their part, African regimes in search for development assistance will be correspondingly inclined to take the path of least resistance. When the choice is accept assistance with conditions that might undermine the basis of one’s own tenure, versus accept assistance with minimal conditions, political leaders single-mindedly concerned about their own ability to remain in office will certainly choose the latter.

This tendency for African leaders to once again own greater latitude for prescribing opposition is compounded by an international climate which stresses rivalry over cooperation and coordination. More specifically, powerful countries such as the United States see China as a rival, indeed a potentially hostile one, for access to Africa’s wealth of resources which are necessary for keeping the motors of global capitalism turning. In order to continue to secure access to markets, especially primary ordered commodities, greater pressure will be placed on

Western governments to lower its good governance expectations. If this scenario plays out,

380

Africans will once again pay a far greater price for the Continent’s subordinate position in the global economy.

Insinuations and Implications

So far, I have managed to articulate the implications that current domestic and international changes might have on the future the CCM’s tenure in Tanzania. What I have so far neglected is to tackle the broader implications of this study beyond the specificities of authoritarian rule per se. In the pages that follow, I provide a brief summary of some thoughts on how the preceding pages might contribute to related areas of study. These include: 1) assessment of the propects of democracy, 2), the assessment of elections as measures of democracy, and 3) the limits of using an analysis of clientele networks as a basis for explaining political loyalties.

Democratic Transition and Consolidation

In the aftermath of the eurphoria of the third wave literature and the pessimism of those that interpreted third wave transitions as transitions to some form of soft-authoritarain rule, has given way to the recent resurgence of the debate over democratic sequencing. As the argument goes, third wave transformations have ensued in a fundamentally different sequence than their first wave counterparts. Whereas early democratizing countries introduced universal suffrage after the formation of the modern state, third wave transitions have introduced universal suffrage prior to the development of the basic institutions of the modern state, including the rule of law, civil society, and accountability (Rose & Shin, 2001). In a short and crude statement, in the first wave, the state precedes democracy while in the third-wave democracy precedes the state (Linz

& Stepan, 1996; Hadenius, 2001).

381

As a consequence of the historically unique temporal sequence, transitions that were initially counted as democratic ones might actually result in the rejection of free and fair elections in exchange for authoritarian induced stability (Rose & Shin, 2001). As pointed out in the first few pages of this piece, rather than a transition toward democracy, the third wave resulted in the rise of other forms of authoritarian rule, highlighting the dangers of pushing for democratization prior to the introduction of necessary pre-conditions (Mansfield & Snyder,

2007). The result of the unique sequence of the third wave is the rise of illiberal multiparty democracy (Zakaria, 1997; Carothers, 2007) and states that prove to be more prone to violence than former authoritarian regimes (Mansfield & Snyder, 1995), as demonstrated in the fragmentation of elite cohesion and the rise of privatized and ethnified violence in the 2007 election in Kenya (Branch & Cheeseman, 2008). Mansfield and Snyder (2007) convincingly argue that, where elections take place in the absence of, among other things, a strong sense of national cohesion, “political competition typically coalesces around and reinforces ethnic and sectarian divisions…” (p. 7). The authors go on to highlight the path dependent risks of pushing for democracy too early: “Once a country starts on an illiberal trajectory, ideas are unleashed and institutions are established that tend to continue propelling it along that trajectory” (p. 7).

Most scholars recognize that the initial optimism surrounding the early third-wave literature was highly misplaced. At the same time, there are those which are skeptical toward the pre- conditions and sequencing arguments and the risks associated with holding multiparty elections prior to the institutionalization of the rule of law and the emergence of a national identity as a centripetal check against the pull of ethnic and regional ones. These skeptics point out that one principle aspect of the modern state is legitimacy, which can hardly ensue where leaders are completely unaccountable to the population in which they govern. As Bratton and Chang (2006)

382 discuss, “the process of state building and democratization are best viewed neither as occurring forwards or backwards but rather reciprocally or together” (p. 1061). In other words, a capable modern state is predicated on legitimacy, best strengthened through institutions like multiparty elections.

Other sequencing skeptics object to the assumption that authoritarian rule is a superior climate for building capable states based on the rule of law (Berman, 1997; Fukuyama, 2007).

As Carothers (2007a) points out, “for every Lee Kuan Yew of Singapore there have been dozens or even hundreds of rapacious, repressive autocrats posing as reformers, leaders for whom the rule of law represents a straightjacket to be avoided at all costs” (p. 15).

The concluding positions of sequence supporters and sequence skeptics are however, not so fundamentally at odds. Most transition theorists today embrace the notion of gradualism, recognizing that early transition researchers failed to appreciate the role that political conditions and historical legacies would play in shaping the nature of democratization. Rather than deterministic pre-conditions, transition theorists have come to talk about facilitating conditions.

The arguments now center on the degree to which low levels of economic development, the concentration of national wealth, and sub-national divisions decrease the prospects of democratic gains rather than precluding the democratic prospects altogether (Carothers, 2007a,

2007b).

It could be argue that the Tanzanian case, as presented in the dissertation, appears to be on a comparatively sound path toward the consolidation of democratic governance. In nearly every sense, these promising prospects are largely products of commitments established during single- party rule. While Ujamaa, which included a development trajectory based on peasant production, an emphasis on social and economic equality, and the cultivation of a profound

383 sense of nationhood, has been widely deemed as an economic disaster (Hyden, 1980; Barkan,

1994), and converse to the spirit of democratic governance, the sense of social cohesion produced out of African Socialism resulted in a degree of political stability found in few other

African countries.

Peace and stability has two implications worth mentioning here. First, over time, the resolution of conflict through peaceful means is positive for the production of trust and mutual respect rather than animosities and hostilities which follow from election violence and repeated deployment of security forces as a primary tool of suppression (Kodmani, 2005; Othman, 2006).

Similarly, national consciousness also provides a glue which binds citizens together into a community whereby subnational violence is taken off the table. Since democratic governance in the liberal sense requires norms of mutual trust and reciprocity to undergird bargaining processes, Tanzania’s past should be considered a highly promising one for the possibilities of democratization.

Secondly, the relatively peaceful atmosphere in Tanzania also has positive implications for economic growth, which according to the arguments in transition literature, also raises the prospects for democratic consolidation. I recall a conversation I had a few years back with an

American embassy official about Tanzania’s comparative prospects at attracting foreign investment. The embassy official proposed a very bleak outlook, citing rampant corruption and inefficiencies as the primary reasons why Tanzania is lagging behind in attracting investment.

My first reaction was to point out that, as far as net investment inflows were concerned, I was not convinced that Tanzania was currently a laggard in reality. Furthermore, while corruption might restrict the inflows of investments from the United States, whose investors are bound by stricter foreign corrupt practices rules, one of the very first things investors from any country

384 look for is the presence of political and economic stability. Stability, or resistance to government collapse and widespread social turmoil, generally is conducive to macro-political and economic predictability and long-term macro-economic growth (Alesina, et. Al., 1996; Feng, 1997;

Gyimah-Brempong & Traynor, 1999). Similarly, periodic outbreaks of violence and social unrest are often taken into account when companies look to invest or tourists look to travel

(Zwass, 1995; Samaraweera, 2008). The irony is that, while African Socialism is often argued to be an economic disaster, in the long-run, in its totality, the policy implications of African

Socialism may have actually created the conditions for stable macro-economic growth.

Measuring the Quality of Elections and Democracy

Whereas this research speaks to the sequencing debate, which, in turn, speaks to the prospects of democracy, the study of party dominance in Tanzania also has a few things to say about elections as measures of the quality of democracy. For starters, there is an undeniable level of criticism over policies that attempt to manage the level of potentially divisive social conflict through mechanisms that limit, sometimes severely so, the scope of political conflict.

Such provisions, to be described in greater detail below, are counter to a universal application of liberal conceptions of open elections as having little or no formal restrictions on the rights of individuals to form political organizations around whatever issue seems important. More specifically, for some state leaders there may be legitimate reasons for constraining the operations of liberal conceptions of democratic elections above and beyond what liberalism recognizes as a legitimate constraint.

This liberal framework proves especially important as a metric for organizations involved in election monitoring, some of which I have participated in. For the Organization for Security and

385

Cooperation in Europe (OSCE) for example, a “genuine” election is conceived as a “celebration of fundamental human rights” characterized by “political competition” amid “confidence, transparency, and accountability” which provides the electorate with “an informed choice between distinct political alternatives” (OSCE, 2005, p. 11). Among other things, a “free” election includes a high degree “freedom of expression, association, assembly, and movement”

(OSCE, 2005, p. 18).

Indeed, election monitoring provides an important venue for keeping the spotlight on the potential for election malpractice, such as ballot-box stuffing, carousel voting, and, for long- term observation, assessing the general fairness of the election environment with the hope of enhancing “accountability and transparency, and thus confidence in the election process”

(OSCE, 2005, 19). For proponents of the universal application of liberal democracy there is repeated criticism over restrictions or disincentives throughout much of the developing world toward regional, ethnic, or religious bases for mobilizing party support. In Kenya for example, the winner in the Presidential election must secure no less than 25 percent of the votes in five out of eight provinces. In Tanzania, political parties must have members in at least ten regions, while presidential candidates need to have signatures from no less than 200 registered voters from 10 regions, 2 of which must be on Zanzibar (TEMCO, 2006). Other restrictions include bans on campaigning around regional, ethnic, or religious issues and in local tongues. In

Tanzania, there have been several cases where the Court of Appeals has actually nullified election results due to repeated expressions of ethnic or racial differences during campaigns

(Chaligha, 2005).

There are a number of specific criticisms to these regulations. First and most commonly, is the objection that regulations which seek to limit the ability of social groups to form political

386 parties violate basic tenets of free and fair elections. Secondly, requiring a regionally diverse membership base in order to form a party and conduct elections is a mechanism by which allows incumbent parties to stay on top since many opposition parties lack the resources for building support throughout the entire country. Finally, ethnicity and religion may be the only “practical” and “normal” method for organizing political parties in certain contexts (Mmuya & Chaligha,

1994, p. 96) and perhaps the only viable way of organizing a competitive political environment.

To be sure, limitations on the ability to form political organization should always be met through the eye of a skeptic. But not a skeptic which shuns anything alien to his or her world view. Assessments of elections and democracy should also consider the larger geopolitical basis behind election assessment: “promoting democratic elections as a key pillar of stability”

(OSCE, 2005, p. 11) not just domestically, but also as a strategy for regional wide peacekeeping efforts. For the OSCE for example, preserving peace and stability in the Balkans was one of the primary motivations for monitoring elections in the first place.

From the perspective of keeping the peace, such restriction seem quite logical, especially under conditions were multiparty competition has the potential for bringing forth the most divisive political discourses with the ability to sow the seeds of violent confrontations for years to come. With this in mind, election assessments might be better served by stating that “given social structural conditions A and B, election performance C should not be surprising, nor necessarily evaluated negatively.”

A second implication of this study is directed toward the democratization literature.

Specifically, there is a marked tendency to associate rising levels of political competition and government turnover as indicators of democratic consolidation, as well as low degrees of electoral competition and prolonged party dominance as a sign of stalled transitions. Because of

387 the hegemony of liberalism in conceptions of democracy the presence of a competitive political arena is almost always assumed to be an indicator for placing regimes on a democratic- authoritarian continuum (Schmitter & Karl, 1991; Diamond, 2002; Levitsky & Way, 2002). The

Tanzanian case I believe, challenges some of these assumptions, suggesting instead that turnover events and levels of competition can indeed be an indicator of democratization, but should never be taken as synonymous with it (Hadenius & Teorell, 2006).

Among the proponents of liberal democracy, there are two statements on the necessity of competition that are worth noting. First, is the somewhat tautological statement that authoritarian regimes with a history of competition are more likely candidates for democratization when compared to regimes without a history of political competition (Bratton

& van de Walle, 1997; Geddes, 1999). This is only logical where definitions of democracy include a heavy emphasis on competition in the first place. Secondly, competition is said to ensure that badly performing governments can be held accountable for their performance

(Schmitter & Karl, 1991; Center for Democratic Institutions, 2003) and helps expand suffrage as political elites compete for votes (Almond, 1991).

While I do not dispute the empirical merit of this latter statement, there are a few limitations worth noting. For starters, competition between factional interests has not always been integral to the definition of democracy. Without appropriate institutional frameworks for example, factional struggles can actually be antithetical to democracy. In a move forewarned by James

Madison, majority factions may emerge out of conflict to simultaneously dominate the legislative, judicial, and executive powers of the state and move to undermine the civil liberties and political rights of other factions. Separately, among those who see democracy as

388 participation in a deliberative process leading to a consensus, competition is not seen as necessary, or necessarily good, for democratic governance (Schmitter & Karl, 1991).

The limitations to this competition-democratization equation are born out in the case studies presented in this piece. The most obvious empirical example of a limitation is found when comparing the level of political competition between Kenya and Tanzania, or better yet Kenya and Botswana. Judging from election margins, Kenya is obviously the most competitive of the all three. KANU won the first two multiparty elections with a mere plurality, only to finally lose the 2002 election. Yet, when referencing the opportunities afforded by the institutions of government, the potential for competition is probably not much different between Kenya and

Tanzania, while Botswana’s liberal democracy, as it’s commonly called (Diamond 2002), has far more openness than nearly any other African polity. Does the appearance of competition warrant labeling Kenya as more democratic or on a clear path toward democratic consolidation vis-à-vis Botswana and Tanzania? Hardly. In many ways, the intensity of competition in Kenya, with the abrasiveness of ethno-centric self-other discourses promoted by ambitious political entrepreneurs, is associated with election conduct that can easily be judged as less democratic than Botswana’s or Tanzania’s for that matter. Furthermore, since democracy also requires parties to “voluntarily make collective decisions binding on the polity as a whole” (Schmitter &

Karl, 1991, p. 79), the social-structural conditions which give rise to political competition in

Kenya also give rise to constraints on cooperation which might not occur in less competitive situations. In short, the qualities of the leadership and identities which feed political competition will have a lot to do with appropriate use of competition as a metric for democratic governance.

Likewise, among transition theorists, talking about democratization without citing regime turnover events has been difficult since the onset of the so called third-wave (Feng, 1997),

389 although many writers do indeed recognize the limitations of using the turnover test as a transition gauge. Even Huntington (1991), who is often associated with the turnover test, was well aware of its limitations. Simply because countries like Japan had not experienced an alternation in power during the Liberal Democratic Party’s tenure between 1955 and 1993 did not mean that Japan was not democratic or less democratic than countries which did experience turnover (p. 27).

Again, the differences between Kenya, where KANU lost the 2002 election, and Tanzania, where the CCM continues to enjoy wide election victories, is a comparison in point. When considering peace and tranquility as necessary conditions for a popular acceptance of the democratic quality of an election, one might easily argue that Tanzania is in a better position for consolidating democracy when compared to the turbulence of Kenya’s elections. Repeated outbreaks of violence during elections or referendums is not, nor should not, be construed as coinciding with steps in more democratic directions. In fact, quite the opposition might occur, since each home burned and each person attacked, beaten, or killed all in the name of ethnic loyalties propagated by ambitious self-interested politicians is only going to solidify a shared sense of mutual hostility and animosity for future elections.

Turnovers or alternations themselves pose an additional problem from the standpoint of democratization. Does a transfer from one party to another party really represent a change in regime? This argument has been, in various ways, applied to transfers from Republicans to

Democrats in the United States during the 1980s and 90s. During this time, the fundamental political orientations between the dominant factions within each party did not differ substantially. Even more questionable is the case of Kenya, where the composition of NARC,

390 which defeated KANU in 2002, contained many members who were only months earlier loyal to KANU.

A manifestation of changing parties without changing regimes tends to pose a real problem for democratic governance. For example, Francis Fox Piven and Richard Cloward’s 2000 publication on voter mobilization in the U.S. argues that, despite the commonly asserted claim that Democrats are more committed to expanding voter turnout, both parties are, in reality, averse to the prospects of bring new voters into the political fray which might ultimately upset the status quo. Similarly, during an interview in Nariobi with Dr. Ludeki Chweya, from the

Department of Political Science and Public Administration at the University of Nairobi, I probed about the ethnic factions within the Kenyan political landscape in the lead-up to the 2005

Constitutional Referendum. Dr. Chweya smiled slightly as if to indicate my somewhat amateurish understanding of Kenyan politics and said “these hostilities you see between elites are more or less showcases. Behind closed doors these is a lot of cooperation between them.” He went on to say that, among all the foul language and personal attacks between them, “all the major political players,” including those in the opposition, “can agree on one thing – to keep class-based contention at bay by dividing them along ethnic lines.” In other words, mobilizing ethnicity is not simply about mobilizing a support base for capturing public office, it is also about diverting attention away from those issues more fundamentally threatening to the status quo. In short, for a number of obvious reasons, a turnover in regime should not be seen as an indicator of democracy, nor should the appearance of politically charged campaign rhetoric.

391

Clientelism

One final implication of this study is the contributions it makes to the study of clientelism as a casual component of party dominance. There are two important political phenomena predicted by the literature on clientelism. First, because executives control clientelistic access to the state, political elites generally coalesce around the presidency and defend their privileges from outsider demands (Amundsen, 1997; van de Walle, 2003; Wantechekon, 2003), similar to the process highlight by Dr. Chweya in reference to elite cohesion in Kenya. In what Rothchild

(1985) has dubbed the “hegemonial exchange” between political elites, or what Bayart (1993) called the “reciprocal assimilation of elites,” political leaders solidify their positions based on an exchange of rewards for support. In short, clientelism functions as an elite adhesive.

Secondly, in regimes heavily dependent on personalized patron-client ties, voters cast their votes for those who offer the highest prospects for distributing rewards, thus privileging those with access to the state and a demonstrated willingness to dispense rewards (Chabal & Daloz,

1999; Wantechekon, 2003; Chandra, 2007). Throughout nearly every multiparty election across

Africa, it is common to hear what are largely empty campaign promises about improving roads and access to social services (van de Walle, 2003). Hence, clientelism functions as a vital link between elites and electorates.

In Malawi, Zambia, Kenya, and Tanzania, patron-client indeed operate as mechanisms for elite cohesion and elite-mass integration. Yet, what surprises me is the extent to which so much of the literature has failed to locate these networks of mutual exchange in the larger social context in which they operate. In fact, much of my work is just a step away from a discussion about the operation of clientelism in different social settings. Let me now relate my work to the concepts of elite cohesion and elite-mass integration.

392

First, as pointed out in chapter two, the ability to effectively exert clientelistic control will be shaped by the level of macro-structural divergence. To be sure, this is the principle claim raised by the “strong states, weak societies” argument articulated extensively by Joel Migdal.

For Migdal, the exodus of colonial rule left in its place societies where social control was highly fragmented (1988, p. 33). Mahmood Mamdani picks up this point with the term “decentralized despotism,” which describes the near universal application of indirect rule under colonialism as a way to subject Africans by constructing a fragmented and repressive native authorities system based upon concocted and racist notions of tribal identity and customary rule (1996, p. 8). In either case, both authors stress the highly fragmented nature of social control at the onset of independence as a consequence of colonial rule and highly disruptive to the consolidation of state power after independence.

The argument of my work parallels those put forth by Migdal and Mamdani. At the time of independence, most African societies were characterized by a centralized administrative state loosely connected to a constellation of relatively powerful social forces throughout rural society with the ability to resist state consolidation. To cite Migdal (2001), the structure of African society, “with its fragmentation of social control, has denied the state the ability to mobilize these clients politically.” As a consequence, the states leaders’ desire for “social change” was traded for the necessity of “survival,” as the regime was compelled to confront thoroughly entrenched conservative interests resistant to change (p. 93). As a consequence, the unwillingness or inability to “de-tribalize” rural society effectively reproduced social structures where ethnic leaders, often called strongmen within the neopatrimonial literature, remained the central organizers of rural society (Mamdani, 1996; Szeftel, 2000).

393

This pattern would have profound consequences in the realm of electoral politics. Where the institutions of colonial social control continued their functions under the post-colonial state, the introduction of elections “meant that ethnic identity was catapulted directly into the electoral arena through clientelist networks speaking for regional or ‘tribal’ factions” (Szeftel, 2000, p.

432).

The case of Tanzania was, in many respects, the exception to this pattern. The consolidation of TANU was not confronted with an enormously powerful array of chiefs, ethnic associations, or yeoman farmers with the ability to block the formulation of transformative aspirations within the leadership stratum. Because TANU was effectively the only game in town, existing elites had little choice but to join the Party’s ranks on the terms largely set by the urban party elite. Stated this way, clientelism in most African countries was based on mutual concessions between rural and urban elites, while in Tanzania, clients signed up, but had little leverage for influencing concessions from the center. Again, to cite Migdal (2001):

Where social structure has not been marked by strongly entrenched strongmen or where such strongmen have been weakened, state leaders have greater opportunities to apply a single set of rules – the state’s rules – and to build channels for widespread, sustained political support (p. 92).

Taking the position which holds that low levels of macro-structural divergence simultaneously allows regimes to pursue ambitious transformative goals bring us to the second point highlighted by clientelism: voters cast their votes toward those parties with the highest prospects for distributing rewards. Again, I will not refute this position, but only suggest that the statement itself begs for an examination of the distribution of material capabilities throughout society. As the Tanzania case illustrates, the historical degree to which the state subsumed society’s productive forces explain the weakened material capacity of political mobilization outside the organizational network of the incumbent party. Randall & Svåsand’s (2002)

394 statement to the effect that neopatrimonial rule “alone” may not suffice as an “explanation for weak parties”, but instead a “combination of patrimonialism and the general scarcity of resources in African societies…” (p. 44) can easily be modified to say that the level of opposition party weakness is a function of the degree to which clientele relations operate in an environment of relative scarcity of material resources outside the patron-client pyramid climaxing in the executive office. Additionally, the intensity, weight, or value of ideational and material spaces for mobilizing dissent clearly impacts the pay-off structures among regime insiders when faced with the choice between defecting from existing political loyalties to or standing one’s ground. In both cases, the scarcity of resources and ideational distinctions leaves little space for the formulation of dissenting views with organizational sustenance.

There is a risk however, in overextending Migdal’s (1988) statement from one that says

“where rural elites are weak, the state has greater latitude in formulating transformative policy,” to one that says “where rural elites are weak the state has greater latitude in implementing transformative policy.” If there is one lesson to learn from Hyden’s seminal 1980 work, it is the lesson about the utter disaster of Ujamaa to translate into actual economic development simply because the state was unable to subjugate a population of free-hold peasants into producing for the market.

While Mamdani is widely critical of Hyden’s work, he indeed gives implicit recognition of the overarching claim made by Hyden. For Mamdani (1996), radical states were in reality no better at bridging the urban-rural divide than conservative ones. Where the latter bridged the divide through clientele networks which utilized chiefs and headman as agents of social control, the former sought to transform “decentralized despotism” into “centralized despotism” which

395 employed administrative ‘experts’ with little or no local ties rather than local collaborators with at least some semblance of local authority as a form of social control over the peasants (p. 26).

Two Models of Party Dominance

The discussion on clientelism can now move to a conclusion which re-articulates the two forms of party-dominance in Africa, outside those cases where harder forms of authoritarian rule remain. On the one hand, the BDP in Botswana emerges from the subjugation of colonial rule as the conservative regime case, where chiefs and elements of the rural petty bourgeoisie were employed for ensuring rural social control. On the other hand, in Tanzania, a regime with more radical ambitions emerged out of colonialism, one committed to replacing chiefs with centrally appointed administrators and the petty bourgeois production with state led peasant collectives.

Each route emerges today with different political and economic qualities.

From the standpoint of macro-economic growth, Botswana emerged as the clearly superior route, as the state was able to employ local collaborators to help steer production for market consumption. In Tanzania on the other hand, peasant free-hold production gave them the capacity to resist state demands, resulting in a wave of desperate measures such as forced villagization. However, aside from growing wealth differentials and high levels of poverty, both countries today are in a comparatively better position for macro-economic gains down the road.

From the standpoint of liberal democratic governance, Botswana is again emerges with a better track record. It remains one of the only African countries where periodic elections coincide with institutions which permit high levels of civil liberties and political rights.

However, there is a certain sour taste in the Botswana case as well, which leads me to conclude that democracy here is highly overrated. Specifically, the reproduction of the market oriented

396 status quo has also required assurances that the population by and large remains passive. I personally have trouble deploying the term “democracy” to describe a situation where passivity is encouraged.

Finally, from the standpoint of party dominance, the Tanzania route has certainly been quite remarkable. Whereas the BDP’s election margins have continually declined since the late 1980s, from the onset of multiparty reform in Tanzania, the CCM has managed to increase its share of the votes. That said, both regimes are presented with somewhat similar challenges: rising income inequalities and unemployment with the potential for exploding into widespread discontent, along with the growth of business and professional interests with the material resources for organizing discontent. Both trends also speak to the possibility of enticing political entrepreneurs inside each ruling party to stake out a political career outside the corridors of the incumbent. At the same time, the international climate appears to less concerned about issues of democratic governance and more concerned about issues of resource extraction. With these contradictory tends in mind, the only genuinely accurate prediction for future of party dominance in each case is ‘party dominance is forever, until it is no more.’

397

REFERENCES CITED

Abrams, Philip. (1982). Historical Sociology. Ithaca, New York: Cornell University Press.

Adar, Korwa G. (1999). The Interface Between Elections and Democracy: Kenya’s Search for a Sustainable Democratic System, 1960s-1990s. In Jonathan Hyslop (Ed.). African Democracy in the Era of Globalization (pp. 340-360). Johannesburg: Witwatersrand University Press.

Africa Watch (1993). Divide and Rule: State-Sponsored Ethnic Violence in Kenya. New York: Human Rights Watch.

African Elections Database. (2007a). Elections in Botswana. Retrieved December 14, 2007 at http://africanelections.tripod.com/bw.html

African Elections Database. (2007b). Elections in Burundi. Retrieved December 14, 2007 at http://africanelections.tripod.com/bi.html.

African Elections Database. (2007c). Elections in Chad. Retrieved December 14, 2007 at http://africanelections.tripod.com/td.html.

African Elections Database. (2007d). Central African Republic. Retrieved December 14, 2007 at http://africanelections.tripod.com/cf.html.

African Elections Database. (2007e). Elections in Congo-Brazzaville. Retrieved December 14, 2007 at http://africanelections.tripod.com/cg.html.

African Elections Database. (2007f). Elections in Congo-Kinshasa. Retrieved December 14, 2007 at http://africanelections.tripod.com/cd.html.

African Elections Database. (2007g). Elections Equatorial Guinea. Retrieved December 14, 2007 at http://africanelections.tripod.com/gq.html.

African Elections Database. (2007h). Elections in Gabon. Retrieved December 14, 2007 at http://africanelections.tripod.com/ga.html.

African Elections Database. (2007i). Elections in Ghana. Retrieved December 14, 2007 at http://africanelections.tripod.com/gh.html.

African Elections Database. (2007j). Elections in Guinea. Retrieved December 14, 2007 at http://africanelections.tripod.com/gn.html.

African Elections Database. (2007k). Elections in Kenya. Retrieved December 14, 2007 at http://africanelections.tripod.com/ke.html

398

African Elections Database. (2007l). Elections in Lesotho. Retrieved December 14, 2007 at http://africanelections.tripod.com/ls.html.

African Elections Database. (2007m). Elections in Malawi. Retrieved December 14, 2007 at http://africanelections.tripod.com/mw.html

African Elections Database. (2007n). Elections in Mauritius. Retrieved December 14, 2007 at http://africanelections.tripod.com/mu.html.

African Elections Database. (2007o). Elections in Nigeria. Retrieved December 14, 2007 at http://africanelections.tripod.com/ng.html.

African Elections Database. (2007p). Elections in Rwanda. Retrieved December 14, 2007 at http://africanelections.tripod.com/rw.html.

African Elections Database. (2007q). Elections in Senegal. Retrieved December 14, 2007 at http://africanelections.tripod.com/sn.html.

African Elections Database. (2007r). Elections in Sierra Leone. Retrieved December 14, 2007 at http://africanelections.tripod.com/sl.html.

African Elections Database. (2007s). Elections in Tanzania. Retrieved December 14, 2007 at http://africanelections.tripod.com/tz.html

African Elections Database. (2007t). Elections in The Gambia. Retrieved December 14, 2007 at http://africanelections.tripod.com/gm.html.

African Elections Database. (2007u). Elections in Togo. Retrieved December 14, 2007 at http://africanelections.tripod.com/tg.html.

African Elections Database. (2007v). Elections in Uganda. Retrieved December 14, 2007 at http://africanelections.tripod.com/ug.html.

African Elections Database. (2007w). Elections in Zambia. Retrieved December 14, 2007 at http://africanelections.tripod.com/zm.html

Afrobarometer. (1999a). Malawi Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round1c.html.

Afrobarometer. (1999b). Zambia Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round1c.html.

Afrobarometer. (2001). Tanzania Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round1c.html.

399

Afrobarometer. (2003). Kenya Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round2c.html.

Afrobarometer. (2005b). Kenya Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round3c.html.

Afrobarometer. (2005c). Malawi Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round3c.html.

Afrobarometer. (2005d). Tanzania Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round3c.html.

Afrobarometer. (2005e). Zambia Data. [Data file]. Retrieved November 29, 2007 from http://afrobarometer.org/round3c.html.

Agonafer, Mulugeta. (1994). Contending Theories of Development in the Contemporary International Disorder/Order: Lessons from Kenya and Tanzania. Lanham, Maryland: University Press of America.

Aidoo, Thomas Maxwell. (2008). Political Participation, Governance, and Neopatrimonial Rule in Africa. Saarbrücken, Germany: VDM Verlag Dr. Muller Aktiengesellschaft

Aitken, Rob. (1996). Neoliberalism and Identity: Redefining State and Society in Mexcio. In Rob Aitken, Nikki Craske, Gareth Jones, & David Stansfield (Eds.). Dismantling the Mexican State? Houndmills, NH: Macmillian Press.

Ake, Claude. (1976). Explanatory Notes on the Political Economy of Africa. The Journal of Modern African Studies, 14(1), 1-23.

Akivaga, S. Kichamu. (2004).Toward a National Movement for Democratic Change in Kenya. In Kioko, Wanza, Lawrence Murugu Mute & S. Kichamu Akivaga (Eds.), Building an Open Society: The Politics of Transition in Kenya (pp. 8-59). Nairobi: Claripress.

Alavi, Hamza. (1972). The State in Post-Colonial Societies: Pakistan and Bangladesh. New Left Review, I/74, 59-81.

Alesina, Alberto, Sule Ozler, Nouriel Roubini, and Phillip Swagel. (1996). Political Instability and Economic Growth. Journal of Economic Growth, 1(2), 189-212.

Alexander, Gerard. (2001). Institutions, Path Dependence, and Democratic Consolidation. Journal of Theoretical Politics, 13(70), 249-270.

Almond, Gabriel A. Capitalism and Democracy. (1991). Political Science and Politics, 24(3), 476-474.

400

Almond, Gabriel & Verba, Sidney. (1963). The Civic Culture: Political Attitudes and Democracy in Five Nations. Princeton: Princeton University Press.

Amundsen, Inge. (1997). In Search of a Counter-Force: State Power and Civil Society in the Struggle for Democracy in Africa. Unpublished doctoral dissertation, University of Tromsø, Tromsø, Norway.

Apter, David E. (1970). Ghana. In Coleman, James S. & Carl G. Rosberg, Jr. (Eds.). Political Parties and National Integration in Tropical Africa (pp. 259-316). Berkeley, CA: University of California Press.

Arrighi, Giovanni & John S. Saul. (1968). Socialism and Economic Development in Tropical Africa. The Journal of Modern African Studies, 6(2), 141-169.

Asher, Mukul G. (1994). Some Aspects of Role of State in Singapore. Economic and Political Weekly, 29(14), 795-804.

Asingo, Patrick O. (2003). The Political Economy of Transition in Kenya. In Oyugi, Walter O., Peter Wanyande & C. Odhiambo Mbai. The Politics of Transition in Kenya: From KANU to NARC (pp. 15-50). Nairobi: Heinrich Böll Foundation.

Askew, Kelly M. (2002). Performing the Nation: Swahili Music and Cultural Politics in Tanzania. Chicago: The University of Chicago Press.

Ayang, S. G. (1970). A History of Zanzibar: A Study in Constitutional Development. Nairobi: Kenya Literature Bureau.

Baccaro, Lucio. (2006). Civil Society Meets the State: Towards Associational Democracy? Socio-Economic Review, 4, 185-208.

Bačová, Viera. (1998). The Construction of National Identity – on Primordialism and Instrumentalism. Human Affairs, 8(1), 29-43.

Bahati, Ben (2005, October 16). Goodbye outstanding African president. Sunday News.

Bakary, Abubakar Khamis. (2006). The Union Question and the Zanzibar Constitution. In Peter, Chris Maina & Haroub Othman (Eds.), Zanzibar and the Union Question (pp. 1-33). Zanzibar: Zanzibar Legal Services Center.

Balasubramaniam, Vejai. (2006). Embedding Ethnic Politics in Malaysia: Economic Growth, its Ramifications and Political Popularity. Asian Journal of Political Science, 14(1), 23-39.

Baregu, Mwesiga. (1994). The Rise and Fall of the One-Party State in Tanzania. In Widner, Jennifer A. (Ed.) Economic Change and Political Liberalization in Sub-Saharan Africa (pp. 159-181). Baltimore: Johns Hopkins University Press.

401

Baregu, Mwesiga. (1997). Political Culture and the Party-State in Tanzania. In Research and Education for Democracy in Tanzania (REDET), Political Culture and Popular Participation in Tanzania (pp. 53-72). Dar es Salaam: Author.

Baregu, Mwesiga & S. S. Mushi (1994). Mobilization, Participation and System Legitimacy. In Mukandala, Rwekaza S. & Haroub Othman (Eds.), Liberalization and Politics: The 1990 Elections in Tanzania (pp. 90-133). Dar es Salaam: Dar es Salaam University Press.

Barkan, Joel D. (1979a). Comparing Politics and Public Policy in Kenya and Tanzania. In Barkan, Joel D. & John J. Okumu (Eds.). Politics and Public Policy in Kenya and Tanzania (pp. 3-39). New York: Praeger.

Barkan, Joel D. (1979b). Legislators, Elections, and Political Linkages. In Barkan, Joel D. & John J. Okumu (Eds.). Politics and Public Policy in Kenya and Tanzania (pp. 64-92). New York: Praeger.

Barkan, Joel D. (1994). Divergence and Convergence in Kenya and Tanzania: Pressures for Reform. In Barkan, Joel D. Beyond Capitalism vs. Socialism in Kenya and Tanzania (pp. 1- 45). Boulder, Colorado: Lynne Rienner.

Barkan, Joel D. and John J. Okumu. (1978). Semi-Competitive Elections, Clientelism, and Political Recruitment in a No-Party State: The Kenyan Experience. In Guy Hermet, Richard Rose and Alain Rouquié (Eds.). Elections Without Choice (pp. 88-107). London: Macmillan Press.

Bauer, Gretchen. (1999). Challenges to Democratic Consolidation in Namibia. In Joseph, Richard. (Ed.), State, Conflict, and Democracy in Africa. Boulder: Lynne Rienner Publishers, Inc.

Bavu, Immanuel. (1990a). Election Management and Democracy. In Haroub, Othman; Immanuel K. Bavu & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 22- 36). Dar Es Salaam: Dar Es Salaam University Press.

Bavu, Immanuel. (1990b). Ilala: The Politics of Ethnicity in a Cosmopolitan Setting. In Haroub, Othman; Immanuel K. Bavu & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 88-102). Dar Es Salaam: Dar Es Salaam University Press.

Bayart, Jean-Francois. (1993). The State in Africa: The Politics of the Belly. New York: Longman Group.

Baylies, Carolyn & Morris Szeftel. (1992). The Fall and Rise of Multi-Party Politics in Zambia. Review of African Political Economy, 54, 75-91.

BBC World Trust Service (2008, April). The Kenyan 2007 elections and their aftermath: the role of the media (Policy Brief No. 1). London: Author.

402

Beblawi, Hazem. (1990). The Rentier State in the Arab World. In Luciani, Giacomo. (Ed.), The Arab State. California: University of California Press.

Bennet, Andrew, Barth, Aharon, & Rutherford, Kenneth R. (2003). Do We Preach What We Practice? A Survey of Methods in Political Science Journals and Curricula. Political Science and Politics, 36, 373-378.

Berman, Sheri. (2007). The Vain Hope for Correct Timing. Journal of Democracy, 18(3), 14-17.

Bienen, Henry. (1970). Tanzania: Party Transformation and Economic Development. Princeton, New Jersey: Princeton University Press.

Boesen, Jannik. (1979). Tanzania: From Ujamaa to Villagization. In Mwansasu, Bismarck & Cranford Pratt (Eds.). Toward Socialism in Tanzania (pp. 125-144). Toronto: University of Toronto Press.

Boone, Catherine. (1994). States and Ruling Classes in Postcolonial Africa. In Migdal, Joel, Atul Kohli & Vivienne Shue (Eds.). State Power and Social Forces (pp. 108-135). Cambridge: Cambridge University Press.

Booth, John A. & Richard, Patricia Bayer. (1995). Repression, Participation and Democratic Norms in Urban Central America. American Journal of Political Science, 40, 1205-1232.

Bratton, Michael. (1994). Civil Society and Political Transition in Africa. Institute for Development Research. (IDR Report Vol. 11, 6).

Bratton, Michael & Eric Chang (2006). State Building and Democratization in Sub-Saharan Africa: Forwards, Backwards, or Together? Comparative Political Studies, 39, 1059-83.

Bratton, Michael & Nicholas van de Walle. (1997). Democratic Experiments in Africa: Regime Transformation in Comparative Perspective. New York: Cambridge University Press.

Brennan, James R. (2006). Youth, the TANU Youth League and Managed Vigilantism in Dar es Salaam, Tanzania, 1925-73. Africa, 67, 221-246.

Brown, David. (1982). Who Are the Tribalists? Social Pluralism and Political Ideology in Ghana. African Affairs, 81(322), 37-69.

Brumberg, Daniel. (2002). The Trap of Liberalized Autocracy. Journal of Democracy, 13(4), 57-68.

Buchert, Lene. (1994). Education in the Development of Tanzania: 1919-1990. London: James Currey.

Burnell, Peter. (2001). The Party System and Party Politics in Zambia: Continuities Past, Present, and Future. African Affairs, 100, 239-263.

403

Byarugaba, E. F. (1998). Ethnopolitics and the State – Lessons from Uganda. In Salih, M.A. Mohamed & John Markakis. Ethnicity and the State in Eastern Africa (pp. 180-189). Uppsala. Nordiska Afrikainstitutet.

Canel, Eduardo. (1997). New Social Movement Theory and Resource Mobilization Theory: The Need for Integration. In Michael Kaufman and Haroldo Dilla Alfonso (Eds.). Community Power and Grassroots Democracy: The Transformation of Social Life (pp. 189-221). London: Zed Books.

Carothers, Thomas. (2002). The End of the Transition Paradigm. Journal of Democracy, 13(1), 5-21.

Carothers, Thomas. (2004). Critical Missions: Essays on Democracy Promotion. Washington, D.C.: Carnegie Endowment for International Peace.

Carothers, Thomas. (2007a). The Sequencing Fallacy. Journal of Democracy. 18(1), 12-27.

Carothers, Thomas. (2007b). Misunderstanding Gradualism. Journal of Democracy, 18(3), 18- 22.

Case, William. (2004). New Uncertainties for an Old Pseudo-Democracy: The Case of Malaysia. Comparative Politics, 37(1), 83-104.

CCM delivers Mrema to the opposition. (1995, 2-4 March). The Express.

CCM moves to enforce discipline. (2005, October 15). The Daily News.

Center for Democratic Institutions. (2003, May). Promoting Transparent and Accountable Governance through Overseas Aid Programs. Canberra: Author.

Chabal, Patrick & Jean-Pascal Daloz. (1999). Africa Works: Disorder as Political Instrument. Bloomington, IN: Indiana University Press.

Chachage, C.S.L. (1998). Nation Building and Ethnicity: Towards a Re-Conceptualization of Democracy in Africa. Utafiti, 4, 151-178.

Chaligha, Amon E. (2005). The State of Political Parties in Tanzania. In Research and Education for Democracy in Tanzania (REDET), Democratic Transition in Africa (pp. 120- 138). Dar es Salaam: Author.

Chambua, Samuel E. (2002). Democratic Participation in Tanzania: the Voices of Workers Representation. Dar es Salaam: Dar es Salaam University Press.

Chandra, Kanchan. (2007). Counting Heads: a theory of voter and elite behavior in patronage democracies. In Kitschelt, Herbert & Steven I Wilkinson (Eds.), Patronage, Clients and

404

Policies: Patterns of Democratic Accountability and Political Competition (pp. 84-109). Cambridge: Cambridge University Press.

Chang, David W. (1968). Nation-Building in Singapore. Asian Survey, 8(9), 761-773.

Chazan, Naomi. (1982). The New Politics of Participation in Tropical Africa. Comparative Politics, 14(2), 169-89.

Chege, Michael. (1994). The Return of Multiparty Politics. In Barkan, Joel D. (Ed.), Beyond Capitalism vs. Socialism in Kenya and Tanzania (pp. 47-74). Boulder, Colorado: Lynne Rienner.

Chehabi, H. E. & Linz, Juan J. (1998). A Theory of Sultanism 1: A Type of Nondemocratic Rule. In Chehabi, H. E. & Linz, Juan J. (Eds.), Sultanistic Regimes (pp. 3-25). Baltimore: Johns Hopkins University Press.

Chikulo, Bornwell C. (1988). The Impact of Elections in Zambia’s One Party Second Republic. Africa Today, 35(2), 37-49.

Chu, Yun-han. (2001). The Legacy of One-Party Hegemony is Taiwan. In Larry Diamond and Richard Gunther (Eds.). Political Parties and Democracy (pp. 266-98). Baltimore: Johns Hopkins University Press.

Clark, Edmund W. (1978). Socialist Development and Public Investment in Tanzania, 1964-73. Toronto: University of Toronto Press.

Clayton, Anthony. (1981). The Zanzibar Revolution and Its Aftermath. London: C. Hurst & Co.

Clignet, Remi P. & Philip J. Foster. (1964). French and British Colonial Education in Africa. Comparative Education Review, 8(2), 191-198.

Cohen, Jean .L. (1982). Class and Civil Society: The Limits of Marxian Critical Theory. Amherst, MA: The University of Massachusetts Press.

Colclough, Christopher & Stephen McCarthy. (1980). The Political Economy of Botswana: A Study of Growth and Distribution. Oxford: Oxford University Press.

Coleman, James S. & Carl G. Rosberg Jr. (1966). Political Parties and National Integration in Tropical Africa. Berkeley: University of California Press.

Collier, Ruth Berins. (1982). Regimes in Tropical Africa: Changing forms of Supremacy, 1945- 1975. Berkeley: University of California Press.

Collier, Ruth Berins & David Collier. (1991). Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics in Latin America. Princeton: Princeton University Press.

405

Commonwealth Human Rights Initiative. (2006). The Police, the People, the Politics: Police Accountability in Tanzania. New Delhi: Author.

Cook, Thomas D. & Donald T Campbell. (1979). Quasi-Experimentation: Design and Analysis Issues for Field Settings. Boston, Massachusetts: Houghton Mifflin.

Cooksey, Brian, David Court & Ben Makau. (1994). Education for Self-Reliance and Harambee. In Barkan, Joel D. (Ed.). Beyond Capitalism vs. Socialism in Kenya and Tanzania (pp. 201-233). Boulder, CO: Lynne Rienner.

Coulson, Andrew. (1982). Tanzania: A Political Economy. Oxford: Clarendon Press.

Court, David & Dharam Ghai (Eds.). (1974). Education, Society, and Development: New Perspectives from Kenya. Nairobi: University of Nairobi Institute of Development Studies.

Court, David & Kabiru Kinyanjui. (1980). Development Policy and Educational Opportunity: The Experience of Kenya and Tanzania (IDS Occasional Paper no. 33). Nairobi: University of Nairobi.

Dahl, Robert. (1956). A Preface to Democratic Theory. Chicago: The University of Chicago Press.

Davison, Jean. (1997). Gender, Lineage, and Ethnicity in Southern Africa. Boulder, CO: Westview Press.

Dessler, David. (1989). What’s at Stake in the Agent-Structure Debate? International Organizations, 43, 441-473.

Dessler, David. (1991). Beyond Correlations: Toward a Causal Theory of War. International Studies Quarterly, 35, 337-355.

Deutsch, Karl A. (1961). Social Mobilization and Political Development. American Political Science Review, 55, 492-514.

Diamond, Larry. (2002). Thinking About Hybrid Regimes. The Journal of Democracy, 13(2), 21-35.

Diamond, Larry, Anthony Kirk-Greene, and Oyeleye Oyediran. (1997). Transition without End: Nigerian Politics and Civil Society under Babangida. Boulder: Lynne Rienner.

Dismas, Paschal (1995, 14 September). Govt. gives Mrema body guards. The Guardian.

Douma, Pyt. (2006). Poverty, Relative Deprivation and Political Exclusion as Drivers of Violent Conflict in Sub Saharan Africa. Journal on Science and World Affairs, 2(2), 59-69.

406

Du Toit, Pierre. (1999). Bridge or Bridgeheads? Comparing the Party Systems of Botswana, Namibia, Zimbabwe, Zambia, and Malawi. In Giliomee, Hermann & Charles Simkins (Eds.), The Awkward Embrace: One-Party Domination and Democracy (pp. 193-217). Amsterdam: Harwood Academic Publishers.

Dzimbiri, Lewis B. (2005). The State and Labour Control in Malawi: Continuities and Discontinuities Between One-Party and Multiparty Systems. Africa Development, 30(4), 53- 85.

Ehrenkranz, Andrew. (2006, 26 July). Zanzibar: Between Mecca and a Hard Place. World Politics Review. Retrieved April 1, 2009, from http://www.worldpoliticsreview.com.

Engels, Frederick. (2003). Socialism: Utopian and Scientific. Marx/Engels Internet Archive. Retrieved January 30, 2009, from http://www.marxists.org.

Englebert, Pierre. (2000). State Legitimacy and Development in Africa. Boulder: Lynne Rienner Publishers.

Erdmann, Gero. (1994). Guided Democratization: Political Perceptions and Attitudes in Tanzania (Africa Discussion Papers, no. 11). Bremen, Germany: University of Bremen.

Erdmann, Gero. (2002). Neo=Patrimonial Rule: Transition to Democracy has not Succeeded. D+C Development and Cooperation, 1, 8-11.

Eriksen, Thomas Hylland. (1991). Ethnicity versus Nationalism. Journal of Peace Research, 28(3), 263-278.

Esman, Milton. (1987). Ethnic Politics and Economic Power. Comparative Politics, 19(4), 395- 418.

Esman, Milton. (1994). Ethnic Politics. Ithaca, NY: Cornell University Press.

Esping-Andersen. (1985). Politics Against Markets. Princeton: Princeton University Press.

Evans, Jocelyn A. J. (2002). In Defence of Sartori: Party System Change, Voter Preference Distributions and Other Competitive Incentives. Party Politics, 8(2), 155-174.

Fainsod, Merle. (1963). How Russia is Ruled (Rev. ed.). Cambridge, Massachusetts: Harvard University Press.

Fangyin, Zhou. (2007). The Role of Ideational and Material Factors in the Qing Dynasty Diplomatic Transition. Chinese Journal of International Politics, 1, 447-474.

Feng, Yi. (1997). Democracy, Political Stability and Economic Growth. British Journal of Political Science, 27, 391-418.

407

Fiorina, Morris. (1981). Retrospective Voting in American Elections. New Haven: Yale University Press.

Firmin-Sellers, Kathryn. (2000). Institutions, Context, and Outcomes: Explaining French and British Rule in West Africa. Comparative Politics, 32(3), 253-272.

Foltz, William J. (1970). The Pragmatic Pluralist Pattern: Senegal. In Coleman, James S. & Carl G. Rosberg, Jr. (Eds.). Political Parties and National Integration in Tropical Africa (pp. 16- 64). Berkeley, CA: University of California Press.

Forster, Peter G., Michael Hitchcock & Francis F. Lyimo. (2000). Race and Ethnicity in East Africa. New York: St. Martin’s Press.

Foucault, Michel. (1995). Discipline and Punish: The Birth of the Prison. New York: Random House.

Foweraker, Joe & Todd Landman. (2000). Citizenship Rights and Social Movements: A Comparative and Statistical Analysis. Oxford: Oxford University Press.

Francisco, Ronald A. (1995). The Relationship Between Coercion and Protest: An Empirical Evaluation in Three Coercive States. The Journal of Conflict Resolution, 39(2), 263-282.

Freedom House. (2005). Freedom in the World Country Ratings: 1972 through 2003 [Data File]. Retrieved August 3, 2006, from www.freedomhouse.org.

Freund, W. M. (1981). Class Conflict, Political Economy and the Struggle for Socialism in Tanzania. African Affairs, 80(321), 483-499.

Friese, Heidrun. (2002). Identity: Desire, Name and Difference. In Friese, Heidrun (Ed). Identities: Time, Difference and Boundaries (pp. 17-55). New York: Berghahn Books.

Fukuyama, Francis (2007). Liberalism Versus State-Building. Journal of Democracy, 18(3), 10- 13.

Gabriel, Jürg Martin. (1998). Cameroon’s Neopatrimonial Dilemma. Zürich: Center for International Studies. (Nr. 20).

Geddes, Barbara. (1999). Authoritarian Breakdown: Empirical Test of a Game Theoretic Argument. Paper presented at Annual Meeting of the American Political Science Association, Atlanta, GA.

Geertz, Clifford. (1963). The Integrative Revolution: Primordial Sentiments and Civil Politics in the New States. In Geertz, Clifford (Ed.). Old Societies and New States. New York: Free Press.

408

Gertzel, Cherry. (1984). Dissent and Authority in the Zambian One-Party State 1973-80. In Gertzel, Cherry (Ed.), Carolyn Baylies & Morris Szeftel. The Dynamics of the One-Party State in Zambia (79-115). Manchester: Manchester University Press.

Gertzel, Cherry, Carolyn Baylies & Morris Szeftel. (1984). Introduction: the Making of the One-Party State. In Gertzel, Cherry (Ed.), Carolyn Baylies & Morris Szeftel. The Dynamics of the One-Party State in Zambia (1-28). Manchester: Manchester University Press.

Giddens, Anthony. (1984). The Constitution of Society: Outline of the Theory of Structuration. Cambridge: Polity.

Giliomee, Hermann & Simkins, Charles. (1999). The Dominant Party Regimes of South Africa, Mexico, Taiwan and Malaysia: A Comparative Perspective. In Giliomee, Hermann & Simkins, Charles. (Eds.), The Awkward Embrace: One-Party Domination and Democracy. Amsterdam: Harwood Academic Publishers.

González, Luis E. (1995). Continuity and Change in the Uruguayan Party System. In Mainwaring, Scott & Timothy R. Scully (Eds.), Building Democratic Institutions: Party Systems in Latin America (pp. 138-163). Stanford: Stanford University Press.

Goodwin, Jim (1995a, 16-19 April). Youth want change. The Express.

Goulbourne, H. (1978). Some Aspects of Ideological Functions in the Development of the Post- Colonial State in Tanzania. Utafiti, 3(2), 377-386.

Gramsci, Antonio. (1977). Selections from Political Writings: 1910-1920. London: Lawrence and Wishart.

Gurr, Robert. (1971). Why Men Rebel. Princeton: Princeton University Press.

Guzzini, Stefano & Anna Leander. (2006). A Relentless Quest for Synthesis. In Stefano Guzzini & Anna Leander. Constructivism and International Relations: Alexander Wendt and his Critics (pp. 73-92). New York: Routledge.

Gyimah-Boadi, E. (1997). Civil Society in Africa. In Diamond, L. et al., (Eds.), Consolidating the Third Wave Democracies (278-292). Baltimore: The Johns Hopkins University Press.

Gyimah-Brempong, K. and T.L. Traynor. (1999). Political Instability, Investment and Economic Growth in Sub-Saharan Africa. Journal of African Economies. 8(1), 52-86.

Hacker, Jacob. (2002). The Divided Welfare State: The Battle over Public and Private Social Benefits in the United States. Cambridge: Cambridge University Press.

Hadenius, Axel. (2001). Institutions and Democratic Citizenship. Oxford: Oxford University Press.

409

Hadenius, Axel & Jan Teorell. (2006). Authoritarian Regimes: Stability, Change, and Pathways to Democracy, 1972-2003. Unpublished manuscript.

Haggard, Stephen & Kaufman, Robert. (1995). The Political Economy of Democratic Transitions. Princeton: Princeton University Press.

Hall, Budd & Stephen Lucas. (1974). The Election as an Exercise in Political Communication. In The Election Study Committee, University of Dar es Salaam. Socialism and Participation: Tanzania’s 1970 National Elections (pp. 165-188). Dar es Salaam: Tanzania Publishing House.

Harbeson, J. (1994). Civil Society and Political Renaissance in Africa. In Harbeson, J., D. Rothchild & Chazan (Eds.), Civil Society and the State in Africa (pp. 1-29). Boulder, Colorado: Lynne Rienner.

Harris, Belle. (1967). The Electoral System. In Cliffe, Lionel (Ed.), One Party Democracy: The 1965 Tanzania General Elections (pp. 20-52). Nairobi: East African Publishing House.

Harris, David E. (2003). Global Zanzibar: East Meets West South on the Indian Ocean. Unpublished Honors Thesis, University of California, Berkeley.

Hatch, John Charles. (1976). Two African Statesmen: Kaunda of Zambia and Nyerere of Tanzania. London: Secker & Warburg.

Havnevik, Kjell J. (1993). Tanzania: The Limits to Development from Above. Sweden: Nordiska Afrikainstitutet.

Hay, Colin. (2001). What Place for Ideas in the Structure-Agency Debate? Globalization as a Process with a Subject. The First Press, Retrieved June 8, 2009 from www.theglobalsite.ac.uk

Hellman, Geoffrey. (1983). Realist Principles. Philosophy of Science, 50, 227-249.

Holmquist, Frank. (1979). Class Structure, Peasant Participation, and Rural Self-Help. In Barkan, Joel D. & John J. Okumu (Eds.). Politics and Public Policy in Kenya and Tanzania (129-153). New York: Praeger.

Hopkins, Raymond F. (1971). Political Roles in a New State: Tanzania’s First Decade. New Haven, Connecticut: Yale University Press.

Horowitz, Donald L. (1985). Ethnic Groups in Conflict. Berkeley: University of California Press.

Huntington, Samuel P. (1991). Democracy’s Third Wave. Journal of Democracy, 2(2), 12-34.

410

Hyden, Goran. (1980). Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry. Berkeley: University of California Press.

Hyden, Goran. (1994). Party, State, and Civil Society: Control versus Openness. In Barkan, Joel D. (Ed.), Beyond Capitalism vs. Socialism in Kenya and Tanzania (pp. 75-99). Boulder, Colorado: Lynne Rienner.

HHwang, In-Won. (2003). Personalized Politics: The Malaysian State under Mahathir. Singapore: Institute of Southeast Asian Studies.

Ihonvbere, Julius O. (1997). From Despotism to Democracy: The Rise of Multiparty Politics in Malawi. Third World Quarterly, 18(2), 225-247.

Ihonvbere, Julius O. (2003). Overcoming a One-Man Dictatorship: Political Liberalization and Democratization in Malawi. In Julius O. Ihonvbere and John M. Mbaku (Eds.). Political Liberalization and Democratization in Africa (pp. 243-276). Portsmouth, NH: Greenwood.

International Finance Corporation. (2006). Micro, Small, and Medium Enterprises: A Collection of Published Data. [Data file]. Retrieved October 30, 2007 from Earthtrends, http://earthtrends.wri.org.

International Institute for Democracy and Electoral Assistance. (2007). Country View: Botswana. Retrieved December 14, 2007 at http://www.idea.int

International Monetary Fund. (2007). World Economic Outlook. [Data file]. Retrieved May 12, 2008, from http://www.imf.org.

Iversen, Torill (2001). Tanzania: General Elections 2000 (Nordem Working Paper 09/01). Oslo, Norwegian Centre of Human Rights.

Jacobs, Lawrence R. & Robert Y. Shapiro. (2000). Politicians Don’t Pander: Political Manipulation and the Loss of Democratic Responsiveness. Chicago: The University of Chicago Press.

Jain, Shail. (1975). Size Distribution of Income: A Compilation of Data. Washington D.C.: The World Bank.

Jenkins, J.C. (1983). Resource Mobilization Theory and the Study of Social Movements. Annual Review of Sociology, 9, 527-53.

Jessop, Bob. (1982). The Capitalist State. New York: New York University Press.

Joseph, Richard (1999). The Reconfiguration of Power in Late Twentieth Century Africa. In Joseph, Richard (Ed.), State Conflict, and Democracy in Africa (pp. 57-80). Boulder: Lynne Rienner Publishers, Inc.

411

Kagwanja, Peter. (2006). Counter-Terrorism in the Horn of Africa: New Security Frontiers, Old Strategies. African Security Review, 15(3), 72-86.

Kahoho, Timothy. (2005, 1 February). Mrema urges court to block opponents. The Gaurdian.

Kalpeni, Ezekiel. (1992). Political Development and Prospects for Democracy in Malawi. Trans Africa Forum, 9(1), 27-40.

Kanyinga, Karuti. (1998). Contestation over Political Space: The State and Demobilization of Opposition Politics in Kenya. In Olukoshi, A. O. (Ed.). The Politics of Opposition in Contemporary Africa (pp. 39-90). Uppsala: Nordic Africa Institute.

Kapinga, Osmund. (1995, 5-8 February). Party financial disposition weak. The Express.

Karl, Terry Lynn. (1995). The Hybrid Regimes of Central America. Journal of Democracy, 6(3), 72-86.

Karugire, Samwiri R. (1996). The Roots of Instability in Uganda. Kampala: Foundation Publishers.

Karume Decries Divisive Politics. (2005, October 16). Sunday News.

Karume to improve retirement benefits. (2005, 28 October). The Daily News.

Kaspin, Deborah. (1995). The Politics of Ethnicity in Malawi’s Democratic Transition. The Journal of Modern African Studies, 33(4), 595-620.

Kasumuni, Ludger (2005a, May 6). It’s a Smooth Ride for Kikwete’s Entry to State House. The Guardian.

Kasumuni, Ludger (2005b, October 29). Mkapa lashes out at opposition in Zanzibar. The Guardian.

Kaunda, Jonathan. (1998). The State and Society in Malawi. Commonwealth & Comparative Politics, 36(1), 48-67.

Keenan, Patrick James. (2008). Curse or Cure? China, Africa, and the Effects of Unconditional Wealth. Berkley Journal of International Law, Forthcoming. Retrieved April 2, 2009, from http://ssrn.com

Kenya National Commission on Human Rights. (2006). Behaving Badly: Deception, Chauvinism and Waste During the Referendum Campaigns. Nairobi: Author.

Kihaule, Emmanuel (2005, October 27). Kikwete: Power sharing possible but…The Guardian.

412

Kikwete, Jakaya Mrisho. (2005, 30 December). Speech on Inaugurating the Fourth Phase Parliament of the United Republic. Dodoma: Author.

Kilimwiko, Rence (1995, 10 September). The campaign fever. The Guardian.

Killian, Bernadeta (1999). The March 1999 Opinion Poll Findings and Discussion: Citizen’s Evaluation of Government and Political Institutions (REDET Opinion Poll Report No. 1). Dar es Salaam: Research and Education for Democracy in Tanzania.

Kilson, Martin L. (1963). Authoritarian and Single-Party Tendencies in African Politics. World Politics, 15, 262-294.

Kiondo, Andrew S. (1994). Economic Power and Electoral Politics in Tanzania. In Mukandala, Rwekaza S. & Haroub Othman (Eds.), Liberalization and Politics: The 1990 Elections in Tanzania (pp. 65-89). Dar Es Salaam: Dar Es Salaam University Press.

Kiondo, Andrew S. (1995). No to Democracy from Above: The Search for a Democratic Transition in Tanzania. In Melin, Mia (Ed.). Democracy in Africa: On Whose Terms. Sweden: Forum Syd.

Kiondo, Andrew S. (2001). Group Differences in Political Orientation: Ethnicity and Class. In Mushi, Samuel S., Rwekaza S. Mukandala & Mwesiga L. Baregu (Eds.). Tanzania’s Political Culture: A Baseline Survey (pp. 252-275). Dar es Salaam: University of Dar es Salaam, Department of Political Science and Public Administration.

Kirchheimer, Otto. (1966). The Transformation of Western European Party Systems. In LaPalombara, Joseph & Myron Weiner. (Eds.), Political Parties and Political Development (pp. 177-200). Princeton, NJ: Princeton University Press.

Kitschelt, Herbert. (1995). The Radical Right in Western Europe. Ann Arbor, MI: The University of Michigan Press.

Klesner, Joseph L. (2005). Electoral Competition and the New Party System in Mexico. Latin American Politics and Society, 47(2), 103-42.

Knack, Stephen. (1992). Civic Norms, Social Sanctions, and Voter Turnout. Rationality and Society, 4(2), 133-56.

Koford, Kenneth & Linda Heckert. (1987). Determinants of the Number of Legislative Parties: Evidence from Postwar France. In Holler, M. J. (Ed.), The Logic of Multiparty Systems (pp. 371-380). Leiden: Martinus Nijhoff.

Kodmani, Bassma. (2005, Oct). The Dangers of Political Exclusion: Egypt’s Islamist Problem. (Democracy and Rule of Law Paper No. 63). Carnegie Endowment for International Peace.

413

Krasner, Stephen D. (1989). Sovereignty: An Institutional Perspective. In James Caporaso (Ed.). The Elusive State (pp. 69-96), Newbury Park: Sage.

Kreckel, Reinhard. (1994). Soziale Integration und nationale Identität. Berliner Journal für Soziologie, 4, 13-20.

Kweka, A. N. (1997). The Pemba Factor in the 1995 General Elections. In Omari, C. K. (Ed.). The Right to Choose a Leader: Reflections on the 1995 Tanzanian General Elections (pp. 128-160). Dar es Salaam: Dar es Salaam University Press.

Kydd, J. & R. Christiansen. (1982). Structural Change in Malawi Since Independence: Consequences of a Development Strategy Based on Large Scale Agriculture. World Development, 10(5), 355-375.

Lakatos, Imre & Alan Musgrave. (1972). Criticism and the Growth of Knowledge. Cambridge: Cambridge University Press.

Lane, Jan-Erik & Svante O. Ersson. (1999). The New Institutional Politics: Performance and Outcomes. New York: Routledge.

Lange, Siri. (1995). From Nation-Building to Popular Culture: The Modernization of Performance in Tanzania. Bergen, Norway: Chr. Michelsen Institute.

Langston, Joy. (2002). Breaking out is Hard to do: Exit, Voice, and Loyalty in Mexico’s One- Party Hegemonic Regime. Latin American Politics and Society, 44(3), 61-88.

Le Vine, Victor T. (1970). The Pragmatic-Pluralistic Pattern: Cameroun. In Coleman, James S. & Carl G. Rosberg, Jr. (Eds.). Political Parties and National Integration in Tropical Africa (pp. 132-184). Berkeley, CA: University of California Press.

Lee, Chris, Sandra Maline and Will H. Moore. (2000). Coercion and Protest: An Empirical Test Revisited. In C.A. Davenport (Ed.). Paths to Sate Repression: Human Rights Violations and Contentious Politics (pp. 127-147). Boulder: Rowman & Littlefield.

Lema, Elieshi, Marjorie Mbilinyi & Rakesh Rajani (Eds.). (2004). Nyerere on Education: Selected Essays and Speeches 1954-1998. Dar es Salaam: Mwalimu Nyerere Foundation.

Lemarchand, René. (1994). Burundi: Ethnocide as Discourse and Practice. New York: Woodrow Wilson Center Press.

Lerner, Daniel. (1958). The Passing of Traditional Society. Glencoe: The Free Press

Levi, Margaret. (1997). Consent, Dissent and Patriotism. New York: Cambridge University Press.

414

Levitsky, Steven & Way, Lucan (2002). The Rise of Competitive Authoritarianism. The Journal of Democracy, 13(2), 55-65.

Lichbach, Mark Irving. (1987). Deterrence of Escalation? The Puzzle of Aggregate Studies of Repression and Dissent. Journal of Conflict Resolution, 31(2), 266-97.

Lindemann, Stephan & James Putzel. (2008). State Resilience in Tanzania. Paper presented at the 2008 Crisis States Research Center summer seminar. Retrieved May 2, 2008, from http://www.crisisstates.com/events/seminars%2007-08.htm.

Linz, Juan and Alfred Stepan. (1996). Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore: Johns Hopkins University Press.

Linz, Juan. (2000). Totalitarian and Authoritarian Regimes. Boulder, Colorado: Lynne Rienner.

Lipset, Seymour Martin. (1981). Political Man: The Social Bases of Politics. Baltimore: Johns Hopkins University Press.

Lipset, Seymour Martin & Stein Rokkan. (1967). Cleavage Structures, Party Systems and Voter Alignments: an Introduction. In Lipset, Seymour Martin & Stein Rokkan (Eds.), Party Systems and Voter Alignments (pp. 1-63). New York: The Free Press.

Liviga, Athumani. (1990). Morogoro Urban: Questioning the Manifesto. In Haroub, Othman; Immanuel K. Bavu & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 121- 134). Dar Es Salaam: Dar Es Salaam University Press.

Lodhi, Abdulaziz & David Westerlund. (1997). African Islam in Tanzania. In David Westerlund & I. Svanberg (Eds.). Islam Outside the Arab World (pp. 97-110). London: Curzon Press.

Lokuji, Alfred Sebit (1995, 7 September). Ethnic strife and political conflict as global issues (part 1). The Guardian.

Lonsdale, John. (1968). The Tanzanian Experiment. African Affairs, 67(269), 330-344.

Lorch, Donatella. (1995, 29 October). A joyful but Anxious Vote in Tanzania. New York Times. Retrieved June 13, 2008, from http://nytimes.com.

Luanda, N. N. (1994). 1990 Parliamentary Elections: Continuity in Change. In Mukandala, Rwekaza S. & Haroub Othman (Eds.), Liberalization and Politics: The 1990 Elections in Tanzania (pp. 257-68). Dar Es Salaam: Dar Es Salaam University Press.

Ludwig, Friefer. (1996). After Ujamaa: Is Religious Revivalism a Threat to Tanzania’s Stability? In Westerlund, David (Ed.). Questioning the State: The Worldwide Resurgence of Religion in Politics (pp. 216-236). London: Hurst & Company.

415

Lüsebrink, Hans- Jürgen. (2002). Historical Culture in (Post-) Colonial Context: The Genesis of National Identification Figures in Francophone Western Africa. In Friese, Heidrun (Ed). Identities: Time, Difference and Boundaries. New York: Berghahn Books.

MacDonald, Alexander. (1966). Tanzania: Young Nation in a Hurry. New York: Hawthorn Books.

Maguire, Andrew G. (1969). Toward ‘Uhuru’ in Tanzania: The Politics of Participation. London: Cambridge University Press.

Mahoney, James & Dietrich Rueschemeyer. (2003). Comparative Historical Analysis: Achievements and Agendas. In Mahoney, James & Dietrich Rueschemeyer (Eds.). Comparative Historical Analysis in the Social Sciences (pp. 3-36). Cambridge: Cambridge University Press.

Mahoney, James (2000). Strategies of Causal Inference in Small-N Analysis. Sociological Methods & Research, 28(4), 387-424.

Maliyamkono, T. L. (2000). The Political Plight of Zanzibar. Dar es Salaam: TEMA.

Maliyamkono, T. L. & F. E. Kanyongolo (Eds.). (2003). When Political Parties Clash. Dar es Salaam: TEMA.

Malkki, Liisa. (1992). National Geographic: The Rooting of Peoples and the Territorizations of National Identity among Scholars and Refuggess. Cultural Anthropology, 7(1), 24-55.

Malupenga, Amos. (2006, 20 September). Chiluba and his Crooks Must be Stopped. The Post (Lusaka).

Mamdani, Mahmood. (1975). Class Struggles in Uganda. Review of African Political Economy, 4, 26-61.

Mamdani, Mahmood. (1996). Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton, New Jersey: Princeton University Press.

Mansfield, Edward D. & Jack Snyder. (1995). Democratization and the Dangers of War. International Security, 20, 5-38.

Mansfield, Edward D. & Jack Snyder. (2007). The Sequencing Fallacy. Journal of Democracy, 18(3), 5-9.

Mapuri, Omar R. (1996). Zanzibar: the 1964 Revolution. Dar es Salaam: Tema.

March, James G. & Olsen, Johan P. (1995). Democratic Governance. New York: Free Press.

416

Marshall, Catherine. (1984). The Wrong Time for Mechanistics in Qualitative Research. Educational Researcher, 13, 26-28.

Martin, Robert. (1975). Untitled [Review of the book Socialism and Participation: Tanzania’s 1970 National Elections]. The Journal of Modern African Studies, 13, 528-532.

Marx, Karl and Frederick Engels. (2000). Manifesto of the Communist Party. Marx/Engels Internet Archive. Retrieved January 30, 2009, from http://www.marxists.org.

Masato, Masato. (2004, 2 February). Salmin slams divisive party members. The Daily News.

Masato, Masato. (2005, 29 April). Poll points to pure merit in coming polls. The Daily News.

Mascarenhas, Adolpho. (1979). After Villagization – What?. In Mwansasu, Bismarck & Cranford Pratt (Eds.). Toward Socialism in Tanzania (pp. 145-165). Toronto: University of Toronto Press.

Mauzy, Diane K. (1983). The 1982 General Elections in Malaysia: A Mandate for Change? Asian Survey, 4, 497-517.

Maxon, Robert M. The Colonial Roots. (1994). In Oyugi, Walter O. (Ed.), Politics and Administration in East Africa (pp. 33-61). Nairobi: East African Educational Publishers.

Mbao, Melvin. (1996). The Third Republic and Its 1991 Constitution. In Sichone, Owen B. & Bornwell Chikulo (Eds.), Democracy in Zambia: Challenges for the Third Republic (pp. 9- 23). Harare: Sapes Books.

Mbogoni, Lawrence E. Y. (2004). The Cross versus the Crescent: Religion and Politics in Tanzania from the 1880s to the 1990s. Dar es Salaam: Mkuki na Nyota.

Mbogora, Alfred (1995, 12 October). Kinana demands evidence. The Guardian.

Mburu, Stephen (2008, 17 August). ‘Secret’ Report Reveals Sponsors of Poll Violence. The Nation (Nairobi).

McCracken, John. (1998). Democracy and Nationalism in Historical Perspective: The Case of Malawi. African Affairs, 97, 231-249.

McCracken, John. (2002). The Ambiguities of Nationalism: Flax Musopole and the Northern Factor in Malawian Politics, 1956-1966. Journal of Southern African Studies, 28(1), 67-87.

McCarthy, JD & M.N. Zald. (1977). Resource Mobilization and Social Movements: A Partial Theory. American Journal of Sociology, 82, 1212-39.

McDonald, Ronald H. & J. Mark Ruhl. (1989). Party Politics and Elections in Latin America. Boulder: Westview Press.

417

McHenry, Dean E. Jr. (1994). Limited Choices: The Political Struggle for Socialism in Tanzania. Boulder, Colorado: Lynne Rienner.

Means, Gordon P. (1996). Soft Authoritarianism in Malaysia. The Journal of Democracy 4(4), 103-107.

Media Council of Tanzania. (2001). Tanzania Elections 2000 Media Monitoring Project. Dar es Salaam: Author.

Meillassoux, Claude. (1970). A Class Analysis of the Bureaucratic Process in Mali. The Journal of Development Studies, 6(2), 97-110

Meinhardt, Heiko and Nandini Patel. (2003). Malawi’s Process of Democratic Transition: An analysis of political developments between 1990 and 2003. Lilongwe: Konrad-Adenauer- Stiftung.

Messiant, Christine. (2001). The Eduardo Dos Santos Foundation: or, How Angola’s Regime is Taking Over Civil Society. African Affairs 100, 287-309.

Mhina, Amos (2001). Partisanship, Organization, Membership and Orientation Towards Political Change. In Mushi, Samuel S., Mukandala, Rwekaza S. & Baregu, Mwesiga L. (Eds.). Tanzania’s Political Culture: A Baseline Survey (pp. 186-213). Dar es Salaam: Department of Political Science and Public Administration.

Mhone, G. C. Z. (1987). Agriculture and Food Policy in Malawi: A Review. In T. Mkandawire & N. Bourevan (Eds.), The State and Agriculture in Africa (pp. 58-80). Dakar: CODESRIA.

Migdal, Joel. (1988). Strong Societies and Weak States: State-Society Relations and State Capabilities in the Third World. Princeton: Princeton University Press.

Migdal, Joel. (2001). State in Society: Studying how States and Societies Transform and Constitute One Another. Cambridge: Cambridge University Press.

Migot-Adholla, S. E. (1979). Rural Development Policy and Equality. In Barkan, Joel D. & John J. Okumu (Eds.). Politics and Public Policy in Kenya and Tanzania (pp. 154-178). New York: Praeger.

Miguel, Edward. (2004). Tribe or Nation? Nation Building and Public Goods in Kenya versus Tanzania. World Politics, 56, 327-362.

Mihangwa, Joseph (2005, October 20). Coalition government only way to avoid bloodshed in Zanzibar. The African.

Mkandawire, R. (1992). The Land Question and Agrarian Change in Malawi. In G. C. Z. Mhone (Ed.), Malawi at the Crossroads: The Post-Colonial Political Economy (pp. 102-33). Harare: Sapes Books.

418

Mkandawire, Thandika. (1999). Crisis Management and the Making of “Choiceless Democracies” in Africa. In Joseph, Richard (Ed.), State Conflict, and Democracy in Africa (pp. 119-136). Boulder: Lynne Rienner Publishers, Inc.

Mkapa: We must erase divisive legacy. (2005, October 20). Daily News.

Mlahagwa, Josiah R. (1999). Contending for the Faith: Spiritual Revival and the Fellowship Church in Tanzania. In Spear, Thomas & Isaria N. Kimambo (Eds.), East African Expressions of Christianity. Dar es Salaam: Mkuki na Nyota.

Mmuya, Max & Amon Chiligha. (1992). Toward Multiparty Politics in Tanzania. Dar es Salaam: Dar es Salaam University Press.

Mmuya, Max & Amon Chaligha. (1994). Political Parties and Democracy in Tanzania. Dar es Salaam: Dar es Salaam University Press.

Mohan, Jitendra. (1966). Varieties of African Socialism. The Socialist Register, 3, 220-266.

Molomo, Mpho. (1991). Botswana’s Political Process. In Molomo, Mpho and Brian T. Mokopakgosi (Eds.). Multiparty Democracy in Botswana (pp. 11-22). Harare: Sapes.

Molutsi, Patrick P. (1991). Political Parties and Democracy in Botswana. In Molomo, Mpho and Brian T. Mokopakgosi (Eds.). Multiparty Democracy in Botswana (pp. 5-9). Harare: Sapes.

Momba, Jotham C. (2003). Democratic Transition and the Crisis of an African Nationalist Party: UNIP, Zambia. In Salih, Mohamed M. A. (Ed.). African Political Parties: Evolution, Institutionalization and Governance (pp. 37-65). London: Pluto Press.

Morgenthau, Ruth Schachter. (1965). African Elections: Tanzania’s Contribution. Africa Report, 10, 12-16.

Mpangala, G. P. (1994). The Organization and Management of the 1990 Parliamentary Elections. In Mukandala, Rwekaza S. & Haroub Othman (Eds.), Liberalization and Politics: The 1990 Elections in Tanzania (pp. 36-53). Dar Es Salaam: Dar Es Salaam University Press.

Mpangala, Gaudens P. (2000). Ethnic Conflicts in the Region of the Great Lakes: Origins and Prospects for Change. Dar es Salaam: Institute of Kiswahili Research.

Mpinga, James & Jamila Cushnie. (1995, 30 July-2 August). Survey begs: Ben who?. The Express.

Muekalia, Domingos Jardo. (2004). Africa and China’s Strategic Partnership. Africa Security Review, 13(1), 5-11.

419

Mueller, Susanne D. (1980). Retarded Capitalism in Tanzania. The Socialist Register, 17, pp. 203-226.

Mukandala, Rwekaza. (1998). Decentralization and Democratization in Tanzania (Occasional Paper No. 46). Iowa City: University of Iowa International Programs.

Munishi, Gasper & Asha Rose Mtengeti-Migiro. (1990). One-Party Democracy. In Haroub, Othman; Immanuel K. Bavu & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 182-201). Dar Es Salaam: Dar Es Salaam University Press.

Murder most foul, again. (2003, September). Africa Confidential, 44(19), 1-3.

Mushi, Samuel S. (1971). Modernization by Traditionalization: Ujamaa Principles Revisited. Taamuli, 1(2), 13-29.

Mushi, Samuel S., Rwekaza S. Mukandala & Mwesiga L. Baregu. (Eds.). (2001). Tanzania’s Political Culture: A Baseline Survey. Dar es Salaam: University of Dar es Salaam, Department of Political Science and Public Administration.

Mutahaba, Gelase & Michael Okema. (1990). Tanzania Presidential Elections: 1962-1985. In Haroub, Othman; Immanuel K. Bavu, & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 65-77). Dar es Salaam: Dar Es Salaam University Press.

Mutalib, Hussin. (2000). Illiberal Democracy and the Future of Opposition in Singapore. Third World Quarterly, 21(2), 313-42.

Mvungi, S. A. D. & Amos Mhina. (1990). Dodoma Urban: Searching for a People’s M.P.? In Haroub, Othman; Immanuel K. Bavu, & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 103-120). Dar Es Salaam: Dar Es Salaam University Press.

Mwaikusa, J. T. (1996). The Limits of Judicial Enterprise: Judicial Powers in the Process of Political Change in Tanzania. Journal of African Law, 40, 243-255.

Mwakikagile, Godfrey. (2004). Tanzania Under Mwalimu Nyerere: Reflections on an African Statesman. United Kingdom: Fultus.

Mwanakatwe, Thobias (2005, October 29). Salma Kikwete warns opposition’s tactics. The Guardian.

Mwanjisi, Jamillah (1995, 10-13 September). Poor Mrema now needs dark glasses. The Express.

Mwansasu, Bismarck U. (1979). The Changing role of the Tanganyika African National Union. In Mwansasu, Bismarck & Cranford Pratt (Eds.). Toward Socialism in Tanzania (pp. 169- 192). Toronto: University of Toronto Press.

420

Mwakyembe, Harrison. (1990). Mbozi: The End of a Honeymoon. In Haroub, Othman; Immanuel K. Bavu, & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 135- 149). Dar Es Salaam: Dar Es Salaam University Press.

Mwanjisi, Jamillah (1995, 10-13 September). Poor Mrema now needs dark glasses. The Express.

Mwita, Sosthenes. (2005, 10 September). CCM decries divisive campaigns. The Daily News.

Mørck, Anna Birgitte. (2006). The United Republic of Tanzania: Presidential and Parliamentary Elections December 2005. Oslo: The Norwegian Centre for Human Rights.

National Bureau of Statistics Tanzania. (2002). Household Budget Survey 2000/01. Dar es Salaam: Author.

National Electoral Commission. (1997). The Report of the National Electoral Commission on the 1995 Presidential and Parliamentary Elections. Dar es Salaam: Author.

National Electoral Commission. (2001). The Report of the National Electoral Commission on the 2000 Presidential, Parliamentary and Councillors’ Elections. Dar es Salaam: Author.

National Electoral Commission. (2006). The Report of the National Electoral Commission on the 2005 Presidential, Parliamentary and Councillors’ Elections. Dar es Salaam: Author.

National Election Monitoring Unit. (1993). The Multi-Party General Elections in Kenya, 1992. Nairobi: Author.

Ndegwa, Stephen N. (1996). The Two Faces of Civil Society: NGOs and Politics in Africa. West Hartford: Kumarian Press, Inc.

Nellis, John. (1972). A Theory of Ideology: The Tanzanian Example. Nairobi: Oxford University Press.

Newbury, Catharine. (1988). The Cohesion of Oppression: Clientship and Ethnicity in Rwanda – 1860-1960. New York: Columbia University Press.

Njozi, Hamza Mustafa. (2003). Muslims and the State in Tanzania. Dar es Salaam: Dar es Salaam University Muslims Trusteeship.

Norris, Pippa. (2003). Electoral Engineering: Voting Rules and Political Behavior. New York: Cambridge University Press.

North, Douglass. (1990). Institutions, Institutional Change and Economic Performance: Cambridge: Cambridge University Press.

421

Nugent, Paul. (2004). Africa Since Independence: A Comparative History. New York: Palgrave Macmillan.

Nursey-Bray, P. F. (1980). Tanzania: The Development Debate. African Affairs, 79(314), pp. 55-78.

Nyagetera, Bartholomew M. (1995). Investment, Foreign Aid and Self-Reliance in Tanzania: A State-of-the-Art Review (Discussion Paper Series No. 001). Dar es Salaam: Economic and Social Research Foundation.

Nyambura-Mwaura, Helen (2007, December 11). Tribalism and Kenya’s Election. Reuters.

Nyamungumi, Penzi (2005a, November 22). VP deplores hate politics. The Daily News.

Nyamungumi, Penzi (2005b, November 25). CCM Promises to uphold multiparty democracy. The Daily News.

Nyanje, Peter. (2005, 1 January). Politicians abusing Pluralism, says Mkapa. The Guardian.

Nyanje, Peter (2005b, October 23). CCM Dar warns CUF saboteurs. The Sunday Citizen.

Nyanje, Peter (2005c, October 24). You see, they are beating people already. The Citizen.

Nyerere, Julius. (1962). Ujamaa: The Basis of African Socialism. Dar es Salaam.

Nyerere, Julius. (1966). Freedom and Unity: Uhuru na Umoja. London: Oxford University Press.

Nyerere, Julius. (1968). Freedom and Socialism: Uhuru na Ujamaa. Oxford: Oxford University Press.

Nyerere’s speech wins Dar hearts. (1995, 2 September). The Guardian.

Nzomo, Maria. (2003). Civil Society in the Kenyan Political Transition: 1992-2002. In Oyugi, Walter O., Peter Wanyande & C. Odhiambo Mbai (Eds.), The Politics of Transition in Kenya: From KANU to NARC (pp. 180-211). Nairobi: Heinrich Böll Foundation.

O’Donnell, Guillermo A. & Schmitter, Philippe C. (1986). Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies. Baltimore: The Johns Hopkins University Press.

Offe, Claus. (1985). New Social Movements: Challenging the Boundaries of Institutional Politics. Social Research, 52(4), 817-68.

422

Okema, Michael. (1990). Some Salient Changes in the Tanzania Parliamentary Systems. In Haroub, Othman; Immanuel K. Bavu & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 37-57). Dar Es Salaam: Dar Es Salaam University Press.

Okema, Michael. (1996). Political Culture of Tanzania. Lewiston, New York: The Edwin Mellen Press.

Okoko, Kimse A. B. (1987). Socialism and Self-Reliance in Tanzania. London: KPI.

Olaniyan, Tejumola. (1995). Scars of Conquest Mask of resistance: The Invention of Cultural Identities in African, African-American, and Caribbean Drama. New York: Oxford University Press.

Organization for Security and Cooperation in Europe. (2005). Election Observation Handbook. Warsaw: Author.

Othman, Haroub. (1990). The Political and Constitutional Developments of Zanzibar and the Case Studies of the 1995 Zanzibar General Elections. In Haroub, Othman; Immanuel K. Bavu, & Michael Okema (Eds.), Tanzania: Democracy in Transition (pp. 150-181). Dar Es Salaam: Dar Es Salaam University Press.

Othman, Haroub. (2006). Tanzania: The Withering Away of the Union? In Peter, Chris Maina & Haroub Othman (Eds.), Zanzibar and the Union Question (pp. 34-72). Zanzibar: Zanzibar Legal Services Center.

Paffenholz, Thania and Christoph Spurk. (2006). Civil Society, Civil Engagement, and Peacebuilding. Washington D.C.: World Bank. (Social Development Papers No. 36).

Papstein, Robert J. (1989). From Ethnic Identity to Tribalism: The Upper Zambezi Region of Zambia, 1830-1981. In Vail, Leroy (Ed.). The Creation of Tribalism in Southern Africa: London: James Currey.

Park, Chong-Min. (1991). Authoritarian Rule in South Korea: Political Support and Governmental Performance. Asian Survey, 31(8), 743-61.

Parker, Ian. (1979). Contradictions in the Transition to Socialism: the case of the National Development Corporation. In Mwansasu, Bismarck & Cranford Pratt (Eds.). Toward Socialism in Tanzania (pp. 46-71). Toronto: University of Toronto Press.

Pastor, Manuel Jr. & Carol Wise. (2005). The Lost Sexenio: Vincente Fox and the New Politics of Economic Reform in Mexico. Latin American Politics and Society, 47(4), 135-60.

Pempel, T. J. (1990). Uncommon Democracies: The One-Party Dominant Regimes. Ithaca: Cornell University Press.

423

Peter, Chris Maina. (1997). Incarcerating the Innocent: Preventive Detention in Tanzania. Human Rights Quarterly, 19, 113-135.

Peters, Guy B. (1998). Comparative Politics: Theory and Methods. New York: New York University Press.

Phiri, Kings and Kenneth R. Ross. (1998). From Totalitarianism to Democracy. In Kings Phiri and Kenneth R. Ross (Eds.). Democratization of Malawi: A Stocktaking (pp 45-59.). Blantyre: Kachere Books

Picard, Louis A. (1987). The Politics of Development in Botswana: A Model for Success?. Boulder, Colorado: Lynne Rienner.

Pierson, Paul. (2004). Politics in Time: Historical Institutions, and Social Analysis. Princeton, NJ: Princeton University Press.

Piven, Frances Fox & Richard A. Cloward. (2000). Why Americans Still Don’t Vote: And Why Politicians Want it that Way. Boston: Beacon Press.

Polletta, Francesca & James M. Jasper. (2001). Collective Identity and Social Movements. Annual Review of Sociology, 27, 283-305.

Posner, Daniel N. (1995). Malawi’s New Dawn. Journal of Democracy, 6(1), 131-145.

Posner, Daniel N. (2005). Institutions and Ethnic Politics in Africa. New York: Cambridge University Press.

Posusney, Marsha Pripstein. (2004). Enduring Authoritarianism: Middle East Lessons for Comparative Theory. Comparative Politics, 36(2), 127-138.

Poulantzas, Nicos. (1973). Political Power and Social Classes. London: New Left Books.

Pratt, Cranford. (1976). The Critical Phase in Tanzania, 1945-67: Nyerere and the Emergence of a Socialist Strategy. Cambridge: Cambridge University Press.

Pratt, Cranford. (1979). Tanzania’s Transition to Socialism: Reflections of a Democratic Socialist. In Mwansasu, Bismarck & Cranford Pratt (Eds.). Toward Socialism in Tanzania (pp. 193-236). Toronto: University of Toronto Press.

Pratt, Cranford. (1999). Julius Nyerere: Reflections on the Legacy of His Socialism. Canadian Journal of African Studies, 33(1), 137-152.

Pratt, Cranford. (2002). The Ethical Foundations of Julius Nyerere’s Legacy. In David McDonald & Aunice N. Sahle (Eds.). The Legacies of Julius Nyerere: Influence on Development Discourse and Practice in Africa (pp. 39-52). Trenton, NJ: Africa World Press.

424

Proctor, J. H. (1968). The House of Chiefs and the Political Development of Botswana. The Journal of Modern African Studies, 6(1), 59-79.

Przeworski, Adam & John Sprague. (1986). Paper Stones: A History of Electoral Socialism. Chicago: The University of Chicago Press.

Putnam, Robert. (1993). Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press.

Ragin, Charles C. (1987). The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies. Berkeley, California: The University of California Press.

Raikes, P. L. (1975). Ujamaa and Rural Socialism. Review of African Political Economy, 3, 32- 52.

Randell, Vicky & Lars Svåsand. (2002). Political Parties and Democratic Consolidation in Africa. Democratization, 9(3), 30-52.

Rasmussen, Thomas. (1969). Political Competition and One-Party Dominance in Zambia. The Journal of Modern African Studies, 7(3), 407-424.

Rawls, John. (1997). The Idea of Public Reason. In Bohman, James and William Rehg (Eds.). Deliberative Democracy: Essays on Reason and Politics (pp. 93-141). Cambridge: The MIT Press.

Republic of Kenya (1965). African Socialism and its Application to Planning in Kenya. Nairobi: Author.

Research and Education for Democracy in Tanzania. (1997). The 1993 Neighbourhood and Village Council Elections in Tanzania (REDET publications no. 3). Dar es Salaam: Department of Political Science and Public Administration, University of Dar es Salaam.

Research and Education for Democracy in Tanzania. (1999a). The First Opinion Poll: The March 1999 Opinion Poll Findings and Discussion. Dar es Salaam: Author.

Research and Education for Democracy in Tanzania. (1999b). The Second Opinion Poll: The December 1999 Opinion Poll. Dar es Salaam: Author.

Research and Education for Democracy in Tanzania. (2000). The Third Opinion Poll: The October 2000 Elections. Dar es Salaam: Author.

Research and Education for Democracy in Tanzania. (2001). The Fourth Opinion Poll: Citizenry Post Election Attitudes. Dar es Salaam: Author.

425

Research and Education for Democracy in Tanzania. (2004, November). People’s Opinion and Preferences for the 2005 General Elections (REDET Opinion Poll Report No. 8). Dar es Salaam: Author.

Roe, Alan R. (1970). [Review of the book Tanzania Second Five-Year Plan for Economic and Social Development 1969-1974]. The Economic Journal, 80(318), 396-398.

Ross, Andrew. (1967). White Africa’s Black Ally. New Left Review, 1(45), 85-95.

Rose, Richard & Doh Chull Shin. (2001). Democratization Backwards: The Problem of Third- Wave Democracies. British Journal of Political Science, 31, 331-54.

Rothchild, Donald. (1985). State-ethnic relations in Middle Africa. In Carter, Gwendolen and Patrick O’Meara (Eds.). African Independence: The First Twenty-Five Years. Bloomington, IN: Indiana University Press.

Rothchild, Donald. (1997). Ethnic Insecurity, Peace Agreements, and State Building. In Joseph, Richard (Ed.), State Conflict, and Democracy in Africa (319-337). Boulder: Lynne Rienner Publishers, Inc.

Safran, William. (2008). Names, Labels, and Identities: Sociopolitical Context and the Question of Ethnic Categorization. Identities, 15, 437-461.

Salih, M. A. Mohamed. (2001). African Democracies and African Politics. London: Pluto Press.

Samaraweera, Dilshani (2008, July 27). War costs Enhance GDP. The Sunday Times. Retrieved March 30, 2009, from http://sundaytimes.lk.

Samatar, Ismail Abdi. (1999). An African Miracle: State and Class Leadership and Colonial Legacy in Botswana Development. Portsmouth, NH: Heinemann.

Sandbrook, Richard. (2000). Closing the Circle: Democratization and Development in Africa. Toronto: Between the Lines.

Sartori, Giovanni. (1976). The Comparative Study of Political Parties. New York: St. Martin’s.

Sarwatt, Sebastian & Esther Mollel (2000). Country Pasture/Forage Resource Profiles: United Republic of Tanzania. Retrieved November 23, 2007, from United Nations Food and Agricultural Organization, http://www.fao.org

Saul, John S. (1974). The State in Post-Colonial Societies: Tanzania. The Socialist Register, 11, 349-372.

Scarritt, James R. (1971). Elite Values, Ideology, and Power in Post-Independence Zambia. African Studies Review, 14(1), 31-54.

426

Scarritt, James & Susan McMillan. (2000). Protests, Democratization, and Human Rights in Africa. In Christian Davenport (Ed.). Paths to State Repression: Human Rights Violations and Contentious Politics (pp.195-216). Rowman & Littlefield

Schedler, Andreas. (2002). The Menu of Manipulation. The Journal of Democracy 13(2), 36-50.

Schmitter, Philippe C. & Terry Lynn Karl. (1991). What Democracy is…and is Not. Journal of Democracy, 2(3), 75-88.

Schubert, Gunter. (2004). Taiwan’s Political Parties and National Identity: The Rise of an Overarching Consensus. Asian Survey, 44(4), 534-554.

Scott, Richard W. (1995). Institutions and Organizations. Thousand Oaks, California: Sage.

Semwaika, Richard. (1995, 9-12 April). Ikulu race ignites political scene. The Express.

Shariff, A. (1994). The Union and the Struggle for Democracy in Zanzibar. In Mukandala, Rwekaza S. & Haroub Othman (Eds.), Liberalization and Politics: The 1990 Elections in Tanzania (pp. 146-174). Dar Es Salaam: Dar Es Salaam University Press.

Shitundu, J. L. M. (2002). Industrialization and the Nyerere Legacy: Missing Links and Lessons for Future Policy in Tanzania. In A. V. Y. Mbelle, G. D. Mjema, and A. A. L. Kilindo (Eds.). The Nyerere Legacy and Economic Policy Making in Tanzania (pp. 134-149). Dar es Salaam: Dar es Salaam University Press.

Shivji, Issa G. (1976). Class Struggles in Tanzania. London: Monthly Review Press.

Shivji, Issa G. (1991). Contradictory Developments in the Teaching and Practice of Human Rights Law in Tanzania. Journal of African Law, 35, 116-127.

Shivji, Issa G. (2006). Let the People Speak: Tanzania Down the Road to Neo-Liberalism. CODESRIA

Sichone, Owen B. & Neo R. Simutanyi. (1996). The Ethnic and Regional Questions, Ethnic Nationalism and the State in Zambia: The Case of Bulozi, 1964-1994. In Sichone, Owen B. & Bornwell Chikulo (Eds.), Democracy in Zambia: Challenges for the Third Republic (pp. 173-195). Harare: Sapes Books.

Sichone, Owen B. (1996). Problems of the State and Civil Society. In Sichone, Owen B. & Bornwell Chikulo (Eds.), Democracy in Zambia: Challenges for the Third Republic (pp. 93- 108). Harare: Sapes Books.

Silverman, David. (1970). The Theory of Organizations: A Sociological Framework. London: Heinemann.

427

Simkins, Charles. (1999). Stability and Competitiveness in the Political Configurations of Semi- Developed Countries. In Giliomee, Hermann & Simkins, Charles. (Eds.), The Awkward Embrace: One-Party Domination and Democracy (pp. 47-60). Amsterdam: Harwood Academic Publishers.

Simutanyi, Neo. (2005, October 5). Political Parties and Party System in Zambia. Background paper prepared for Friedrich Ebert Stiftung, Zambia.

Skidmore, Max J. (1993). Ideologies: Politics in Action (2nd ed.). United States: Harcourt Brace Jovanovich, Inc.

Skocpol, Theda. (1979). States and Social Revolutions: A Comparative Analysis of France, Russia, and China. Cambridge, UK: Cambridge University Press.

Skowronek, Stephen. (1997). The Politics Presidents Make: Leadership from John Adams to Bill Clinton. Cambridge, MA: Harvard University Press.

Smith, Benjamin. (2004). Oil Wealth and Regime Survival in the Developing World, 1960- 1999. American Journal of Political Science 48(2), 232-246.

Sobel, Michael E. (1995). Causal Inference in the Social and Behavioral Sciences. In Arminger, Gerhard, Clifford C. Clogg, and Michael E. Sobel (Eds.). Handbook of Statistical Modeling for the Social and Behavioral Sciences (pp. 1-31). New York: Plenum.

Socialism ‘crucified to death’. (1995, 20 September). The Guardian.

Sokhulu, Themba M. (2004, July). Is Botswana Advancing or Regressing in Its Democracy? (Occasional Paper 1/2004). South Africa: ACCORD.

Spencer, John. (1985). The Kenya African Union. London: KPI

Straub, Jürgen. (2002). Personal and Collective Identity: A Conceptual Analysis. In Friese, Heidrun. (Ed.), Identities: Time, Difference and Boundaries (pp. 56-76). New York: Berghahn Books.

Stephen, Dassu. (2005, 12 December). Mrema promises to revamp agriculture. The Daily News.

Stinchombe, Arthur. (1968). Constructing Social Theories. New York: Harcourt, Brace, and World.

Story, Dale. (1986). The Mexican Ruling Party: Stability and Authority. New York: Praeger.

Sued, Hilal (2000, 26 August). Setting a Muslim to catch a Muslim. The African.

Sundet, Geir (1996). Democracy in Transition: 1995 Elections n Tanzania (Human Rights Report No. 8). Oslo: Norwegian Institute of Human Rights.

428

Suspicions raised over Kazakh deaths. (2006, February 14). The BBC. Retrieved July August 5, 2007, from http://news.bbc.co.uk/2/hi/asia-pacific/4712674.stm

Szeftel, Morris. (2000). Clientelism, Corruption & Catastrophe. Review of African Political Economy, 27(85), 427-41.

Tambiah, Stanley Jeyaraja. (1985). Culture, Thought, and Social Action: An Anthropological Perspective. Cambridge, MA: Harvard University Press.

Tangri, Roger K. (1968). The Rise of Nationalism in Colonial Africa: The Case of Colonial Malawi. Comparative Studies in Society and History, 10(2), 142-61.

Tanzania Bankers Association. (2005, March 8). The Land Reform in Tanzania: Opportunities for Agriculture and Mortgage Finance. Paper presented at the 2005 Tanzania Bankers Association conference. Retrieved January 23, 2007, from http://www.tanzaniabankers.org/land%20reform.pdf.

Tanzania Election Media Monitoring-2005. (2006). Final Report. Dar es Salaam: Author.

Tanzania Election Monitoring Committee. (1997). The 1995 Elections in Tanzania. Dar es Salaam: Author.

Tanzania Election Monitoring Committee (2001). The 2000 General Elections in Tanzania. Dar es Salaam: Author.

Tanzania Election Monitoring Committee. (2006). The 2005 Elections in Tanzania. Dar es Salaam: Author.

Tarimo, Judica. (2003, December 13). TPDF: Unruly soldiers to be punished. The Guardian. Retrieved December 20, 2006, from http://www.ippmedia.com/ipp/old/2003/12/13.html.

Tarrow, Sidney. (1998). Power in Movement: Social Movements and Contentious Politics. New York: Cambridge University Press.

Taylor, Scott. (1999). Race, Class, and Neopatrimonialism in Zimbabwe. In Joseph, Richard (Ed.), State Conflict, and Democracy in Africa (pp. 239-266). Boulder: Lynne Rienner Publishers, Inc.

Temba, Pudenciana. (2005, 2 September). Thousands witness Kikwete launch. The Daily News.

Tenga, R. (1996). Land Tenure in Tanzania. In N. Leader-Williams, J.A. Kayera and G.L. Overton (Eds.) Community-based Conservation in Tanzania (pp. 19-28). Gland, Switzerland: International Union for Conservation of Nature.

Thomas, Juma (1995, 3 September). NCCR crowds take Dar roads. The Guardian.

429

Thomas, Simon. (1975). Economic Developments in Malawi Since Independence. Journal of Southern African Studies, 2(1), 30-51.

Throup, David. (1993). Elections and Political Legitimacy in Kenya. Journal of the International African Institute, 63, 371-396.

Throup, David & Charles Hornsby. (1998). Multi-Party Politics in Kenya. Nairobi: East African Educational Publishers.

Trinh, T. Minh-ha. (1989). Women, Native, Other: Writing Postcoloniality and Feminism. Bloomington, IN: Indiana University Press.

Tordoff, William. (1988). Political Parties in Zambia. In Randall, Vicky (Ed.), Political Parties in the Third World (pp. 7-29).

Tordoff, William. (1993). Government and Politics in Africa (3rd ed.). Hong Kong: MacMillan Press.

Tsebelis, George and John Sprague. (1989). Coercion and Revolution: Variations on a Predator- Prey Model. Mathematical and Computer Modeling, 12, 547-559.

United Nations Food and Agriculture Organization. (2006). Fertilizer Use Intensity [Data file]. Retrieved August 14, 2007, from Earthtrends, http://earthtrends.wri.org.

United Republic of Tanzania. (1992). The Presidential Commission on Single Party or Multiparty System in Tanzania, 1991. (Vol. 1). Dar es Salaam: Author.

United Republic of Tanzania. (1998). The Constitution of the United Republic of Tanzania of 1977. Dar es Salaam: Author.

U.S. Department of State. (1994a). 1993 Country Reports on Human Rights Practices: Botswana. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994b). 1993 Country Reports on Human Rights Practices: Cameroon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994c). 1993 Country Reports on Human Rights Practices: Gabon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994d). 1993 Country Reports on Human Rights Practices: Kazakhstan. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994e). 1993 Country Reports on Human Rights Practices: Kenya. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

430

U.S. Department of State. (1994f). 1993 Country Reports on Human Rights Practices: Malaysia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994g). 1993 Country Reports on Human Rights Practices: Tanzania. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994h). 1993 Country Reports on Human Rights Practices: Togo. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1994i). 1993 Country Reports on Human Rights Practices: Zambia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995a). 1994 Country Reports on Human Rights Practices: Botswana. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995b). 1994 Country Reports on Human Rights Practices: Cameroon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995c). 1994 Country Reports on Human Rights Practices: Gabon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995d). 1994 Country Reports on Human Rights Practices: Kazakhstan. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995e). 1994 Country Reports on Human Rights Practices: Kenya. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995f). 1994 Country Reports on Human Rights Practices: Malaysia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995g). 1994 Country Reports on Human Rights Practices: Tanzania. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995h). 1994 Country Reports on Human Rights Practices: Togo. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1995i). 1994 Country Reports on Human Rights Practices: Zambia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996a). 1995 Country Reports on Human Rights Practices: Botswana. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996b). 1995 Country Reports on Human Rights Practices: Cameroon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

431

U.S. Department of State. (1996c). 1995 Country Reports on Human Rights Practices: Gabon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996d). 1995 Country Reports on Human Rights Practices: Kazakhstan. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996e). 1995 Country Reports on Human Rights Practices: Kenya. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996f). 1995 Country Reports on Human Rights Practices: Malaysia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996g). 1995 Country Reports on Human Rights Practices: Tanzania. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996h). 1995 Country Reports on Human Rights Practices: Togo. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1996i). 1995 Country Reports on Human Rights Practices: Zambia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997a). 1996 Country Reports on Human Rights Practices: Botswana. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997b). 1996 Country Reports on Human Rights Practices: Cameroon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997c). 1996 Country Reports on Human Rights Practices: Gabon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997d). 1996 Country Reports on Human Rights Practices: Kazakhstan. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997e). 1996 Country Reports on Human Rights Practices: Kenya. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997f). 1996 Country Reports on Human Rights Practices: Malaysia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997g). 1996 Country Reports on Human Rights Practices: Tanzania. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1997h). 1996 Country Reports on Human Rights Practices: Togo. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

432

U.S. Department of State. (1997i). 1996 Country Reports on Human Rights Practices: Zambia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998a). 1997 Country Reports on Human Rights Practices: Botswana. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998b). 1997 Country Reports on Human Rights Practices: Cameroon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998c). 1997 Country Reports on Human Rights Practices: Gabon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998d). 1997 Country Reports on Human Rights Practices: Kazakhstan. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998e). 1997 Country Reports on Human Rights Practices: Kenya. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998f). 1997 Country Reports on Human Rights Practices: Malaysia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998g). 1997 Country Reports on Human Rights Practices: Tanzania. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998h). 1997 Country Reports on Human Rights Practices: Togo. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1998i). 1997 Country Reports on Human Rights Practices: Zambia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999a). 1998 Country Reports on Human Rights Practices: Botswana. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999b). 1998 Country Reports on Human Rights Practices: Cameroon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999c). 1998 Country Reports on Human Rights Practices: Gabon. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999d). 1998 Country Reports on Human Rights Practices: Kazakhstan. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999e). 1998 Country Reports on Human Rights Practices: Kenya. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

433

U.S. Department of State. (1999f). 1998 Country Reports on Human Rights Practices: Malaysia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999g). 1998 Country Reports on Human Rights Practices: Tanzania. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999h). 1998 Country Reports on Human Rights Practices: Togo. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (1999h). 1998 Country Reports on Human Rights Practices: Zambia. Retrieved January 20, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000a). 1999 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000b). 1999 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000c). 1999 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000d). 1999 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000e). 1999 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000f). 1999 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000g). 1999 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000h). 1999 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2000h). 1999 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001a). 2000 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001b). 2000 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

434

U.S. Department of State. (2001c). 2000 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001d). 2000 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001e). 2000 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001f). 2000 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001g). 2000 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001h). 2000 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2001i). 2000 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002a). 2001 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002b). 2001 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002c). 2001 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002d). 2001 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002e). 2001 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002f). 2001 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002g). 2001 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2002h). 2001 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

435

U.S. Department of State. (2002i). 2001 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003a). 2002 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003b). 2002 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003c). 2002 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003d). 2002 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003e). 2002 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003f). 2002 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003g). 2002 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003h). 2002 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2003i). 2002 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004a). 2003 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004b). 2003 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004c). 2003 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004d). 2003 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004e). 2003 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

436

U.S. Department of State. (2004f). 2003 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004g). 2003 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004h). 2003 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2004i). 2003 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005a). 2004 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005b). 2004 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005c). 2004 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005d). 2004 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005e). 2004 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005f). 2004 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005g). 2004 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005h). 2004 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2005i). 2004 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006a). 2005 Country Reports on Human Rights Practices: Botswana. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006b). 2005 Country Reports on Human Rights Practices: Cameroon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

437

U.S. Department of State. (2006c). 2005 Country Reports on Human Rights Practices: Gabon. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006d). 2005 Country Reports on Human Rights Practices: Kazakhstan. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006e). 2005 Country Reports on Human Rights Practices: Kenya. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006f). 2005 Country Reports on Human Rights Practices: Malaysia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006g). 2005 Country Reports on Human Rights Practices: Tanzania. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006h). 2005 Country Reports on Human Rights Practices: Togo. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

U.S. Department of State. (2006i). 2005 Country Reports on Human Rights Practices: Zambia. Retrieved April 8, 2007, from http://www.state.gov/g/drl/hr/

Vail, Leroy & Landeg White. (1989). Tribalism in the Political History of Malawi. In Vail, Leroy (Ed.). The Creation of Tribalism in Southern Africa (pp. 151-92). London: James Currey.

Van de Walle, Nicolas. (2002). Africa’s Range of Regimes. Journal of Democracy, 13(2), 66- 80.

Van de Walle, Nicolas. (2003). Presidentialism and Clientelism in Africa’s emerging Party Systems. Journal of Modern African Studies, 41(2), 297-321.

Van Dijk, Richard A. (1992). Young Puritan Preachers in Post-Independence Malawi. Africa, 62(2), 159-81.

Van Dijk, Teun A. (2004). Politics, Ideology and Discourse. Encyclopedia of Language and Linguistics, Section Language and Politics. Retrieved July 2, 2008 from http://www.discursos.org/unpublished%20articles/Politics,%20ideology%20and%20discour se%20(ELL).htm.

Van Dijk, Teun A. (2008). [Review of the book Manipulation and ideologies in the twentieth century]. Language in Society 37, 294-299.

Van Donge, Jan Kees. (1995). Kamuzu’s Legacy: The Democratization of Malawi: or Searching for the Rules in the Game in African Politics. African Affairs, 94(375), 227-257.

438

Van Donge, Jan Kees & Athumani J. Liviga. (1989). The 1985 Tanzanian Parliamentary Elections: A Conservative Election. African Affairs, 88, 47-62.

Wang, Jian-Ye. (2007). What Drives China’s Growing Role in Africa? IMF Working Papers, 07/211.

Wanjohi, Nick G. (2003). Sustainability of Political Parties in Kenya. In Salih, Mohamed M. A. (Ed.). African Political Parties: Evolution, Institutionalization and Governance (pp. 239- 256). London: Pluto Press.

Wantchekon, Leonard. (2001). Clientelism and Voting Behavior: Evidence from a Field Experiment in Benin. World Politics, 55, 399-422.

Warren, Mark E. (1989). What is Political Theory/Philosophy? Political Science and Politics, 22, 606-612.

Weldon, Jeffrey. (1997). The Political Sources of Presidencialismo in Mexico. In Scott Mainwaring and Matthew S. Shugart. (Eds.). Presidentialism and Democracy in Latin America (pp. 225-58). Cambridge: Cambridge University Press.

Wendt, Alexander. (1994). Collective Identity Formation and the International State. The American Political Science Review, 88(2), 384-396.

Wendt, Alexander. (1999). Social Theory of International Politics. Cambridge, UK: Cambridge University Press.

Whitaker, Beth. Shades of Compliance: Africa and the Counter-Terrorism Regime. Paper presented at the annual meeting of the American Political Science Association, Hyatt Regency Chicago and the Sheraton Chicago Hotel and Towers, Chicago, IL, Aug 30, 2007.

Whitehead, Richard. (2000). The Institutionalization of the Tanzanian Opposition Parties: How Stable Are They?, Unpublished master’s thesis, University of Bergen, Bergen, Norway.

Widner, Jennifer A. (1992). The Rise of a Party-State in Kenya: From “Harambee” to “Nyayo!”. Berkeley, CA: University of California Press.

Wikicommons. (2008). Regions of Tanzania. [Image File]. Retrieved March 1, 2009 from www.wikimedia.org

World Bank. (1987). World Development Report. New York: Oxford University Press.

World Bank. (2005). The Political Economy of Tanzania. Washington, D.C.: Author.

World Bank. (2007a). World Development Indicators Online: Employment in Agriculture. [Data file]. Retrieved December 5, 2007 from www.worldbank.org.

439

World Bank. (2007b). World Development Indicators Online: Total External Debt. [Data file]. Retrieved December 1, 2007 from www.worldbank.org.

World Bank. (2007c). World Development Indicators Online: Aid (% of GNI). [Data file]. Retrieved December 1, 2007 from www.worldbank.org.

World Bank. (2007d). World Development Indicators Online: Gini Index. [Data file]. Retrieved December 2, 2007 from www.worldbank.org.

World Bank. (2007e). World Development Indicators Online: Share of total Income. [Data file]. Retrieved December 2, 2007 from www.worldbank.org.

Yom, Sean L. and Mohammad H. Al-Momani. (2008). The International Dimensions of Authoritarian Regime Stability: Jordan in the Post-Cold War Era. Arab Studies Quarterly, Winter, 2008

You’ll live to regret, Mrema warns. (1995, 5 September). The Guardian.

Zald, Mayer N. & John D. McCarthy. (1987). Social Movements in an Organizational Society: New Brunswick, NJ: Transaction Books.

Z’bar will not turn into another Rwanda. (2000, 25 August). The African.

Zolberg, Aristide R. (1966). Creating Political Order: The Party States of West Africa. Chicago: Rand McNally.

Zucker, Lynne G. (1988). Where do Institutional Patterns Come From? Organizations as Actors in Social Systems. In Zucker, Lynne G. (Ed.), Institutional Patterns and Organizations: Culture and Environment (pp. 23-49). Cambridge: Ballinger Publishing Company.

Zakaria, Fareed. (2007). The Rise of Illiberal Democracy. Foreign Affairs, 76, 22-43.

Zwass, Adam. (1996). From Failed Communism to Underdeveloped Capitalism: Transformation of Eastern Europe, the Post-Soviet Union, and China. New York: M.E. Sharpe.

440

APPENDIX A DATA FROM COUNTRY REPORTS ON HUMAN RIGHTS PRACTICES

Since the United States Department of State provides one of the most systematic and uniform cross sectional accounts of human rights conditions, I found it valuable to inform my analysis of political repression by comparing State Department human rights reports from a selection of countries from 1994 through 2006. With the limited exception of Kenya, each country can be described as a case of prolonged party tenure amid multiparty elections. The following tables within this appendix list the raw data from the findings of my review of these reports.

Several methodological points need to be noted. First, within most of the categories, I counted cases rather than individuals effected. For example, in Kenya during 1993, there were large death tolls as a result of politically motivated ethnic conflict. This conflict is captured in the category entitled “Deaths in Political Conflict”. However, since death tolls are often unreported, I found it necessary to count deaths in political conflict as an instance rather than by the number of deaths. The parenthetical “cases” highlights the categories where data reflects instances.

Secondly, the figures accumulated from the reports should not be taken as a collection of precise indicators with the ability to make distinctions between regimes with similar human rights records. Numeric values were almost totally absent is most of the reports. Instead, I relied on the fallible interpretation of adjectives and nouns within the reports, and carefully assigned numeric value to them. For example, phrases such as applications for rallies were “routinely denied”, “there were reports of summary police executions”, and “some opposition leaders” reported wiretaps, are highly ambiguous about the number of instances. However, it can be

441 argued that adverbs, adjectives, and nouns reports different degrees of severity and intensity.

Therefore, words like “some,” “occasionally”, “reports,” and “cases” were assigned values of 2,

“several” was interpreted as 3, “many” as 4, while “most,” “routinely,” “rarely permitted,” and

“widespread” given value of 5.

Finally, since I am attempting to place the level of repression in Tanzania into a comparative perspective on the correlation between regime survival and the level of repression, all of the cases (save for Kenya) reflect prolonged party dominance. In that sense, even the inclusion of Kenya is of interest, as KANU did survive the 1992 and 1997 election.

442

Tanzania (All) 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically related Arrests, Detentions, or 1 4 7 5 6 8 12 11 2 1 9 4 1 71 Exiles (Cases) Extended Politically Related Detainees 0 0 0 0 17 18 0 0 0 0 0 0 0 35 (individuals at year end) Restrictions on announcing public 0 0 0 0 0 0 0 0 0 0 0 0 0 0 meetings or purchasing advertizing time on radio Interference with Media 3 4 5 1 3 4 10 2 0 9 3 11 5 60 Disbandment or Denial of Registration 1 1 1 1 1 1 1 0 2 0 0 0 0 9 Action against Political Assembly 0 2 2 1 1 5 7 3 1 0 0 4 0 26 Politically Related Extrajudicial Killings 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Deaths in Political Conflicts (number of 0 2 0 0 0 0 0 3 0 0 2 1 0 8 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 1 4 3 4 5 6 16 7 0 0 7 10 0 63 State Officials Surveillance 1 0 1 1 1 1 1 2 1 0 1 1 1 12 Destruction of Property 0 0 0 0 0 0 0 2 0 0 4 0 0 6 Threats of dismissal for party affiliation 0 0 0 2 0 0 0 0 0 0 0 0 0 2 or preference Actual Denial of Public Rallies 2 5 5 3 3 0 4 0 3 1 2 3 1 32 Restrictions on Speech Freedom (Non- 0 0 0 1 2 2 2 1 2 1 0 0 0 11 media) Restricting Academic Freedom 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total 9 22 24 19 39 45 53 31 11 12 28 34 8 335 Proportion of Violations to Total for all 8.11 23.40 22.86 14.73 21.67 30.61 41.41 24.41 7.48 9.84 35.90 15.32 8.25 16.57 Cases Yearly contribution to total number of 2.69 6.57 7.16 5.67 11.64 13.43 15.82 9.25 3.82 3.57 8.36 10.15 2.39 100 country violations (%) Over/under average yearly contribution -16.8 -3.8 -1.8 -6.8 13.2 19.2 27.2 5.2 -14.8 -13.8 2.2 8.2 -17.8 @ 25.77 cases

443

Tanzania (Excluding Zanzibar) 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically related Arrests, Detentions, or 1 2 2 1 1 3 1 6 2 0 7 2 0 28 Exiles (Cases) Extended Politically Related Detainees 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (individuals at year end) Restrictions on announcing public 0 0 0 0 0 0 0 0 0 0 0 0 0 0 meetings or purchasing advertizing time on radio Interference with Media 3 3 3 0 3 3 1 2 0 4 2 8 1 33 Disbandment or Denial of Registration 1 1 1 1 1 1 1 0 2 0 0 0 0 9 Action against Political Assembly 0 2 0 1 1 3 3 2 1 0 0 3 0 16 Politically Related Extrajudicial Killings 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Deaths in Political Conflicts (number of 0 2 0 0 0 0 0 1 0 0 1 0 0 4 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 1 0 1 0 0 1 5 2 0 0 3 4 0 17 State Officials Surveillance 1 0 1 1 1 1 1 1 1 0 1 1 1 11 Destruction of Property 0 0 0 0 0 0 0 0 0 0 1 0 0 1 Threats of dismissal for party affiliation 0 0 0 0 0 0 0 0 0 0 0 0 0 0 or preference Actual Denial of Public Rallies 1 0 0 0 0 0 1 0 0 1 0 0 0 3 Restrictions on Speech Freedom (Non- 0 0 0 0 1 2 2 1 2 0 0 0 0 8 media) Restricting Academic Freedom 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total 8 10 8 4 8 14 15 15 8 5 15 18 2 130 Proportion of Violations to Total for all 6.72 9.62 7.08 3.01 4.26 8.70 10.49 10.56 5.16 3.94 16.13 7.50 2.02 7.15 Cases Yearly contribution to total number of 6.15 7.69 6.15 3.08 6.15 10.77 11.54 11.54 6.15 3.85 11.54 13.85 1.54 100 country violations (%) Over/under average yearly contribution -2 0 -2 -6 -2 4 5 5 -2 -5 5 8 -8 @ 10 cases

444

Kenya 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 1 0 0 0 0 1 0 0 0 0 0 0 2 4 Politically related Arrests, Detentions, or 6 10 5 2 2 12 10 9 8 3 0 2 2 71 Exiles (Cases) Extended Politically Related Detainees 0 0 0 0 0 1 0 0 0 0 0 0 0 1 (individuals at year end) Restrictions on announcing public 0 0 0 0 0 0 0 0 0 0 0 0 0 0 meetings or purchasing advertizing time on radio Interference with Media 3 7 11 7 18 12 13 3 5 4 2 7 4 96 Disbandment or Denial of Registration 0 0 0 1 0 0 0 0 0 0 0 0 0 1 Action against Political Assembly 4 3 21 12 4 9 7 7 7 0 2 4 0 80 Politically Related Extrajudicial Killings 1 1 1 0 1 0 1 1 0 1 0 0 0 7 Deaths in Political Conflicts (number of 0 0 0 4 1 0 0 0 0 0 1 2 0 8 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 0 2 5 12 3 6 1 1 2 3 2 0 0 37 State Officials Surveillance 2 2 1 2 2 2 2 2 2 1 0 0 0 18 Destruction of Property 0 3 1 1 1 0 0 0 0 0 0 1 1 8 Threats of dismissal for party affiliation 4 2 0 0 0 0 0 0 1 0 0 0 0 7 or preference Actual Denial of Public Rallies 7 19 3 5 0 0 1 1 4 1 0 1 0 42 Restrictions on Speech Freedom (Non- 1 1 1 1 1 1 1 1 1 1 1 1 1 13 media) Restricting Academic Freedom 2 1 0 1 1 1 1 1 1 1 0 0 0 10 Total 31 51 49 48 34 45 37 26 31 15 8 18 10 403 Proportion of Violations to Total for all 34.83 78.46 61.25 48.00 18.38 30.61 25.69 19.70 24.41 12.61 8.16 7.56 10.53 19.93 Cases Yearly contribution to total number of 7.69 12.66 12.16 11.91 8.44 11.17 9.18 6.45 7.69 3.72 1.99 4.47 2.48 100 country violations (%) Over/under average yearly contribution 0.0 20.0 18.0 17.0 3.0 14.0 6.0 -5.0 0.0 -16.0 -23.0 -13.0 -21.0 @ 31 cases

445

Botswana 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically related Arrests, Detentions, or 0 0 0 0 0 0 0 0 0 0 0 0 0 Exiles (Cases) Extended Politically Related Detainees 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (individuals at year end) Restrictions on announcing public 0 0 0 0 0 0 0 0 0 0 0 0 0 0 meetings or purchasing advertizing time on radio Interference with Media 0 0 0 0 0 0 0 2 0 0 1 4 0 7 Disbandment or Denial of Registration 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Action against Political Assembly 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically Related Extrajudicial Killings 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Deaths in Political Conflicts (number of 0 0 0 0 0 0 0 0 0 0 0 0 0 0 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 0 0 0 0 0 0 0 0 0 0 0 0 0 0 State Officials Surveillance 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Destruction of Property 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Threats of dismissal for party affiliation 0 0 0 0 0 0 0 0 0 0 0 0 0 0 or preference Actual Denial of Public Rallies 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Restrictions on Speech Freedom (Non- 0 0 0 0 0 0 0 0 0 0 0 0 0 0 media) Restricting Academic Freedom 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total 0 0 0 0 0 0 0 2 0 0 1 4 0 7 Proportion of Violations to Total for all 0 0 0 0 0 0 0 1.28 0 0 0.95 1.59 0 0.35 Cases Yearly contribution to total number of 0 0 0 0 0 0 0 28.57 0 0 14.29 57.14 100 0 country violations (%) Over/under average yearly contribution -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 1.5 -0.5 -0.5 0.5 3.5 -0.5 @ 0.54 cases

446

Gabon 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically related Arrests, Detentions, or 5 0 0 0 3 0 0 0 0 3 3 1 2 17 Exiles (Cases) Extended Politically Related Detainees 0 0 0 0 0 0 0 0 0 0 1 15 0 16 (individuals at year end) Restrictions on announcing public 0 0 0 0 0 0 0 0 0 0 0 0 0 0 meetings or purchasing advertizing time on radio Interference with Media 3 4 0 3 5 6 1 4 5 8 1 3 1 44 Disbandment or Denial of Registration 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Action against Political Assembly 2 0 0 0 0 0 0 1 0 0 1 2 0 6 Politically Related Extrajudicial Killings 1 0 0 0 0 0 0 0 0 0 0 1 1 3 Deaths in Political Conflicts (number of 5 0 0 1 0 0 0 1 0 0 0 1 0 8 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 1 0 0 0 0 0 0 0 0 0 1 2 0 4 State Officials Surveillance 1 1 1 1 1 1 1 1 1 1 1 1 1 13 Destruction of Property 1 0 0 0 0 0 0 0 0 0 0 0 0 1 Threats of dismissal for party affiliation 0 0 0 0 0 0 0 1 0 0 0 2 0 3 or preference Actual Denial of Public Rallies 1 0 0 0 0 0 0 0 0 0 0 5 0 6 Restrictions on Speech Freedom (Non- 0 0 0 0 0 0 0 0 0 0 0 0 0 0 media) Restricting Academic Freedom 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total 20 5 1 5 9 7 2 8 6 12 8 33 5 121 Proportion of Violations to Total for all 20.00 4.50 0.78 3.50 4.29 3.78 1.12 5.33 3.95 9.84 8.16 14.80 5.00 5.98 Cases Yearly contribution to total number of 16.53 4.13 0.83 4.13 7.44 5.79 1.65 6.61 4.96 9.92 6.61 27.27 4.13 100 country violations (%) Over/under average yearly contribution 10.7 -4.3 -8.3 -4.3 -0.3 -2.3 -7.3 -1.3 -3.3 2.7 -1.3 23.7 -4.3 @ 9.31 cases

447

Malaysia 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically related Arrests, Detentions, or 2 3 1 1 8 7 7 4 1 1 2 3 1 41 Exiles (Cases) Extended Politically Related Detainees 0 0 0 0 1 2 1 6 15 7 1 2 2 37 (individuals at year end) Restrictions on announcing public meetings 0 0 0 0 0 1 0 0 0 0 1 0 0 2 or purchasing advertizing time on radio Interference with Media 5 4 0 4 3 6 7 5 8 5 2 7 4 60 Disbandment or Denial of Registration 0 0 0 0 0 1 1 1 1 1 1 1 2 9 Action against Political Assembly 0 0 0 0 4 6 6 2 2 2 2 3 1 28 Politically Related Extrajudicial Killings 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Deaths in Political Conflicts (number of 0 0 0 0 0 0 0 0 0 0 0 0 0 0 cases) Use of Internal Security Laws 1 1 1 6 3 1 6 4 2 0 0 2 2 29 Harassment, beatings by incumbent or 0 1 1 2 6 2 2 1 2 0 1 1 1 20 State Officials Surveillance 0 0 0 0 0 0 0 0 0 0 0 0 1 1 Destruction of Property 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Threats of dismissal for party affiliation or 0 0 0 0 2 2 2 1 0 0 0 1 0 8 preference Actual Denial of Public Rallies 0 0 1 0 2 4 4 2 1 1 0 0 0 15 Restrictions on Speech Freedom (Non- 1 2 0 2 0 5 5 1 3 2 2 2 6 31 media) Restricting Academic Freedom 0 0 2 0 5 3 2 3 2 2 1 2 1 23 Total 9 11 6 15 34 40 43 30 37 21 13 24 21 304 Proportion of Violations to Total for all 8.11 10.48 4.88 11.28 18.38 26.32 31.16 23.44 30.58 18.58 13.98 10.34 25.00 15.03 Cases Yearly contribution to total number of 2.96 3.62 1.97 4.93 11.18 13.16 14.14 9.87 12.17 6.91 4.28 7.89 6.91 100 country violations (%) Over/under average yearly contribution -14.4 -12.4 -17.4 -8.4 10.6 16.6 19.6 6.6 13.6 -2.4 -10.4 0.6 -2.4 @ 23.38 cases

448

Kazakhstan 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Politically related Arrests, Detentions, or 2 2 6 8 6 5 4 1 5 8 9 15 10 81 Exiles (Cases) Extended Politically Related Detainees 0 1 0 3 1 0 2 1 2 2 1 2 1 16 (individuals at year end) Restrictions on announcing public meetings 0 0 0 0 0 0 0 0 0 0 0 0 0 0 or purchasing advertizing time on radio Interference with Media 6 4 8 2 12 7 6 7 14 15 13 29 18 141 Disbandment or Denial of Registration 3 0 1 0 0 0 0 0 0 4 0 3 1 12 Action against Political Assembly 1 1 2 8 4 3 6 0 1 1 3 8 6 44 Politically Related Extrajudicial Killings 0 0 0 0 0 0 0 1 0 0 0 0 3 4 Deaths in Political Conflicts (number of 0 0 0 0 0 0 0 0 0 0 0 0 0 0 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 2 2 4 4 8 7 2 4 5 1 3 5 6 53 State Officials Surveillance 0 0 1 1 3 2 0 3 1 1 0 1 0 13 Destruction of Property 0 1 0 0 1 0 1 0 2 0 0 1 2 8 Threats of dismissal for party affiliation or 0 0 0 0 1 0 2 0 0 0 0 0 0 3 preference Actual Denial of Public Rallies 0 1 2 3 4 7 0 2 2 2 1 0 1 25 Restrictions on Speech Freedom (Non- 0 2 0 1 0 0 1 1 2 0 3 8 0 18 media) Restricting Academic Freedom 0 0 0 0 3 5 3 1 1 0 0 0 0 13 Total 14 14 24 30 43 36 27 21 35 34 33 72 48 431 Proportion of Violations to Total for all 13.21 13.73 22.86 25.42 24.43 23.08 17.53 15.33 28.46 34.00 45.21 39.13 84.21 21.32 Cases Yearly contribution to total number of 3.25 3.25 5.57 6.96 9.98 8.35 6.26 4.87 8.12 7.89 7.66 16.71 11.14 100 country violations (%) Over/under average yearly contribution -19.2 -19.2 -9.2 -3.2 9.8 2.8 .2 -6 -12.2 1.8 0.8 -0.2 38.8 14.8 @ 33.15 cases

449

Togo 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 7 0 0 0 0 0 0 0 6 0 0 5 0 18 Politically related Arrests, Detentions, or 2 4 8 14 9 5 3 15 10 13 0 2 4 89 Exiles (Cases) Extended Politically Related Detainees 3 0 5 0 1 1 1 2 2 8 2 9 6 40 (individuals at year end) Restrictions on announcing public meetings 0 0 0 0 0 0 0 0 0 0 0 0 0 0 or purchasing advertizing time on radio Interference with Media 3 5 4 7 6 3 5 7 8 7 6 16 1 78 Disbandment or Denial of Registration 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Action against Political Assembly 4 0 2 2 5 2 4 8 3 1 2 11 1 45 Politically Related Extrajudicial Killings 7 1 2 2 9 0 0 0 0 0 0 0 0 21 Deaths in Political Conflicts (number of 1 2 1 0 2 1 1 0 1 4 0 10 0 23 cases) Use of Internal Security Laws 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Harassment, beatings by incumbent or 4 0 2 1 21 2 2 3 4 2 5 15 1 62 State Officials Surveillance 1 1 1 1 1 0 1 0 0 0 0 0 0 6 Destruction of Property 3 0 0 0 3 2 0 1 0 0 0 2 0 11 Threats of dismissal for party affiliation or 0 0 0 0 0 0 0 0 0 0 0 0 0 0 preference Actual Denial of Public Rallies 2 0 0 4 2 1 1 3 3 5 0 1 0 22 Restrictions on Speech Freedom (Non- 0 0 0 0 0 1 0 0 0 0 0 0 0 1 media) Restricting Academic Freedom 0 0 0 0 1 1 1 1 1 0 0 0 0 5 Total 37 13 25 31 60 19 19 40 38 40 15 71 13 421 Proportion of Violations to Total for all 45.50 12.72 24.57 28.52 48.07 12.98 13.50 37.82 35.50 43.68 14.40 51.27 9.83 20.82 Cases Yearly contribution to total number of 8.79 3.09 5.94 7.36 14.25 4.51 4.51 9.50 9.03 9.50 3.56 16.86 3.09 100 country violations (%) Over/under average yearly contribution 4.6 -19.4 -7.4 -1.4 27.6 -13.4 -13.4 7.6 5.6 7.6 -17.4 38.6 -19.4 @ 32.38 cases

450

Sums (Excluding Zanzibar) 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 8 0 0 0 0 1 0 0 6 0 0 5 2 22 Politically related Arrests, Detentions, or 18 21 22 26 29 32 25 35 26 28 21 25 19 327 Exiles (Cases) Extended Politically Related Detainees 3 1 5 3 3 4 4 9 19 17 5 28 9 110 (individuals at year end) Restrictions on announcing public meetings 0 0 0 0 0 1 0 0 0 0 1 0 0 2 or purchasing advertizing time on radio Interference with Media 23 27 26 23 47 37 33 30 40 43 27 74 29 459 Disbandment or Denial of Registration 4 1 2 2 1 2 2 1 3 5 1 4 3 31 Action against Political Assembly 11 6 25 23 18 23 26 20 14 4 10 31 8 219 Politically Related Extrajudicial Killings 9 2 3 2 10 0 1 2 0 1 0 1 4 35 Deaths in Political Conflicts (number of 6 4 1 5 3 1 1 2 1 4 2 13 0 43 cases) Use of Internal Security Laws 1 1 1 6 3 1 6 4 2 0 0 2 2 29 Harassment, beatings by incumbent or 8 5 13 19 38 18 12 11 13 6 15 27 8 193 State Officials Surveillance 5 4 5 6 8 6 5 7 5 3 2 3 3 62 Destruction of Property 4 4 1 1 5 2 1 1 2 0 1 4 3 29 Threats of dismissal for party affiliation or 4 2 0 0 3 2 4 2 1 0 0 3 0 21 preference Actual Denial of Public Rallies 11 20 6 12 8 12 7 8 10 10 1 7 1 113 Restrictions on Speech Freedom (Non- 2 5 1 4 2 9 9 4 8 3 6 11 7 71 media) Restricting Academic Freedom 2 1 2 1 10 10 7 6 5 3 1 2 1 51 Total 119 104 113 133 188 161 143 142 155 127 93 240 99 1817

451

Sums (Including Zanzibar) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Sum Disappearances 8 0 0 0 0 1 0 0 6 0 0 5 2 22 Politically related Arrests, Detentions, or 18 23 27 30 34 37 36 40 26 29 23 27 20 370 Exiles (Cases) Extended Politically Related Detainees 3 1 5 3 20 22 4 9 19 17 5 28 9 145 (individuals at year end) Restrictions on announcing public meetings 0 0 0 0 0 1 0 0 0 0 1 0 0 2 or purchasing advertizing time on radio Interference with Media 23 28 28 24 47 38 42 30 40 48 28 77 33 486 Disbandment or Denial of Registration 4 1 2 2 1 2 2 1 3 5 1 4 3 31 Action against Political Assembly 11 6 27 23 18 25 30 21 14 4 10 32 8 229 Politically Related Extrajudicial Killings 9 2 3 2 10 0 1 2 0 1 0 1 4 35 Deaths in Political Conflicts (number of 6 4 1 5 3 1 1 4 1 4 3 14 0 47 cases) Use of Internal Security Laws 1 1 1 6 3 1 6 4 2 0 0 2 2 29 Harassment, beatings by incumbent or 8 9 15 23 43 23 23 16 13 6 19 33 8 239 State Officials Surveillance 5 4 5 6 8 6 5 8 5 3 2 3 3 63 Destruction of Property 4 4 1 1 5 2 1 3 2 0 4 4 3 34 Threats of dismissal for party affiliation or 4 2 0 2 3 2 4 2 1 0 0 3 0 23 preference Actual Denial of Public Rallies 12 25 11 15 11 12 10 8 13 10 3 10 2 142 Restrictions on Speech Freedom (Non- 2 5 1 5 3 9 9 4 8 4 6 11 7 74 media) Restricting Academic Freedom 2 1 2 1 10 10 7 6 5 3 1 2 1 51 Total 120 116 129 148 219 192 181 158 158 134 106 256 105 2022

452

APPENDIX B ELECTION TURNOUT RATES FOR SELECTED COUNTRIES

Legend: Column A: Year. Column B: Voter turnout as a percentage of eligible voting population. Column C: Voter turnout as a percentage of registered population. Single -party election Regime turnover Non-African case Notes: IDEA as source unless otherwise noted Presidential elections unless otherwise noted with **

Kenya Tanzania Zambia Malawi A B C A B C A B C A B C **1963 72 % 1962 **1969 46 % 1965 52 % 83 % **1968 50 % 82 % 1974 1970 52 % 70 % 1973 39 % 1964 1979 1975 65 % 82 % 1978 67 % 1978 1983 1980 73 % 86 % 1983 64 % 1983 1988 1985 54 % 75 % 1988 1987 1992 46 % 68 % 1990 53 % 73 % 1991 34 % 46 % 1992 1997 48 % 65 % 1995 59 % 73 % 1996 28 % 58 % 1994 68 % 81 % 2002 56 % 2000 68 % 84 % 2001 68 % 1999 94 % 2007 2005 68 % 71 % 2006 71 % 2004 54 % Mean 47 % 61 % 60 % 77 % 37 % 62 % 68 % 76 %

Botswana Benin Cameroon Cape Verde A B C A B C A B C A B C **1965 69 % 75 % **1969 37 % 55 % 1960 **1974 26 % 31 % 1964 1973 1975 **1979 46 % 58 % ***1968 73 % 1978 1980 **1984 54 % 78 % 1970 1983 1985 **1989 48 % 68 % 1991 57 % 64 % **1988 67 % 90 % 1991 61 % **1994 45 % 77 % 1996 73 % 78 % 1992 55 % 72 % 1996 55 % **1999 42 % 77 % 2001 88 % 1997 81 % 2001 63 % 52 % **2004 76 % 2006 82 % 2004 80 % 2006 53 % Mean 46 % 66 % 65 % 77 % 61 % 81 % 63 % 55 %

Côte d'Ivoire Djibouti The Gambia Ghana A B C A B C A B C A B C 1960 99 % 1956 50 % 1965 **1972 44 % 76 % 1960 1970 **1977 68 % 82 % 1965 1975 **1982 57 % **1969 42 % 63 % 1980 1981 85 % **1987 54 % 80 % **1979 37 % 35 % 1985 **1987 87 % **1992 55 % 56 % 1992 60 % 50 % 1990 66 % 70 % 1993 26 % 50 % 1996 73 % 88 % 1996 82 % 78 % 1995 31 % 45 % 1997 29 % 58 % 2001 90 % 2000 62 % 2000 28 % 37 % 2005 72 % 2006 59 % 2004 85 % Mean 42 % 63 % 28 % 70 % 58 % 76 % 55 % 60 %

Nigeria Senegal Gabon Togo A B C A B C A B C A B C 1963 79 % 86 % 1968 94 % 1963 1978 42 % 62 % ***1972 99 % 1983 40 % 58 % 1979 1959 43 % 80 % 1988 38 % 59 % 1985 70 % 79 % 1979 41 % 32 % 1993 33 % 52 % 1990 75 % **1983 58 % 39 % 1998 28 % 39 % 1994 65 % 1993 28 % 2000 37 % 62 % 1993 67 % 88 % 1998 70 % 1999 57 % 52 % 2001 67 % 1998 55 % 54 % 2002 67 % 2003 69 % 2007 71 % 2005 63 % 2005 64 % Mean 46 % 54 % 42 % 65 % 61 % 68 % 73 % 74 %

Zimbabwe Kazakhstan Malaysia Mexico A B C A B C A B C A B C 1946 22 % 90 % 1952 29 % 74 % *1979 59 % 64 % 1958 52 % 71 % *1980 84 % **1974 40 % 1964 57 % 69 % *1985 75 % **1978 57 % 1970 65 % 64 % *1990 49 % **1982 61 % 72 % 1976 58 % 65 % *1995 26 % 31 % **1986 64 % 74 % 1982 66 % 75 % 1996 27 % 32 % **1995 23 % 76 % **1990 63 % 70 % 1988 42 % 50 % *2000 48 % 1999 71 % 86 % **1995 64 % 72 % 1994 66 % 79 % 2002 54 % 2005 77 % **1999 50 % 69 % 2000 60 % 64 % *2005 48 % **2007 68 % **2004 74 % 2006 59 % Mean 53 % 46 % 47 % 77 % 57 % 72 % 58 % 66 %

454

APPENDIX C REGIONAL RESULTS FOR THE 1995 PRESIDENTIAL ELECTION

Legend: Column A: Candidate (Party). Column B: Total votes cast for each party. Column C: Percentage of B within each party. Tells how well the party did relative to other parties. Column D: Percentage of B in proportion to the total votes received by each party. This column tells which regions contributed the most votes for each party. Column E: Comparison between regional percentage won by each party, versus national percentage points won. Tells the degree to which a region can be considered a party stronghold percentage won by each party.

Data A B C D E

Arusha Benjimin Mkapa (CCM) 261,342 60.46% 6.49% -1.36% A.L. Mrema (NCCR-M) 159,770 36.96% 8.83% 9.19% I.H. Lipumba (CUF) 7,518 1.74% 1.79% -4.69% John Cheyo (UDP) 3,610 0.84% 1.40% -3.14%

Total 432,240 100.00% 6.64% 0.00% Coast Benjimin Mkapa (CCM) 132,717 71.67% 3.30% 9.85% A.L. Mrema (NCCR-M) 24,652 13.31% 1.36% -14.46% I.H. Lipumba (CUF) 23,388 12.63% 5.58% 6.20% John Cheyo (UDP) 4,411 2.38% 1.70% -1.59%

Total 185,168 100.00% 2.84% 0.00% Dar es Salaam Benjimin Mkapa (CCM) 223,922 71.97% 5.56% 10.15% A.L. Mrema (NCCR-M) 66,068 21.23% 3.65% -6.54% I.H. Lipumba (CUF) 18,988 6.10% 4.53% -0.33% John Cheyo (UDP) 2,151 0.69% 0.83% -3.28%

Total 311,129 100.00% 4.78% 0.00%

Dodoma* Benjimin Mkapa (CCM) 222,408 71.89% 5.52% 10.07% A.L. Mrema (NCCR-M) 53,150 17.18% 2.94% -10.59% I.H. Lipumba (CUF) 18,253 5.90% 4.36% -0.53% John Cheyo (UDP) 15,562 5.03% 6.01% 1.06%

Total 309,373 100.00% 4.75% 0.00% Iringa* Benjimin Mkapa (CCM) 254,544 67.35% 6.32% 5.52% A.L. Mrema (NCCR-M) 96,741 25.59% 5.35% -2.18% I.H. Lipumba (CUF) 11,395 3.01% 2.72% -3.42% John Cheyo (UDP) 15,289 4.05% 5.91% 0.07%

Total 377,969 100.00% 5.80% 0.00% Kagera Benjimin Mkapa (CCM) 206,307 56.06% 5.12% -5.76% A.L. Mrema (NCCR-M) 145,953 39.66% 8.07% 11.89% I.H. Lipumba (CUF) 10,832 2.94% 2.59% -3.49% John Cheyo (UDP) 4,894 1.33% 1.89% -2.64%

Total 367,986 100.00% 5.65% 0.00% Kigoma Benjimin Mkapa (CCM) 128,845 56.41% 3.20% -5.42% A.L. Mrema (NCCR-M) 68,651 30.06% 3.80% 2.28% I.H. Lipumba (CUF) 27,301 11.95% 6.52% 5.52% John Cheyo (UDP) 3,620 1.58% 1.40% -2.39%

Total 228,417 100.00% 3.51% 0.00% Kilimanjaro Benjimin Mkapa (CCM) 82,914 20.39% 2.06% -41.43% A.L. Mrema (NCCR-M) 315,291 77.54% 17.43% 49.77% I.H. Lipumba (CUF) 6,641 1.63% 1.59% -4.80% John Cheyo (UDP) 1,756 0.43% 0.68% -3.54%

Total 406,602 100.00% 6.24% 0.00%

456

Lindi Benjimin Mkapa (CCM) 139,191 86.14% 3.46% 24.32% A.L. Mrema (NCCR-M) 12,701 7.86% 0.70% -19.91% I.H. Lipumba (CUF) 7,043 4.36% 1.68% -2.07% John Cheyo (UDP) 2,650 1.64% 1.02% -2.33%

Total 161,585 100.00% 2.48% 0.00% Mara Benjimin Mkapa (CCM) 155,270 56.07% 3.86% -5.76% A.L. Mrema (NCCR-M) 109,719 39.62% 6.07% 11.85% I.H. Lipumba (CUF) 5,579 2.01% 1.33% -4.42% John Cheyo (UDP) 6,376 2.30% 2.46% -1.67%

Total 276,944 100.00% 4.25% 0.00% Mbeya Benjimin Mkapa (CCM) 220,178 57.79% 5.47% -4.03% A.L. Mrema (NCCR-M) 148,602 39.01% 8.22% 11.24% I.H. Lipumba (CUF) 6,434 1.69% 1.54% -4.74% John Cheyo (UDP) 5,753 1.51% 2.22% -2.46%

Total 380,967 100.00% 5.85% 0.00% Morogoro Benjimin Mkapa (CCM) 217,347 64.72% 5.40% 2.89% A.L. Mrema (NCCR-M) 98,003 29.18% 5.42% 1.41% I.H. Lipumba (CUF) 12,149 3.62% 2.90% -2.82% John Cheyo (UDP) 8,351 2.49% 3.23% -1.49%

Total 335,850 100.00% 5.16% 0.00% Mtwara Benjimin Mkapa (CCM) 274,582 89.51% 6.82% 27.68% A.L. Mrema (NCCR-M) 14,650 4.78% 0.81% -22.99% I.H. Lipumba (CUF) 12,150 3.96% 2.90% -2.47% John Cheyo (UDP) 5,386 1.76% 2.08% -2.22%

Total 306,768 100.00% 4.71% 0.00%

457

Mwanza Benjimin Mkapa (CCM) 262,279 56.51% 6.51% -5.32% A.L. Mrema (NCCR-M) 109,327 23.55% 6.04% -4.22% I.H. Lipumba (CUF) 25,818 5.56% 6.16% -0.87% John Cheyo (UDP) 66,716 14.37% 25.79% 10.40%

Total 464,140 100.00% 7.13% 0.00% Rukwa Benjimin Mkapa (CCM) 97,924 61.13% 2.43% -0.69% A.L. Mrema (NCCR-M) 57,346 35.80% 3.17% 8.03% I.H. Lipumba (CUF) 2,983 1.86% 0.71% -4.57% John Cheyo (UDP) 1,927 1.20% 0.74% -2.77%

Total 160,180 100.00% 2.46% 0.00% Ruvuma Benjimin Mkapa (CCM) 150,544 79.99% 3.74% 18.17% A.L. Mrema (NCCR-M) 31,180 16.57% 1.72% -11.20% I.H. Lipumba (CUF) 4,232 2.25% 1.01% -4.18% John Cheyo (UDP) 2,247 1.19% 0.87% -2.78%

Total 188,203 100.00% 2.89% 0.00% Shinyanga Benjimin Mkapa (CCM) 224,646 52.23% 5.58% -9.59% A.L. Mrema (NCCR-M) 84,949 19.75% 4.70% -8.02% I.H. Lipumba (CUF) 19,277 4.48% 4.60% -1.95% John Cheyo (UDP) 101,226 23.54% 39.12% 19.56%

Total 430,098 100.00% 6.60% 0.00% Singida Benjimin Mkapa (CCM) 162,733 75.21% 4.04% 13.38% A.L. Mrema (NCCR-M) 37,639 17.40% 2.08% -10.38% I.H. Lipumba (CUF) 10,584 4.89% 2.53% -1.54% John Cheyo (UDP) 5,422 2.51% 2.10% -1.47%

Total 216,378 100.00% 3.32% 0.00%

458

Tabora Benjimin Mkapa (CCM) 144,083 55.40% 3.58% -6.42% A.L. Mrema (NCCR-M) 87,680 33.72% 4.85% 5.94% I.H. Lipumba (CUF) 18,706 7.19% 4.46% 0.76% John Cheyo (UDP) 9,592 3.69% 3.71% -0.28%

Total 260,061 100.00% 3.99% 0.00% Tanga Benjimin Mkapa (CCM) 258,418 75.53% 6.42% 13.71% A.L. Mrema (NCCR-M) 52,807 15.43% 2.92% -12.34% I.H. Lipumba (CUF) 25,981 7.59% 6.20% 1.16% John Cheyo (UDP) 4,920 1.44% 1.90% -2.53%

Total 342,126 100.00% 5.25% 0.00% Pemba North Benjimin Mkapa (CCM) 6,920 11.22% 0.17% -50.60% A.L. Mrema (NCCR-M) 349 0.57% 0.02% -27.20% I.H. Lipumba (CUF) 54,133 87.76% 12.92% 81.33% John Cheyo (UDP) 278 0.45% 0.11% -3.52%

Total 61,680 100.00% 0.95% 0.00% Pemba South Benjimin Mkapa (CCM) 13,473 25.06% 0.33% -36.76% A.L. Mrema (NCCR-M) 394 0.73% 0.02% -27.04% I.H. Lipumba (CUF) 39,449 73.39% 9.42% 66.95% John Cheyo (UDP) 440 0.82% 0.17% -3.15%

Total 53,756 100.00% 0.83% 0.00% Unguja North Benjimin Mkapa (CCM) 66,321 73.80% 1.65% 11.97% A.L. Mrema (NCCR-M) 542 0.60% 0.03% -27.17% I.H. Lipumba (CUF) 22,215 24.72% 5.30% 18.29% John Cheyo (UDP) 792 0.88% 0.31% -3.09%

Total 89,870 100.00% 1.38% 0.00%

459

Unguja South Benjimin Mkapa (CCM) 27,441 80.71% 0.68% 18.89% A.L. Mrema (NCCR-M) 453 1.33% 0.03% -26.44% I.H. Lipumba (CUF) 5,728 16.85% 1.37% 10.41% John Cheyo (UDP) 376 1.11% 0.15% -2.87%

Total 33,998 100.00% 0.52% 0.00% Urban West Benjimin Mkapa (CCM) 79,544 60.60% 1.98% -1.22% A.L. Mrema (NCCR-M) 1,045 0.80% 0.06% -26.97% I.H. Lipumba (CUF) 49,727 37.89% 11.87% 31.45% John Cheyo (UDP) 941 0.72% 0.36% -3.26%

Total 131,257 100.00% 2.02% 0.00%

National Total Benjimin Mkapa (CCM) 4,026,422 61.82% A.L. Mrema (NCCR-M) 1,808,616 27.77% I.H. Lipumba (CUF) 418,973 6.43% John Cheyo (UDP) 258,734 3.97%

Total 6,512,745 100.00%

Aggregate Statistics for C: Percentage of votes for each party

Range of total votes cast Benjimin Mkapa (CCM) High=89.5% (Mtwara), Low=11.2% (Pemba North) A.L. Mrema (NCCR-M) High=77.5% (Kilimanjaro), Low=0.6% (Pemba North) I.H. Lipumba (CUF) High= 87.8% (Pemba North), Low=1.6% (Kilimanjaro) John Cheyo (UDP) High=23.5% (Shinyanga), Low=0.4% (Kilimanjaro)

460

Mean Median Skew Benjimin Mkapa (CCM) 61.51% 61.13% -1.20 A.L. Mrema (NCCR-M) 21.93% 19.75% 1.12 I.H. Lipumba (CUF) 13.44% 4.89% 2.68 John Cheyo (UDP) 3.12% 1.51% 3.39

Aggregate Statistics for D: Regional contribution to party support

Range of total votes cast Benjimin Mkapa (CCM) High=6.8% (Mtwara), Low=0.2% (Pemba North) A.L. Mrema (NCCR-M) High=17.4% (Kilimanjaro), Low=0.0% (Pemba North) I.H. Lipumba (CUF) High= 12.9% (Pemba North), Low=0.7% (Rukwa) John Cheyo (UDP) High=39.1% (Shinyanga), Low=0.1% (Pemba North)

Mean Median Skew Benjimin Mkapa (CCM) 3.91% 3.86% -0.37 A.L. Mrema (NCCR-M) 3.83% 3.17% 1.94 I.H. Lipumba (CUF) 4.31% 2.90% 1.34 John Cheyo (UDP) 4.48% 1.89% 3.23

Notes * estimated based upon percentages from parliamentary election

461

APPENDIX D REGIONAL RESULTS FOR THE 1995 PARLIAMENTARY ELECTIONS

Data A B C D E

Arusha 1) CCM 252,528 59.02% 6.62% -0.31% 2) NCCR-M 123,058 28.76% 8.75% 6.89% 3) CHADEMA 45,161 10.56% 11.38% 4.38% 4) UDP 3,237 0.76% 1.51% -2.57% 5) CUF 2,599 0.61% 0.80% -4.42% 6) Other 1,253 0.29% 0.46% -3.97% Total 427,836 100.00% 6.66% 0.00% Coast 1) CCM 117,062 64.01% 3.07% 4.68% 2) NCCR-M 17,628 9.64% 1.25% -12.24% 3) CHADEMA 16,170 8.84% 4.07% 2.67% 4) UDP 3,112 1.70% 1.46% -1.62% 5) CUF 8,695 4.75% 2.69% -0.28% 6) Other 20,206 11.05% 7.37% 6.78% Total 182,873 100.00% 2.84% 0.00% Dar es Salaam 1) CCM 184,381 60.28% 4.83% 0.95% 2) NCCR-M 76,322 24.95% 5.43% 3.08% 3) CHADEMA 13,095 4.28% 3.30% -1.89% 4) UDP 3,731 1.22% 1.75% -2.10% 5) CUF 9,516 3.11% 2.94% -1.92% 6) Other 18,810 6.15% 6.86% 1.89% Total 305,855 100.00% 4.76% 0.00%

Dodoma 1) CCM 231,377 69.96% 6.07% 10.63% 2) NCCR-M 50,430 15.25% 3.59% -6.63% 3) CHADEMA 10,095 3.05% 2.54% -3.12% 4) UDP 10,262 3.10% 4.80% -0.22% 5) CUF 13,113 3.97% 4.05% -1.07% 6) Other 15,438 4.67% 5.63% 0.40% Total 330,715 100.00% 5.14% 0.00% Iringa 1) CCM 264,914 65.55% 6.95% 6.22% 2) NCCR-M 96,173 23.80% 6.84% 1.92% 3) CHADEMA 24,868 6.15% 6.27% -0.02% 4) UDP 9,075 2.25% 4.25% -1.08% 5) CUF 4,930 1.22% 1.52% -3.81% 6) Other 4,155 1.03% 1.52% -3.24% Total 404,115 100.00% 6.29% 0.00% Kagera 1) CCM 214,821 58.64% 5.63% -0.69% 2) NCCR-M 101,957 27.83% 7.25% 5.96% 3) CHADEMA 21,686 5.92% 5.46% -0.25% 4) UDP 4,401 1.20% 2.06% -2.12% 5) CUF 15,139 4.13% 4.68% -0.90% 6) Other 8,331 2.27% 3.04% -1.99% Total 366,335 100.00% 5.70% 0.00% Kigoma 1) CCM 125,807 55.75% 3.30% -3.58% 2) NCCR-M 30,614 13.57% 2.18% -8.31% 3) CHADEMA 47,189 20.91% 11.89% 14.74% 4) UDP 3,672 1.63% 1.72% -1.70% 5) CUF 5,132 2.27% 1.59% -2.76% 6) Other 13,234 5.86% 4.83% 1.60% Total 225,648 100.00% 3.51% 0.00% Kilimanjaro 1) CCM 99,106 24.63% 2.60% -34.70% 2) NCCR-M 217,689 54.11% 15.48% 32.23% 3) CHADEMA 72,484 18.02% 18.27% 11.84% 4) UDP 840 0.21% 0.39% -3.12% 5) CUF 382 0.09% 0.12% -4.94% 6) Other 11,813 2.94% 4.31% -1.33% Total 402,314 100.00% 6.26% 0.00%

463

Lindi 1) CCM 125,762 79.43% 3.30% 20.10% 2) NCCR-M 11,810 7.46% 0.84% -14.42% 3) CHADEMA 5,749 3.63% 1.45% -2.54% 4) UDP -3.32% 5) CUF 8,718 5.51% 2.69% 0.47% 6) Other 6,298 3.98% 2.30% -0.29% Total 158,337 100.00% 2.46% 0.00% Mara 1) CCM 128,936 47.07% 3.38% -12.26% 2) NCCR-M 122,368 44.67% 8.70% 22.80% 3) CHADEMA 10,073 3.68% 2.54% -2.50% 4) UDP 4,087 1.49% 1.91% -1.83% 5) CUF 5,183 1.89% 1.60% -3.14% 6) Other 3,283 1.20% 1.20% -3.07% Total 273,930 100.00% 4.26% 0.00% Mbeya 1) CCM 216,457 57.60% 5.68% -1.73% 2) NCCR-M 124,047 33.01% 8.82% 11.13% 3) CHADEMA 11,651 3.10% 2.94% -3.07% 4) UDP 5,531 1.47% 2.59% -1.85% 5) CUF 2,975 0.79% 0.92% -4.24% 6) Other 15,133 4.03% 5.52% -0.24% Total 375,794 100.00% 5.85% 0.00% Morogoro 1) CCM 198,727 59.25% 5.21% -0.08% 2) NCCR-M 82,102 24.48% 5.84% 2.60% 3) CHADEMA 17,956 5.35% 4.52% -0.82% 4) UDP 8,397 2.50% 3.93% -0.82% 5) CUF 7,449 2.22% 2.30% -2.81% 6) Other 20,756 6.19% 7.57% 1.92% Total 335,387 100.00% 5.22% 0.00% Mtwara 1) CCM 257,071 82.96% 6.74% 23.62% 2) NCCR-M 21,343 6.89% 1.52% -14.99% 3) CHADEMA 11,576 3.74% 2.92% -2.44% 4) UDP -3.32% 5) CUF 13,002 4.20% 4.02% -0.84% 6) Other 6,900 2.23% 2.52% -2.04% Total 309,892 100.00% 4.82% 0.00%

464

Mwanza 1) CCM 246,266 54.71% 6.46% -4.62% 2) NCCR-M 67,993 15.11% 4.83% -6.77% 3) CHADEMA 29,762 6.61% 7.50% 0.44% 4) UDP 61,725 13.71% 28.88% 10.39% 5) CUF 25,876 5.75% 8.00% 0.72% 6) Other 18,474 4.10% 6.74% -0.16% Total 450,096 100.00% 7.00% 0.00% Rukwa 1) CCM 95,935 59.98% 2.52% 0.65% 2) NCCR-M 47,169 29.49% 3.35% 7.61% 3) CHADEMA 6,038 3.77% 1.52% -2.40% 4) UDP 4,002 2.50% 1.87% -0.82% 5) CUF 3,236 2.02% 1.00% -3.01% 6) Other 3,576 2.24% 1.30% -2.03% Total 159,956 100.00% 2.49% 0.00% Ruvuma 1) CCM 148,138 77.74% 3.88% 18.40% 2) NCCR-M 26,767 14.05% 1.90% -7.83% 3) CHADEMA 3,078 1.62% 0.78% -4.56% 4) UDP -3.32% 5) CUF 4,834 2.54% 1.49% -2.50% 6) Other 7,750 4.07% 2.83% -0.20% Total 190,567 100.00% 2.96% 0.00% Shinyanga 1) CCM 222,864 52.69% 5.84% -6.64% 2) NCCR-M 52,140 12.33% 3.71% -9.55% 3) CHADEMA 28,069 6.64% 7.07% 0.46% 4) UDP 78,991 18.68% 36.96% 15.35% 5) CUF 23,323 5.51% 7.21% 0.48% 6) Other 17,573 4.15% 6.41% -0.11% Total 422,960 100.00% 6.58% 0.00% Singida 1) CCM 151,438 71.87% 3.97% 12.54% 2) NCCR-M 28,129 13.35% 2.00% -8.53% 3) CHADEMA 10,049 4.77% 2.53% -1.40% 4) UDP 2,051 0.97% 0.96% -2.35% 5) CUF 3,499 1.66% 1.08% -3.37% 6) Other 15,550 7.38% 5.67% 3.12% Total 210,716 100.00% 3.28% 0.00%

465

Tabora 1) CCM 132,216 52.17% 3.47% -7.16% 2) NCCR-M 58,053 22.90% 4.13% 1.03% 3) CHADEMA 5,683 2.24% 1.43% -3.93% 4) UDP 8,037 3.17% 3.76% -0.15% 5) CUF 5,089 2.01% 1.57% -3.02% 6) Other 44,376 17.51% 16.19% 13.24% Total 253,454 100.00% 3.94% 0.00% Tanga 1) CCM 249,539 73.60% 6.54% 14.27% 2) NCCR-M 49,511 14.60% 3.52% -7.27% 3) CHADEMA 4,580 1.35% 1.15% -4.82% 4) UDP 2,173 0.64% 1.02% -2.68% 5) CUF 16,588 4.89% 5.13% -0.14% 6) Other 16,671 4.92% 6.08% 0.65% Total 339,062 100.00% 5.27% 0.00% Pemba North 1) CCM 6,654 10.89% 0.17% -48.44% 2) NCCR-M 139 0.23% 0.01% -21.65% 3) CHADEMA 49 0.08% 0.01% -6.09% 4) UDP -3.32% 5) CUF 53,520 87.61% 16.54% 82.58% 6) Other 726 1.19% 0.26% -3.08% Total 61,088 100.00% 0.95% 0.00% Pemba South 1) CCM 12,713 24.06% 0.33% -35.27% 2) NCCR-M 428 0.81% 0.03% -21.07% 3) CHADEMA -6.17% 4) UDP 71 0.13% 0.03% -3.19% 5) CUF 39,062 73.92% 12.07% 68.89% 6) Other 569 1.08% 0.21% -3.19% Total 52,843 100.00% 0.82% 0.00% Unguja North 1) CCM 29,098 72.33% 0.76% 13.00% 2) NCCR-M -21.88% 3) CHADEMA 1,029 2.56% 0.26% -3.62% 4) UDP -3.32% 5) CUF 9,520 23.66% 2.94% 18.63% 6) Other 584 1.45% 0.21% -2.81% Total 40,231 100.00% 0.63% 0.00%

466

Unguja South 1) CCM 27,142 81.48% 0.71% 22.15% 2) NCCR-M 118 0.35% 0.01% -21.52% 3) CHADEMA 213 0.64% 0.05% -5.53% 4) UDP -3.32% 5) CUF 5,150 15.46% 1.59% 10.43% 6) Other 688 2.07% 0.25% -2.20% Total 33,311 100.00% 0.52% 0.00% Urban West 1) CCM 75,216 65.19% 1.97% 5.86% 2) NCCR-M 337 0.29% 0.02% -21.58% 3) CHADEMA 530 0.46% 0.13% -5.71% 4) UDP 304 0.26% 0.14% -3.06% 5) CUF 36,982 32.05% 11.43% 27.02% 6) Other 2,005 1.74% 0.73% -2.53% Total 115,374 100.00% 1.79% 0.00%

National Total 1) CCM 3,814,168 59.33% 2) NCCR-M 1,406,325 21.88% 3) CHADEMA 396,833 6.17% 4) UDP 213,699 3.32% 5) CUF 323,512 5.03% 6) Other 274,152 4.26%

Total 6,428,689 100.00%

Aggregate Statistics for C: Percentage of votes for each party

Range of total votes cast CCM High=81.5% (Unguja South), Low=10.9% (Pemba North) NCCR-M High=54.1% (Kilimanjaro), Low=Uncontested (Unguja North) CHADEMA High= 20.9% (Kigoma), Low=Uncontested (Pemba South) UDP High=18.7% (Shinyanga), Low=0.0% (Uncontested Areas) CUF High=87.6% (Pemba North), Low=0.0% (Uncontested areas) Other

467

Mean Median Skew CCM 59.23% 59.98% -1.22 NCCR-M 18.25% 14.85% 0.82 CHADEMA 5.33% 3.76% 1.99 UDP 3.03% 1.49% 2.77 CUF 11.67% 3.97% 2.80 Other 4.15% 3.98% 2.25

Aggregate Statistics for D: Regional contribution to party support

Range of total votes cast CCM High=7.0% (Iringa), Low=0.2% (Pemba North) NCCR-M High=15.5% (Kilimanjaro), Low=Uncontested (Unguja North) CHADEMA High= 18.3% (Kilimanjaro), Low=Uncontested (Pemba North) UDP High=37.0% (Shinyanga), Low=0.0% (Uncontested areas) CUF High=16.5% (Pemba North), Low=0.0% (Uncontested areas) Other

Mean Median Skew CCM 4.00% 3.88% -0.31 NCCR-M 4.17% 3.55% 1.26 CHADEMA 4.17% 2.73% 1.77 UDP 5.26% 1.87% 2.80 CUF 4.00% 2.69% 1.81 Other 4.00% 3.04% 1.53

468

APPENDIX E REGIONAL RESULTS FROM THE 2000 PRESIDENTIAL ELECTION

Data A B C D E

Arusha Benjimin Mkapa (CCM) 407,371 82.39% 6.95% 10.65% A.L. Mrema (TLP) 33,795 6.84% 5.30% -0.96% I.H. Lipumba (CUF) 44,524 9.01% 3.35% -7.26% John Cheyo (UDP) 8,746 1.77% 2.55% -2.43%

Total 494,436 100.00% 6.05% 0.00% Coast Benjimin Mkapa (CCM) 162,258 67.51% 2.77% -4.24% A.L. Mrema (TLP) 5,630 2.34% 0.88% -5.45% I.H. Lipumba (CUF) 68,369 28.44% 5.14% 12.18% John Cheyo (UDP) 4,101 1.71% 1.20% -2.49%

Total 240,358 100.00% 2.94% 0.00% Dar es Salaam Benjimin Mkapa (CCM) 431,712 55.66% 7.36% -16.08% A.L. Mrema (TLP) 81,187 10.47% 12.74% 2.67% I.H. Lipumba (CUF) 257,227 33.17% 19.35% 16.90% John Cheyo (UDP) 5,440 0.70% 1.59% -3.49%

Total 775,566 100.00% 9.49% 0.00%

Dodoma Benjimin Mkapa (CCM) 344,025 82.03% 5.87% 10.29% A.L. Mrema (TLP) 13,357 3.18% 2.10% -4.61% I.H. Lipumba (CUF) 54,184 12.92% 4.08% -3.34% John Cheyo (UDP) 7,810 1.86% 2.28% -2.33%

Total 419,376 100.00% 5.13% 0.00% Iringa Benjimin Mkapa (CCM) 372,341 90.18% 6.35% 18.44% A.L. Mrema (TLP) 19,668 4.76% 3.09% -3.03% I.H. Lipumba (CUF) 15,779 3.82% 1.19% -12.44% John Cheyo (UDP) 5,088 1.23% 1.48% -2.96%

Total 412,876 100.00% 5.05% 0.00% Kagera Benjimin Mkapa (CCM) 301,515 66.59% 5.14% -5.15% A.L. Mrema (TLP) 49,371 10.90% 7.75% 3.11% I.H. Lipumba (CUF) 90,506 19.99% 6.81% 3.73% John Cheyo (UDP) 11,373 2.51% 3.32% -1.68%

Total 452,765 100.00% 5.54% 0.00% Kigoma Benjimin Mkapa (CCM) 227,345 73.41% 3.88% 1.67% A.L. Mrema (TLP) 6,378 2.06% 1.00% -5.74% I.H. Lipumba (CUF) 69,990 22.60% 5.27% 6.34% John Cheyo (UDP) 5,962 1.93% 1.74% -2.27%

Total 309,675 100.00% 3.79% 0.00% Kilimanjaro Benjimin Mkapa (CCM) 179,001 48.88% 3.05% -22.86% A.L. Mrema (TLP) 151,244 41.30% 23.74% 33.50% I.H. Lipumba (CUF) 32,065 8.76% 2.41% -7.51% John Cheyo (UDP) 3,894 1.06% 1.14% -3.13%

Total 366,204 100.00% 4.48% 0.00%

470

Lindi Benjimin Mkapa (CCM) 173,085 82.23% 2.95% 10.49% A.L. Mrema (TLP) 2,059 0.98% 0.32% -6.82% I.H. Lipumba (CUF) 31,411 14.92% 2.36% -1.34% John Cheyo (UDP) 3,927 1.87% 1.15% -2.33%

Total 210,482 100.00% 2.58% 0.00% Mara Benjimin Mkapa (CCM) 263,116 77.09% 4.49% 5.35% A.L. Mrema (TLP) 38,380 11.25% 6.02% 3.45% I.H. Lipumba (CUF) 26,640 7.81% 2.00% -8.46% John Cheyo (UDP) 13,157 3.86% 3.84% -0.34%

Total 341,293 100.00% 4.18% 0.00% Mbeya Benjimin Mkapa (CCM) 568,458 92.40% 9.70% 20.65% A.L. Mrema (TLP) 24,524 3.99% 3.85% -3.81% I.H. Lipumba (CUF) 18,382 2.99% 1.38% -13.28% John Cheyo (UDP) 3,859 0.63% 1.13% -3.57%

Total 615,223 100.00% 7.53% 0.00% Morogoro Benjimin Mkapa (CCM) 286,422 73.19% 4.89% 1.44% A.L. Mrema (TLP) 32,684 8.35% 5.13% 0.56% I.H. Lipumba (CUF) 64,777 16.55% 4.87% 0.29% John Cheyo (UDP) 7,471 1.91% 2.18% -2.29%

Total 391,354 100.00% 4.79% 0.00% Mtwara Benjimin Mkapa (CCM) 322,549 89.74% 5.50% 17.99% A.L. Mrema (TLP) 3,277 0.91% 0.51% -6.88% I.H. Lipumba (CUF) 27,565 7.67% 2.07% -8.59% John Cheyo (UDP) 6,037 1.68% 1.76% -2.52%

Total 359,428 100.00% 4.40% 0.00%

471

Mwanza Benjimin Mkapa (CCM) 396,682 65.34% 6.77% -6.40% A.L. Mrema (TLP) 41,419 6.82% 6.50% -0.97% I.H. Lipumba (CUF) 85,304 14.05% 6.42% -2.21% John Cheyo (UDP) 83,678 13.78% 24.40% 9.59%

Total 607,083 100.00% 7.43% 0.00% Rukwa Benjimin Mkapa (CCM) 140,422 74.30% 2.39% 2.55% A.L. Mrema (TLP) 32,796 17.35% 5.15% 9.56% I.H. Lipumba (CUF) 12,236 6.47% 0.92% -9.79% John Cheyo (UDP) 3,550 1.88% 1.04% -2.32%

Total 189,004 100.00% 2.31% 0.00% Ruvuma Benjimin Mkapa (CCM) 196,035 82.79% 3.34% 11.05% A.L. Mrema (TLP) 7,505 3.17% 1.18% -4.63% I.H. Lipumba (CUF) 29,360 12.40% 2.21% -3.86% John Cheyo (UDP) 3,881 1.64% 1.13% -2.56%

Total 236,781 100.00% 2.90% 0.00% Shinyanga Benjimin Mkapa (CCM) 331,581 80.92% 5.66% 9.18% A.L. Mrema (TLP) 38,380 9.37% 6.02% 1.57% I.H. Lipumba (CUF) 26,640 6.50% 2.00% -9.76% John Cheyo (UDP) 13,157 3.21% 3.84% -0.98%

Total 409,758 100.00% 5.01% 0.00% Singida Benjimin Mkapa (CCM) 198,905 83.48% 3.39% 11.74% A.L. Mrema (TLP) 5,716 2.40% 0.90% -5.40% I.H. Lipumba (CUF) 27,262 11.44% 2.05% -4.82% John Cheyo (UDP) 6,376 2.68% 1.86% -1.52%

Total 238,259 100.00% 2.92% 0.00%

472

Tabora Benjimin Mkapa (CCM) 200,667 61.83% 3.42% -9.91% A.L. Mrema (TLP) 8,986 2.77% 1.41% -5.03% I.H. Lipumba (CUF) 98,859 30.46% 7.44% 14.20% John Cheyo (UDP) 16,010 4.93% 4.67% 0.74%

Total 324,522 100.00% 3.97% 0.00% Tanga Benjimin Mkapa (CCM) 320,009 77.56% 5.46% 5.82% A.L. Mrema (TLP) 9,674 2.34% 1.52% -5.45% I.H. Lipumba (CUF) 78,248 18.97% 5.89% 2.70% John Cheyo (UDP) 4,655 1.13% 1.36% -3.07%

Total 412,586 100.00% 5.05% 0.00% Pemba North Benjimin Mkapa (CCM) 8,510 13.24% 0.15% -58.51% A.L. Mrema (TLP) 666 1.04% 0.10% -6.76% I.H. Lipumba (CUF) 54,964 85.49% 4.14% 69.22% John Cheyo (UDP) 156 0.24% 0.05% -3.95%

Total 64,296 100.00% 0.79% 0.00% Pemba South Benjimin Mkapa (CCM) 22,694 36.94% 0.39% -34.80% A.L. Mrema (TLP) 236 0.38% 0.04% -7.41% I.H. Lipumba (CUF) 38,399 62.51% 2.89% 46.24% John Cheyo (UDP) 104 0.17% 0.03% -4.03%

Total 61,433 100.00% 0.75% 0.00% Unguja North Benjimin Mkapa (CCM) 50,884 76.27% 0.87% 4.52% A.L. Mrema (TLP) 532 0.80% 0.08% -7.00% I.H. Lipumba (CUF) 14,897 22.33% 1.12% 6.07% John Cheyo (UDP) 405 0.61% 0.12% -3.59%

Total 66,718 100.00% 0.82% 0.00%

473

Unguja South Benjimin Mkapa (CCM) 40,613 84.71% 0.69% 12.97% A.L. Mrema (TLP) 167 0.35% 0.03% -7.45% I.H. Lipumba (CUF) 6,946 14.49% 0.52% -1.78% John Cheyo (UDP) 217 0.45% 0.06% -3.74%

Total 47,943 100.00% 0.59% 0.00% Urban West Benjimin Mkapa (CCM) 120,217 94.28% 2.05% 22.54% A.L. Mrema (TLP) 1,163 0.91% 0.18% -6.88% I.H. Lipumba (CUF) 5,148 4.04% 0.39% -12.23% John Cheyo (UDP) 978 0.77% 0.29% -3.43%

Total 127,506 100.00% 1.56% 0.00%

National Total Benjimin Mkapa (CCM) 5,863,201 71.74% A.L. Mrema (TLP) 637,115 7.80% I.H. Lipumba (CUF) 1,329,077 16.26% John Cheyo (UDP) 342,891 4.20%

Total 8,172,284 100.00%

Aggregate Statistics for C: Percentage of votes for each party

Range of total votes cast Benjimin Mkapa (CCM) High=94.3% (Mbeya), Low=13.2% (Pemba North) A.L. Mrema (TLP) High=41.3% (Kilimanjaro), Low=0.4% (Unguja South) I.H. Lipumba (CUF) High= 85.5% (Pemba North), Low=3.0% (Mbeya) John Cheyo (UDP) High=13.8% (Mwanza), Low=0.2% (Pemba South)

474

Mean Median Skew Benjimin Mkapa (CCM) 72.52% 77.09% -1.72 A.L. Mrema (TLP) 6.20% 3.17% 3.18 I.H. Lipumba (CUF) 19.11% 14.05% 2.45 John Cheyo (UDP) 2.17% 1.71% 3.71

Aggregate Statistics for D: Regional contribution to party support

Range of total votes cast Benjimin Mkapa (CCM) High=9.7% (Urban West), Low=0.2% (Pemba North) A.L. Mrema (TLP) High=24.0% (Kilimanjaro), Low=0.0% (Unguja South) I.H. Lipumba (CUF) High= 19.4% (Dar es Salaam), Low=0.4% (Urban West) John Cheyo (UDP) High=24.4% (Mwanza), Low=0.0% (Pemba South)

Mean Median Skew Benjimin Mkapa (CCM) 4.08% 3.88% 0.22 A.L. Mrema (TLP) 3.89% 1.52% 2.57 I.H. Lipumba (CUF) 3.82% 2.36% 2.89 John Cheyo (UDP) 2.63% 1.48% 4.29

475

APPENDIX F REGIONAL RESULTS FOR THE 2000 PARLIAMENTARY ELECTIONS

Legend: Column A: Party. Column B: Total votes cast for each party. Column C: Percentage of B within each party. Tells how well the party did relative to other parties. Column D: Percentage of B in proportion to the total votes received by each party. This column tells which regions contributed the most votes for each party. Column E: Comparison between regional percentage won by each party, versus national percentage points won. Tells the degree to which a region can be considered a party stronghold percentage won by each party.

Data A B C D E

Arusha 1) CCM 101,719 60.82% 2.20% -4.37% 2) CUF 4,230 2.53% 0.48% -10.01% 3) TLP 30,594 18.29% 4.69% 9.10% 4) CHADEMA 29,320 17.53% 9.75% 13.30% 5) UDP 793 0.47% 0.25% -3.97% 6) NCCR-M 597 0.36% 0.23% -3.26% 7) Other 0.00% 0.00% -3.61% Total 167,253 100.00% 2.36% 0.80%

Dar es Salaam 1) CCM 404,751 54.95% 8.75% -10.24% 2) CUF 252,532 34.28% 28.37% 21.75% 3) TLP 53,424 7.25% 8.19% -1.94% 4) CHADEMA 2,100 0.29% 0.70% -3.95% 5) UDP 3,755 0.51% 1.19% -3.93% 6) NCCR-M 8,972 1.22% 3.50% -2.40% 7) Other 11,105 1.51% 19.65% -2.11% Total 736,639 100.00% 10.38% -0.71%

Dodoma 1) CCM 171,421 76.16% 3.70% 10.98% 2) CUF 30,090 13.37% 3.38% 0.83% 3) TLP 3,525 1.57% 0.54% -7.62% 4) CHADEMA 16,354 7.27% 5.44% 3.03% 5) UDP 1,888 0.84% 0.60% -3.60% 6) NCCR-M 0.00% 0.00% -3.61% 7) Other 1,788 0.79% 3.16% -2.82% Total 225,066 100.00% 3.17% 0.00%

Iringa 1) CCM 198,985 83.78% 4.30% 18.59% 2) CUF 4,197 1.77% 0.47% -10.77% 3) TLP 15,280 6.43% 2.34% -2.76% 4) CHADEMA 2,784 1.17% 0.93% -3.06% 5) UDP 0.00% 0.00% -4.44% 6) NCCR-M 13,340 5.62% 5.20% 2.00% 7) Other 2,936 1.24% 5.20% -2.38% Total 237,522 100.00% 3.35% -0.44%

Kagera 1) CCM 235,414 57.57% 5.09% -7.61% 2) CUF 51,920 12.70% 5.83% 0.16% 3) TLP 101,141 24.74% 15.50% 15.54% 4) CHADEMA 0.00% 0.00% -4.23% 5) UDP 0.00% 0.00% -4.44% 6) NCCR-M 17,563 4.30% 6.84% 0.68% 7) Other 2,850 0.70% 5.04% -2.92% Total 408,888 100.00% 5.76% 0.10%

Kigoma 1) CCM 88,850 50.07% 1.92% -15.12% 2) CUF 0.00% 0.00% -12.54% 3) TLP 113 0.06% 0.02% -9.13% 4) CHADEMA 47,646 26.85% 15.85% 22.62% 5) UDP 528 0.30% 0.17% -4.14% 6) NCCR-M 39,889 22.48% 15.55% 18.86% 7) Other 430 0.24% 0.76% -3.37% Total 177,456 100.00% 2.50% 0.55%

Kilimanjaro 1) CCM 179,242 49.53% 3.87% -15.66% 2) CUF 4,697 1.30% 0.53% -11.24% 3) TLP 76,472 21.13% 11.72% 11.94% 4) CHADEMA 83,892 23.18% 27.91% 18.95% 5) UDP 0.00% 0.00% -4.44% 6) NCCR-M 17,324 4.79% 6.75% 1.17% 7) Other 273 0.08% 0.48% -3.54% Total 361,900 100.00% 5.10% 0.72%

477

Lindi 1) CCM 71,555 75.36% 1.55% 10.17% 2) CUF 19,476 20.51% 2.19% 7.98% 3) TLP 0.00% 0.00% -9.19% 4) CHADEMA 1,988 2.09% 0.66% -2.14% 5) UDP 929 0.98% 0.29% -3.46% 6) NCCR-M 999 1.05% 0.39% -2.56% 7) Other 0.00% 0.00% -3.61% Total 94,947 100.00% 1.34% 0.80% Mara 1) CCM 206,599 61.99% 4.46% -3.19% 2) CUF 2,735 0.82% 0.31% -11.72% 3) TLP 84,017 25.21% 12.88% 16.02% 4) CHADEMA 0.00% 0.00% -4.23% 5) UDP 0.00% 0.00% -4.44% 6) NCCR-M 39,909 11.98% 15.55% 8.36% 7) Other 0.00% 0.00% -3.61% Total 333,260 100.00% 4.69% 0.80% Mbeya 1) CCM 212,155 79.73% 4.58% 14.54% 2) CUF 836 0.31% 0.09% -12.22% 3) TLP 30,365 11.41% 4.65% 2.22% 4) CHADEMA 2,438 0.92% 0.81% -3.32% 5) UDP 886 0.33% 0.28% -4.11% 6) NCCR-M 17,080 6.42% 6.66% 2.80% 7) Other 2,337 0.88% 4.14% -2.74% Total 266,097 100.00% 3.75% -0.08% Morogoro 1) CCM 168,232 64.89% 3.63% -0.29% 2) CUF 50,135 19.34% 5.63% 6.80% 3) TLP 25,328 9.77% 3.88% 0.58% 4) CHADEMA 3,787 1.46% 1.26% -2.77% 5) UDP 2,160 0.83% 0.69% -3.61% 6) NCCR-M 7,803 3.01% 3.04% -0.60% 7) Other 1,798 0.69% 3.18% -2.92% Total 259,243 100.00% 3.65% 0.10%

478

Mtwara 1) CCM 100,383 89.02% 2.17% 23.83% 2) CUF 5,246 4.65% 0.59% -7.88% 3) TLP 2,023 1.79% 0.31% -7.40% 4) CHADEMA 0.00% 0.00% -4.23% 5) UDP 538 0.48% 0.17% -3.96% 6) NCCR-M 4,575 4.06% 1.78% 0.44% 7) Other 0.00% 0.00% -3.61% Total 112,765 100.00% 1.59% 0.80%

Mwanza 1) CCM 350,495 64.84% 7.57% -0.35% 2) CUF 47,375 8.76% 5.32% -3.77% 3) TLP 19,759 3.66% 3.03% -5.54% 4) CHADEMA 26,262 4.86% 8.74% 0.62% 5) UDP 88,517 16.38% 28.07% 11.93% 6) NCCR-M 2,262 0.42% 0.88% -3.20% 7) Other 5,877 1.09% 10.40% -2.53% Total 540,547 100.00% 7.61% -0.29%

Pwani 1) CCM 125,233 72.63% 2.71% 7.44% 2) CUF 42,270 24.51% 4.75% 11.98% 3) TLP 766 0.44% 0.12% -8.75% 4) CHADEMA 0.00% 0.00% -4.23% 5) UDP 2,394 1.39% 0.76% -3.05% 6) NCCR-M 0.00% 0.00% -3.61% 7) Other 1,774 1.03% 3.14% -2.59% Total 172,437 98.97% 2.43% -0.23%

Rukwa 1) CCM 52,204 72.50% 1.13% 7.31% 2) CUF 3,632 5.04% 0.41% -7.49% 3) TLP 14,953 20.77% 2.29% 11.58% 4) CHADEMA 0.00% 0.00% -4.23% 5) UDP 0.00% 0.00% -4.44% 6) NCCR-M 738 1.02% 0.29% -3.42% 7) Other 476 0.66% 0.19% -2.95% Total 72,003 100.00% 1.01% -3.65%

479

Ruvuma 1) CCM 83,675 70.65% 1.81% 5.46% 2) CUF 10,180 8.59% 1.14% -3.94% 3) TLP 19,050 16.08% 2.92% 6.89% 4) CHADEMA 5,537 4.67% 1.84% 0.44% 5) UDP 0.00% 0.00% -4.44% 6) NCCR-M 0.00% 0.00% -3.61% 7) Other 0.00% 0.00% -3.61% Total 118,442 100.00% 1.67% 0.80% Shinyanga 1) CCM 197,299 53.75% 4.26% -11.43% 2) CUF 15,952 4.35% 1.79% -8.19% 3) TLP 880 0.24% 0.13% -8.95% 4) CHADEMA 23,944 6.52% 7.97% 2.29% 5) UDP 125,383 34.16% 39.77% 29.72% 6) NCCR-M 1,635 0.45% 0.64% -4.00% 7) Other 1,943 0.53% 0.76% -3.08% Total 367,036 100.00% 5.17% -3.65% Singida 1) CCM 173,357 85.12% 3.75% 19.93% 2) CUF 19,113 9.38% 2.15% -3.15% 3) TLP 1,916 0.94% 0.29% -8.25% 4) CHADEMA 2,739 1.34% 0.91% -2.89% 5) UDP 2,480 1.22% 0.79% -3.22% 6) NCCR-M 1,622 0.80% 0.63% -2.82% 7) Other 2,445 1.20% 4.33% -2.41% Total 203,672 100.00% 2.87% -0.40%

Tabora 1) CCM 148,885 63.57% 3.22% -1.61% 2) CUF 58,832 25.12% 6.61% 12.58% 3) TLP 882 0.38% 0.14% -8.81% 4) CHADEMA 850 0.36% 0.28% -3.87% 5) UDP 15,297 6.53% 4.85% 2.09% 6) NCCR-M 3,059 1.31% 1.19% -2.31% 7) Other 6,388 2.73% 11.31% -0.89% Total 234,193 100.00% 3.30% -1.93%

480

Tanga 1) CCM 171,752 73.89% 3.71% 8.70% 2) CUF 52,210 22.46% 5.87% 9.92% 3) TLP 5,128 2.21% 0.79% -6.98% 4) CHADEMA 0.00% 0.00% -4.23% 5) UDP 724 0.31% 0.23% -4.13% 6) NCCR-M 0.00% 0.00% -3.61% 7) Other 2,629 1.13% 4.65% -2.48% Total 232,443 100.00% 3.27% -0.34%

481

APPENDIX G REGIONAL RESULTS FOR THE 2005 PRESIDENTIAL ELECTION

Data A B C D E

Arusha 1) J.M. Kikwete (CCM) 361,237 81.96% 3.96% 1.69% 2) F.A. Mbowe (CHADEMA) 67,618 15.34% 10.11% 9.46% 3) I.H. Lipumba (CUF) 5,324 1.21% 0.40% -10.47% 4) E.S. Mvungi (NCCR-M) 876 0.20% 1.57% -0.29% 5) A.L. Mrema (TLP) 3,571 0.81% 4.21% 0.06% 6) Other 2,111 0.48% 2.02% -0.44% Total 440,737 100.00% 3.88% 0.00%

Dar es Salaam 1) J.M. Kikwete (CCM) 671,134 70.59% 7.36% -9.68% 2) F.A. Mbowe (CHADEMA) 52,979 5.57% 7.92% -0.31% 3) I.H. Lipumba (CUF) 216,700 22.79% 16.33% 11.11% 4) E.S. Mvungi (NCCR-M) 1,038 0.11% 1.86% -0.38% 5) A.L. Mrema (TLP) 5,129 0.54% 6.04% -0.21% 6) Other 3,802 0.40% 3.63% -0.52% Total 950,782 100.00% 8.37% 0.00%

Dodoma 1) J.M. Kikwete (CCM) 491,792 87.47% 5.39% 7.20% 2) F.A. Mbowe (CHADEMA) 15,777 2.81% 2.36% -3.08% 3) I.H. Lipumba (CUF) 44,971 8.00% 3.39% -3.68% 4) E.S. Mvungi (NCCR-M) 1,978 0.35% 3.54% -0.14% 5) A.L. Mrema (TLP) 2,724 0.48% 3.21% -0.26% 6) Other 4,983 0.89% 4.76% -0.03% Total 562,225 100.00% 4.95% 0.00%

Iringa 1) J.M. Kikwete (CCM) 472,612 94.20% 5.18% 13.93% 2) F.A. Mbowe (CHADEMA) 17,691 3.53% 2.65% -2.36% 3) I.H. Lipumba (CUF) 4,776 0.95% 0.36% -10.73% 4) E.S. Mvungi (NCCR-M) 2,108 0.42% 3.78% -0.07% 5) A.L. Mrema (TLP) 1,165 0.23% 1.37% -0.52% 6) Other 3,339 0.67% 3.19% -0.26% Total 501,691 100.00% 4.42% 0.00% Kagera 1) J.M. Kikwete (CCM) 572,498 84.43% 6.28% 4.16% 2) F.A. Mbowe (CHADEMA) 49,867 7.35% 7.46% 1.47% 3) I.H. Lipumba (CUF) 39,710 5.86% 2.99% -5.83% 4) E.S. Mvungi (NCCR-M) 4,690 0.69% 8.40% 0.20% 5) A.L. Mrema (TLP) 5,489 0.81% 6.47% 0.06% 6) Other 5,809 0.86% 5.55% -0.06% Total 678,063 100.00% 5.97% 0.00% Kigoma 1) J.M. Kikwete (CCM) 345,298 79.61% 3.79% -0.66% 2) F.A. Mbowe (CHADEMA) 41,100 9.48% 6.15% 3.59% 3) I.H. Lipumba (CUF) 34,410 7.93% 2.59% -3.75% 4) E.S. Mvungi (NCCR-M) 7,685 1.77% 13.77% 1.28% 5) A.L. Mrema (TLP) 1,092 0.25% 1.29% -0.50% 6) Other 4,145 0.96% 3.96% 0.03% Total 433,730 100.00% 3.82% 0.00% Kilimanjaro 1) J.M. Kikwete (CCM) 344,539 72.42% 3.78% -7.85% 2) F.A. Mbowe (CHADEMA) 98,072 20.61% 14.66% 14.73% 3) I.H. Lipumba (CUF) 9,437 1.98% 0.71% -9.70% 4) E.S. Mvungi (NCCR-M) 1,896 0.40% 3.40% -0.09% 5) A.L. Mrema (TLP) 16,585 3.49% 19.53% 2.74% 6) Other 5,208 1.09% 4.98% 0.17% Total 475,737 100.00% 4.19% 0.00% Lindi 1) J.M. Kikwete (CCM) 215,717 73.99% 2.37% -6.28% 2) F.A. Mbowe (CHADEMA) 4,644 1.59% 0.69% -4.29% 3) I.H. Lipumba (CUF) 66,775 22.90% 5.03% 11.22% 4) E.S. Mvungi (NCCR-M) 1,061 0.36% 1.90% -0.13% 5) A.L. Mrema (TLP) 927 0.32% 1.09% -0.43% 6) Other 2,438 0.84% 2.33% -0.08% Total 291,562 100.00% 2.57% 0.00%

483

Manyara 1) J.M. Kikwete (CCM) 321,026 86.35% 3.52% 6.08% 2) F.A. Mbowe (CHADEMA) 39,834 10.71% 5.96% 4.83% 3) I.H. Lipumba (CUF) 7,171 1.93% 0.54% -9.75% 4) E.S. Mvungi (NCCR-M) 1,214 0.33% 2.17% -0.16% 5) A.L. Mrema (TLP) 988 0.27% 1.16% -0.48% 6) Other 1,536 0.41% 1.47% -0.51% Total 371,769 100.00% 3.27% 0.00% Mara 1) J.M. Kikwete (CCM) 354,613 77.95% 3.89% -2.33% 2) F.A. Mbowe (CHADEMA) 56,493 12.42% 8.45% 6.53% 3) I.H. Lipumba (CUF) 31,533 6.93% 2.38% -4.75% 4) E.S. Mvungi (NCCR-M) 4,425 0.97% 7.93% 0.48% 5) A.L. Mrema (TLP) 3,562 0.78% 4.20% 0.04% 6) Other 4,322 0.95% 4.13% 0.03% Total 454,948 100.00% 4.00% 0.00% Mbeya 1) J.M. Kikwete (CCM) 568,458 88.87% 6.23% 8.60% 2) F.A. Mbowe (CHADEMA) 24,524 3.83% 3.67% -2.05% 3) I.H. Lipumba (CUF) 18,382 2.87% 1.39% -8.81% 4) E.S. Mvungi (NCCR-M) 3,859 0.60% 6.91% 0.11% 5) A.L. Mrema (TLP) 18,028 2.82% 21.23% 2.07% 6) Other 6,427 1.00% 6.14% 0.08% Total 639,678 100.00% 5.63% 0.00% Morogoro 1) J.M. Kikwete (CCM) 543,837 85.71% 5.96% 5.44% 2) F.A. Mbowe (CHADEMA) 15,668 2.47% 2.34% -3.42% 3) I.H. Lipumba (CUF) 60,051 9.46% 4.52% -2.22% 4) E.S. Mvungi (NCCR-M) 2,167 0.34% 3.88% -0.15% 5) A.L. Mrema (TLP) 8,192 1.29% 9.65% 0.54% 6) Other 4,566 0.72% 4.36% -0.20% Total 634,481 100.00% 5.59% 0.00% Mtwara 1) J.M. Kikwete (CCM) 350,117 79.09% 3.84% -1.18% 2) F.A. Mbowe (CHADEMA) 9,145 2.07% 1.37% -3.82% 3) I.H. Lipumba (CUF) 73,667 16.64% 5.55% 4.96% 4) E.S. Mvungi (NCCR-M) 3,174 0.72% 5.69% 0.23% 5) A.L. Mrema (TLP) 2,270 0.51% 2.67% -0.23% 6) Other 4,281 0.97% 4.09% 0.05% Total 442,654 100.00% 3.90% 0.00%

484

Mwanza 1) J.M. Kikwete (CCM) 713,088 79.37% 7.82% -0.90% 2) F.A. Mbowe (CHADEMA) 59,142 6.58% 8.84% 0.70% 3) I.H. Lipumba (CUF) 106,558 11.86% 8.03% 0.18% 4) E.S. Mvungi (NCCR-M) 4,690 0.52% 8.40% 0.03% 5) A.L. Mrema (TLP) 2,571 0.29% 3.03% -0.46% 6) Other 12,385 1.38% 11.84% 0.46% Total 898,434 100.00% 7.91% 0.00% Pwani 1) J.M. Kikwete (CCM) 251,068 72.93% 2.75% -7.34% 2) F.A. Mbowe (CHADEMA) 5,642 1.64% 0.84% -4.25% 3) I.H. Lipumba (CUF) 83,821 24.35% 6.32% 12.67% 4) E.S. Mvungi (NCCR-M) 937 0.27% 1.68% -0.22% 5) A.L. Mrema (TLP) 672 0.20% 0.79% -0.55% 6) Other 2,122 0.62% 2.03% -0.30% Total 344,262 100.00% 3.03% 0.00% Rukwa 1) J.M. Kikwete (CCM) 259,421 89.87% 2.84% 9.60% 2) F.A. Mbowe (CHADEMA) 13,035 4.52% 1.95% -1.37% 3) I.H. Lipumba (CUF) 7,207 2.50% 0.54% -9.19% 4) E.S. Mvungi (NCCR-M) 1,620 0.56% 2.90% 0.07% 5) A.L. Mrema (TLP) 5,086 1.76% 5.99% 1.01% 6) Other 2,284 0.79% 2.18% -0.13% Total 288,653 100.00% 2.54% 0.00% Ruvuma 1) J.M. Kikwete (CCM) 304,611 84.57% 3.34% 4.30% 2) F.A. Mbowe (CHADEMA) 13,973 3.88% 2.09% -2.01% 3) I.H. Lipumba (CUF) 36,426 10.11% 2.74% -1.57% 4) E.S. Mvungi (NCCR-M) 1,890 0.52% 3.39% 0.03% 5) A.L. Mrema (TLP) 868 0.24% 1.02% -0.51% 6) Other 2,401 0.67% 2.29% -0.25% Total 360,169 100.00% 3.17% 0.00% Shinyanga 1) J.M. Kikwete (CCM) 621,975 74.54% 6.82% -5.73% 2) F.A. Mbowe (CHADEMA) 54,631 6.55% 8.17% 0.66% 3) I.H. Lipumba (CUF) 131,526 15.76% 9.91% 4.08% 4) E.S. Mvungi (NCCR-M) 4,231 0.51% 7.58% 0.02% 5) A.L. Mrema (TLP) 2,179 0.26% 2.57% -0.49% 6) Other 19,864 2.38% 18.99% 1.46% Total 834,406 100.00% 7.35% 0.00%

485

Singida 1) J.M. Kikwete (CCM) 319,870 90.90% 3.51% 10.63% 2) F.A. Mbowe (CHADEMA) 12,977 3.69% 1.94% -2.20% 3) I.H. Lipumba (CUF) 14,705 4.18% 1.11% -7.50% 4) E.S. Mvungi (NCCR-M) 1,193 0.34% 2.14% -0.15% 5) A.L. Mrema (TLP) 815 0.23% 0.96% -0.52% 6) Other 2,334 0.66% 2.23% -0.26% Total 351,894 100.00% 3.10% 0.00% Tabora 1) J.M. Kikwete (CCM) 343,549 76.02% 3.77% -4.26% 2) F.A. Mbowe (CHADEMA) 7,618 1.69% 1.14% -4.20% 3) I.H. Lipumba (CUF) 95,047 21.03% 7.16% 9.35% 4) E.S. Mvungi (NCCR-M) 1,643 0.36% 2.94% -0.13% 5) A.L. Mrema (TLP) 691 0.15% 0.81% -0.59% 6) Other 3,396 0.75% 3.25% -0.17% Total 451,944 100.00% 3.98% 0.00% Tanga 1) J.M. Kikwete (CCM) 472,504 84.77% 5.18% 4.50% 2) F.A. Mbowe (CHADEMA) 7,225 1.30% 1.08% -4.59% 3) I.H. Lipumba (CUF) 69,260 12.43% 5.22% 0.74% 4) E.S. Mvungi (NCCR-M) 2,469 0.44% 4.42% -0.05% 5) A.L. Mrema (TLP) 1,949 0.35% 2.30% -0.40% 6) Other 3,997 0.72% 3.82% -0.20% Total 557,404 100.00% 4.91% 0.00% Pemba North 1) J.M. Kikwete (CCM) 7,777 12.03% 0.09% -68.24% 2) F.A. Mbowe (CHADEMA) 130 0.20% 0.02% -5.69% 3) I.H. Lipumba (CUF) 55,360 85.66% 4.17% 73.98% 4) E.S. Mvungi (NCCR-M) 181 0.28% 0.32% -0.21% 5) A.L. Mrema (TLP) 80 0.12% 0.09% -0.62% 6) Other 1,096 1.70% 1.05% 0.77% Total 64,624 100.00% 0.57% 0.00% Pemba South 1) J.M. Kikwete (CCM) 12,844 21.15% 0.14% -59.12% 2) F.A. Mbowe (CHADEMA) 130 0.21% 0.02% -5.67% 3) I.H. Lipumba (CUF) 46,791 77.05% 3.53% 65.37% 4) E.S. Mvungi (NCCR-M) 213 0.35% 0.38% -0.14% 5) A.L. Mrema (TLP) 69 0.11% 0.08% -0.63% 6) Other 682 1.12% 0.65% 0.20% Total 60,729 100.00% 0.53% 0.00%

486

Unguja North 1) J.M. Kikwete (CCM) 49,566 75.53% 0.54% -4.74% 2) F.A. Mbowe (CHADEMA) 206 0.31% 0.03% -5.57% 3) I.H. Lipumba (CUF) 14,968 22.81% 1.13% 11.13% 4) E.S. Mvungi (NCCR-M) 296 0.45% 0.53% -0.04% 5) A.L. Mrema (TLP) 77 0.12% 0.09% -0.63% 6) Other 508 0.77% 0.49% -0.15% Total 65,621 100.00% 0.58% 0.00% Unguja South 1) J.M. Kikwete (CCM) 47,545 88.58% 0.52% 8.31% 2) F.A. Mbowe (CHADEMA) 128 0.24% 0.02% -5.65% 3) I.H. Lipumba (CUF) 5,717 10.65% 0.43% -1.03% 4) E.S. Mvungi (NCCR-M) 94 0.18% 0.17% -0.32% 5) A.L. Mrema (TLP) 32 0.06% 0.04% -0.69% 6) Other 158 0.29% 0.15% -0.63% Total 53,674 100.00% 0.47% 0.00% Urban West 1) J.M. Kikwete (CCM) 102,256 68.03% 1.12% -12.24% 2) F.A. Mbowe (CHADEMA) 507 0.34% 0.08% -5.55% 3) I.H. Lipumba (CUF) 46,832 31.16% 3.53% 19.48% 4) E.S. Mvungi (NCCR-M) 191 0.13% 0.34% -0.36% 5) A.L. Mrema (TLP) 90 0.06% 0.11% -0.69% 6) Other 430 0.29% 0.41% -0.63% Total 150,306 100.00% 1.32% 0.00%

National Total 1) J.M. Kikwete (CCM) 9,118,952 80.27% 2) F.A. Mbowe (CHADEMA) 668,756 5.89% 3) I.H. Lipumba (CUF) 1,327,125 11.68% 4) E.S. Mvungi (NCCR-M) 55,819 0.49% 5) A.L. Mrema (TLP) 84,901 0.75% 6) Other 104,624 0.92% Total 11,360,177 100.00%

487

Aggregate Statistics for C: Percentage of votes for each party

Range of total votes cast J.M. Kikwete (CCM) High=94.2% (Iringa), Low=12.0% (Pemba North) F.A. Mbowe (CHADEMA) High=20.6% (Kilimanjaro), Low=0.2% (Pemba North) I.H. Lipumba (CUF) High= 85.6% (Pemba North), Low=1.0% (Iringa) E.S. Mvungi (NCCR-M) High=1.8% (Kigoma), Low=0.1% (Dar es Salaam) A.L. Mrema (TLP) High=3.5% (Kilimanjaro), Low=0.1% (Unguja South) Other

Mean Median Skew J.M. Kikwete (CCM) 76.19% 79.49% -2.64 F.A. Mbowe (CHADEMA) 4.96% 3.61% 1.59 I.H. Lipumba (CUF) 16.89% 10.38% 2.51 E.S. Mvungi (NCCR-M) 0.47% 0.38% 2.72 A.L. Mrema (TLP) 0.64% 0.28% 2.43 Other 0.86% 0.78% 1.81

Aggregate Statistics for D: Regional contribution to party support

Range of total votes cast J.M. Kikwete (CCM) High=7.8% (Mwanza), Low=0.1% (Pemba North) F.A. Mbowe (CHADEMA) High=14.7% (Kilimanjaro), Low=0.2% (Pemba North) I.H. Lipumba (CUF) High= 16.3% (Dar es Salaam), Low=0.4% (Iringa) E.S. Mvungi (NCCR-M) High=13.8% (Kigoma), Low=0.2% (Unguja South) A.L. Mrema (TLP) High=21.2% (Mbeya), Low=0.0% (Unguja South) Other

Mean Median Skew J.M. Kikwete (CCM) 3.85% 3.78% -0.10 F.A. Mbowe (CHADEMA) 3.85% 2.22% 1.08 I.H. Lipumba (CUF) 3.85% 3.19% 1.83 E.S. Mvungi (NCCR-M) 3.85% 3.16% 1.29 A.L. Mrema (TLP) 3.85% 1.83% 2.40 Other

488

APPENDIX H REGIONAL RESULTS FOR THE 2005 PARLIAMENTARY ELECITONS

Data A B C D E

Arusha 1) CCM 252,794 64.16% 3.31% -5.99% 2) CHADEMA 87,904 22.31% 10.10% 14.31% 3) CUF 4,972 1.26% 0.32% -12.96% 4) NCCR-M 56 0.01% 0.02% -2.19% 5) Other 48,254 12.25% 8.18% 6.82%

Total 393,980 100.00% 3.62% 0.00% Dar es Salaam 1) CCM 597,961 61.21% 7.84% -8.94% 2) CHADEMA 91,760 9.39% 10.54% 1.39% 3) CUF 233,644 23.92% 15.10% 9.69% 4) NCCR-M 10,254 1.05% 4.28% -1.15% 5) Other 43,340 4.44% 7.35% -0.99%

Total 976,959 100.00% 8.98% 0.00% Dodoma 1) CCM 450,667 83.13% 5.91% 12.98% 2) CHADEMA 17,659 3.26% 2.03% -4.75% 3) CUF 56,530 10.43% 3.65% -3.80% 4) NCCR-M 1,180 0.22% 0.49% -1.98% 5) Other 16,112 2.97% 2.73% -2.45%

Total 542,148 100.00% 4.98% 0.00%

Iringa 1) CCM 418,278 84.77% 5.48% 14.62% 2) CHADEMA 32,557 6.60% 3.74% -1.41% 3) CUF 16,956 3.44% 1.10% -10.79% 4) NCCR-M 647 0.13% 0.27% -2.07% 5) Other 25,004 5.07% 4.24% -0.36%

Total 493,442 100.00% 4.54% 0.00% Kagera 1) CCM 397,439 67.40% 5.21% -2.75% 2) CHADEMA 50,332 8.54% 5.78% 0.53% 3) CUF 56,501 9.58% 3.65% -4.64% 4) NCCR-M 19,112 3.24% 7.99% 1.04% 5) Other 66,306 11.24% 11.24% 5.82%

Total 589,690 100.00% 5.42% 0.00% Kigoma 1) CCM 244,869 57.65% 3.21% -12.50% 2) CHADEMA 89,123 20.98% 10.24% 12.98% 3) CUF 11,266 2.65% 0.73% -11.57% 4) NCCR-M 77,271 18.19% 32.29% 15.99% 5) Other 2,245 0.53% 0.38% -4.90%

Total 424,774 100.00% 3.90% 0.00% Kilimanjaro 1) CCM 305,542 65.15% 4.00% -5.00% 2) CHADEMA 102,576 21.87% 11.78% 13.87% 3) CUF 8,904 1.90% 0.58% -12.32% 4) NCCR-M 9,440 2.01% 3.94% -0.19% 5) Other 42,522 9.07% 7.21% 3.64%

Total 468,984 100.00% 4.31% 0.00% Lindi 1) CCM 203,142 70.66% 2.66% 0.51% 2) CHADEMA 6,313 2.20% 0.72% -5.81% 3) CUF 65,081 22.64% 4.21% 8.41% 4) NCCR-M 0.00% -2.20% 5) Other 12,968 4.51% 2.20% -0.91%

Total 287,504 100.00% 2.64% 0.00%

490

Manyara 1) CCM 260,687 71.25% 3.42% 1.10% 2) CHADEMA 66,189 18.09% 7.60% 10.09% 3) CUF 3,571 0.98% 0.23% -13.25% 4) NCCR-M 29,492 8.06% 12.32% 5.86% 5) Other 5,934 1.62% 1.01% -3.80%

Total 365,873 100.00% 3.36% 0.00% Mara 1) CCM 159,427 54.18% 2.09% -15.97% 2) CHADEMA 68,190 23.17% 7.83% 15.17% 3) CUF 40,656 13.82% 2.63% -0.41% 4) NCCR-M 10,862 3.69% 4.54% 1.49% 5) Other 15,112 5.14% 2.56% -0.29%

Total 294,247 100.00% 2.70% 0.00% Mbeya 1) CCM 439,611 79.87% 5.76% 9.72% 2) CHADEMA 449 0.08% 0.05% -7.92% 3) CUF 24,273 4.41% 1.57% -9.81% 4) NCCR-M 18,411 3.34% 7.69% 1.15% 5) Other 67,698 12.30% 11.47% 6.88%

Total 550,442 100.00% 5.06% 0.00% Morogoro 1) CCM 486,128 76.84% 6.37% 6.69% 2) CHADEMA 22,777 3.60% 2.62% -4.40% 3) CUF 92,463 14.61% 5.98% 0.39% 4) NCCR-M 1,034 0.16% 0.43% -2.04% 5) Other 30,263 4.78% 5.13% -0.64%

Total 632,665 100.00% 5.82% 0.00% Mtwara 1) CCM 336,770 75.74% 4.41% 5.59% 2) CHADEMA 12,582 2.83% 1.44% -5.17% 3) CUF 77,229 17.37% 4.99% 3.15% 4) NCCR-M 16,985 3.82% 7.10% 1.62% 5) Other 1,077 0.24% 0.18% -5.18%

Total 444,643 100.00% 4.09% 0.00%

491

Mwanza 1) CCM 599,447 68.15% 7.85% -2.00% 2) CHADEMA 57,204 6.50% 6.57% -1.50% 3) CUF 177,396 20.17% 11.46% 5.95% 4) NCCR-M 0.00% -2.20% 5) Other 45,505 5.17% 7.71% -0.25%

Total 879,552 100.00% 8.08% 0.00% Pwani 1) CCM 224,334 66.42% 2.94% -3.73% 2) CHADEMA 16,821 4.98% 1.93% -3.02% 3) CUF 88,532 26.21% 5.72% 11.99% 4) NCCR-M 0.00% -2.20% 5) Other 8,072 2.39% 1.37% -3.03%

Total 337,759 100.00% 3.10% 0.00% Rukwa 1) CCM 192,847 68.15% 2.53% -1.99% 2) CHADEMA 40,721 14.39% 4.68% 6.39% 3) CUF 9,223 3.26% 0.60% -10.96% 4) NCCR-M 28,783 10.17% 12.03% 7.97% 5) Other 11,380 4.02% 1.93% -1.40%

Total 282,954 100.00% 2.60% 0.00% Ruvuma 1) CCM 271,835 75.67% 3.56% 5.52% 2) CHADEMA 24,078 6.70% 2.77% -1.30% 3) CUF 56,835 15.82% 3.67% 1.60% 4) NCCR-M 2,629 0.73% 1.10% -1.47% 5) Other 3,844 1.07% 0.65% -4.35%

Total 359,221 100.00% 3.30% 0.00% Shinyanga 1) CCM 526,892 63.47% 6.90% -6.67% 2) CHADEMA 52,831 6.36% 6.07% -1.64% 3) CUF 128,456 15.48% 8.30% 1.25% 4) NCCR-M 1,557 0.19% 0.65% -2.01% 5) Other 120,343 14.50% 20.40% 9.07%

Total 830,079 100.00% 7.63% 0.00%

492

Singida 1) CCM 295,770 86.09% 3.88% 15.94% 2) CHADEMA 18,570 5.41% 2.13% -2.60% 3) CUF 25,134 7.32% 1.62% -6.91% 4) NCCR-M 1,102 0.32% 0.46% -1.88% 5) Other 2,966 0.86% 0.50% -4.56%

Total 343,542 100.00% 3.16% 0.00% Tabora 1) CCM 321,959 71.31% 4.22% 1.16% 2) CHADEMA 5,651 1.25% 0.65% -6.75% 3) CUF 105,180 23.29% 6.80% 9.07% 4) NCCR-M 7,548 1.67% 3.15% -0.53% 5) Other 11,179 2.48% 1.89% -2.95%

Total 451,517 100.00% 4.15% 0.00% Tanga 1) CCM 433,499 79.78% 5.68% 9.63% 2) CHADEMA 4,492 0.83% 0.52% -7.18% 3) CUF 97,352 17.92% 6.29% 3.69% 4) NCCR-M 2,057 0.38% 0.86% -1.82% 5) Other 5,997 1.10% 1.02% -4.32%

Total 543,397 100.00% 4.99% 0.00% Pemba North 1) CCM 7,702 12.10% 0.10% -58.05% 2) CHADEMA 798 1.25% 0.09% -6.75% 3) CUF 53,918 84.70% 3.48% 70.48% 4) NCCR-M 301 0.47% 0.13% -1.73% 5) Other 939 1.48% 0.16% -3.95%

Total 63,658 100.00% 0.59% 0.00% Pemba South 1) CCM 12,506 20.94% 0.16% -49.21% 2) CHADEMA 219 0.37% 0.03% -7.64% 3) CUF 46,998 78.69% 3.04% 64.47% 4) NCCR-M 0.00% -2.20% 5) Other 0.00% -5.42%

Total 59,723 100.00% 0.55% 0.00%

493

Unguja North 1) CCM 47,978 74.77% 0.63% 4.62% 2) CHADEMA 292 0.46% 0.03% -7.55% 3) CUF 15,898 24.78% 1.03% 10.55% 4) NCCR-M 0.00% -2.20% 5) Other 0.00% -5.42%

Total 64,168 100.00% 0.59% 0.00% Unguja South 1) CCM 45,399 86.25% 0.59% 16.10% 2) CHADEMA 52 0.10% 0.01% -7.91% 3) CUF 6,041 11.48% 0.39% -2.75% 4) NCCR-M 58 0.11% 0.02% -2.09% 5) Other 1,085 2.06% 0.18% -3.36%

Total 52,635 100.00% 0.48% 0.00% Urban West 1) CCM 98,114 67.43% 1.29% -2.72% 2) CHADEMA 619 0.43% 0.07% -7.58% 3) CUF 44,335 30.47% 2.87% 16.25% 4) NCCR-M 533 0.37% 0.22% -1.83% 5) Other 1,895 1.30% 0.32% -4.12%

Total 145,496 100.00% 1.34% 0.00%

National Total 1) CCM 7,631,597 70.15% 2) CHADEMA 870,759 8.00% 3) CUF 1,547,344 14.22% 4) NCCR-M 239,312 2.20% 5) Other 590,040 5.42%

Total 10,879,052 100.00%

494

Aggregate Statistics for C: Percentage of votes for each party

Range of total votes cast CCM High=86.3% (Unguja South), Low=12.1% (Pemba North) CHADEMA High=23.2% (Mara), Low=0.1% (Mbeya) CUF High= 84.7% (Pemba North), Low=1.0% (Manyara) NCCR-M High=18.2% (Kagera), Low=0.0% (Uncontested areas) Other

Mean Median Skew CCM 67.41% 69.41% -2.06 CHADEMA 7.38% 5.19% 1.07 CUF 18.71% 15.04% 2.38 NCCR-M 2.78% 0.73% 2.53 Other 4.61% 3.50% 1.18

Aggregate Statistics for D: Regional contribution to party support

Range of total votes cast CCM High=7.9% (Mwanza), Low=0.1% (Pemba North) CHADEMA High=11.8% (Kilimanjaro), Low=0.0% (Unguja South) CUF High= 15.1% (Dar es Salaam), Low=0.2% (Manyara) NCCR-M High=12.3% (Manyara), Low=0.0% (Uncontested areas) Other

Mean Median Skew CCM 3.85% 3.72% 0.02 CHADEMA 3.85% 2.37% 0.76 CUF 3.85% 3.26% 1.58 NCCR-M 3.85% 0.57% 3.09 Other 3.85% 1.91% 1.89

495

APPENDIX I PROVINCIAL RESULTS FOR THE 1992 KENYAN PRESIDENTIAL ELECTION

Data A B C D E

Nairobi Danial A. Moi (KANU) 62,402 16.71% 3.19% -19.83% Kenneth Matiba (FORD-A) 165,533 44.31% 11.79% 18.10% Mwai Kibaki (DP) 69,715 18.66% 6.64% -0.95% Oginga Odinga (FORD-K) 75,898 20.32% 8.04% 2.69%

Total 373,548 100.00% 6.97% 0.00%

Coast Province Danial A. Moi (KANU) 200,596 64.61% 10.25% 28.07% Kenneth Matiba (FORD-A) 35,598 11.47% 2.54% -14.75% Mwai Kibaki (DP) 23,766 7.65% 2.26% -11.96% Oginga Odinga (FORD-K) 50,516 16.27% 5.35% -1.36%

Total 310,476 100.00% 5.80% 0.00%

North Eastern Province Danial A. Moi (KANU) 57,400 78.23% 2.93% 41.69% Kenneth Matiba (FORD-A) 7,440 10.14% 0.53% -16.08% Mwai Kibaki (DP) 3,297 4.49% 0.31% -15.12% Oginga Odinga (FORD-K) 5,237 7.14% 0.55% -10.49%

Total 73,374 100.00% 1.37% 0.00%

Eastern Province Danial A. Moi (KANU) 290,494 37.11% 14.84% 0.57% Kenneth Matiba (FORD-A) 80,515 10.29% 5.73% -15.93% Mwai Kibaki (DP) 398,727 50.94% 37.95% 31.32% Oginga Odinga (FORD-K) 13,064 1.67% 1.38% -15.96%

Total 782,800 100.00% 14.62% 0.00% Central Province Danial A. Moi (KANU) 21,882 2.13% 1.12% -34.41% Kenneth Matiba (FORD-A) 621,368 60.51% 44.25% 34.29% Mwai Kibaki (DP) 372,937 36.31% 35.50% 16.70% Oginga Odinga (FORD-K) 10,765 1.05% 1.14% -16.58%

Total 1,026,952 100.00% 19.17% 0.00% Rift Valley Province Danial A. Moi (KANU) 994,844 67.96% 50.84% 31.42% Kenneth Matiba (FORD-A) 274,011 18.72% 19.51% -7.50% Mwai Kibaki (DP) 111,098 7.59% 10.57% -12.03% Oginga Odinga (FORD-K) 83,945 5.73% 8.89% -11.89%

Total 1,463,898 100.00% 27.33% 0.00% Western Province Danial A. Moi (KANU) 217,375 41.47% 11.11% 4.93% Kenneth Matiba (FORD-A) 192,859 36.79% 13.73% 10.57% Mwai Kibaki (DP) 19,115 3.65% 1.82% -15.97% Oginga Odinga (FORD-K) 94,851 18.09% 10.05% 0.47%

Total 524,200 100.00% 9.79% 0.00% Nyanza Province Danial A. Moi (KANU) 111,873 13.97% 5.72% -22.56% Kenneth Matiba (FORD-A) 26,922 3.36% 1.92% -22.86% Mwai Kibaki (DP) 51,962 6.49% 4.95% -13.13% Oginga Odinga (FORD-K) 609,921 76.18% 64.60% 58.55%

Total 800,678 100.00% 14.95% 0.00%

497

National Total Danial A. Moi (KANU) 1,956,866 36.54% Kenneth Matiba (FORD-A) 1,404,246 26.22% Mwai Kibaki (DP) 1,050,617 19.62% Oginga Odinga (FORD-K) 944,197 17.63%

Total 5,355,926 100.00%

498

APPENDIX J PROVINCIAL RESULTS FOR THE 2002 KENYAN PRESIDENTIAL ELECTION

Source:

Data A B C D E

Nairobi Uhuru M. Kenyatta (KANU) 76,001 20.78% 4.14% -10.59% Mwai Kibaki (NARC) 279,705 76.49% 7.67% 14.18% Simeon Nyachae (FORD-P) 8,775 2.40% 2.54% -3.50% James A. Orengo (SDP) 891 0.24% 3.63% -0.18% David N. Waweru (CCU) 301 0.08% 2.99% -0.34%

Total 365,673 100.00% 6.25% -0.42% Coast Province Uhuru M. Kenyatta (KANU) 121,645 33.36% 6.63% 1.99% Mwai Kibaki (NARC) 228,915 62.78% 6.28% 0.47% Simeon Nyachae (FORD-P) 11,716 3.21% 3.39% -2.69% James A. Orengo (SDP) 1,539 0.42% 6.28% 0.00% David N. Waweru (CCU) 823 0.23% 8.18% -0.19%

Total 364,638 100.00% 6.23% -0.42% North Eastern Province Uhuru M. Kenyatta (KANU) 83,358 67.06% 4.54% 35.69% Mwai Kibaki (NARC) 34,916 28.09% 0.96% -34.22% Simeon Nyachae (FORD-P) 5,660 4.55% 1.64% -1.34% James A. Orengo (SDP) 297 0.24% 1.21% -0.18% David N. Waweru (CCU) 73 0.06% 0.73% -0.11%

Total 124,304 100.00% 2.12% -0.17%

Eastern Province Uhuru M. Kenyatta (KANU) 270,060 26.17% 14.71% -5.20% Mwai Kibaki (NARC) 748,273 72.51% 20.52% 10.20% Simeon Nyachae (FORD-P) 7,854 0.76% 2.28% -5.14% James A. Orengo (SDP) 3,465 0.34% 14.13% -0.08% David N. Waweru (CCU) 2,247 0.22% 22.33% 0.05%

Total 1,031,899 100.00% 17.63% -0.17% Central Province Uhuru M. Kenyatta (KANU) 308,012 30.26% 16.78% -1.11% Mwai Kibaki (NARC) 701,916 68.96% 19.25% 6.65% Simeon Nyachae (FORD-P) 4,441 0.44% 1.29% -5.46% James A. Orengo (SDP) 1,443 0.14% 5.88% -0.28% David N. Waweru (CCU) 2,053 0.20% 20.41% 0.03%

Total 1,017,865 100.00% 17.39% -0.17% Rift Valley Province Uhuru M. Kenyatta (KANU) 769,242 53.26% 41.90% 21.89% Mwai Kibaki (NARC) 624,501 43.24% 17.13% -19.07% Simeon Nyachae (FORD-P) 45,145 3.13% 13.08% -2.77% James A. Orengo (SDP) 3,826 0.26% 15.60% -0.15% David N. Waweru (CCU) 1,624 0.11% 16.14% -0.06%

Total 1,444,338 100.00% 24.68% -0.17% Western Province Uhuru M. Kenyatta (KANU) 143,101 21.54% 7.79% -9.84% Mwai Kibaki (NARC) 506,999 76.30% 13.90% 13.99% Simeon Nyachae (FORD-P) 9,073 1.37% 2.63% -4.53% James A. Orengo (SDP) 3,443 0.52% 14.04% 0.10% David N. Waweru (CCU) 1,825 0.27% 18.14% 0.10%

Total 664,441 100.00% 11.35% -0.17% Nyanza Province Uhuru M. Kenyatta (KANU) 64,471 7.60% 3.51% -23.78% Mwai Kibaki (NARC) 521,052 61.39% 14.29% -0.92% Simeon Nyachae (FORD-P) 252,488 29.75% 73.15% 23.85% James A. Orengo (SDP) 9,620 1.13% 39.23% 0.71% David N. Waweru (CCU) 1,115 0.13% 11.08% -0.04%

Total 848,746 100.00% 14.50% -0.17%

500

National Total Uhuru M. Kenyatta (KANU) 1,835,890 31.37% Mwai Kibaki (NARC) 3,646,277 62.31% Simeon Nyachae (FORD-P) 345,152 5.90% James A. Orengo (SDP) 24,524 0.42% David N. Waweru (CCU) 10,061 0.17%

Total 5,851,843 100.00%

501

APPENDIX K REGIONAL RESULTS FOR THE 1994 MALAWI PRESIDENTIAL ELECTION

Source: Kalipeni, 1997

Data A B C D E

North Hastings Banda (MCP) 33,650 7.30% 3.38% -26.14% Bakili Muluzi (UDF) 20,837 4.52% 1.48% -42.63% Chakufwa Chihana (AFORD) 404,837 87.80% 71.92% 68.91% Kamlepo Kalua (MDP) 1,754 0.38% 11.23% -0.14%

Total 461,078 100.00% 15.47% 0.00% Central Hastings Banda (MCP) 743,739 64.27% 74.65% 30.83% Bakili Muluzi (UDF) 321,581 27.79% 22.89% -19.36% Chakufwa Chihana (AFORD) 86,766 7.50% 15.42% -11.39% Kamlepo Kalua (MDP) 5,161 0.45% 33.03% -0.08%

Total 1,157,247 100.00% 38.84% 0.00% South Hastings Banda (MCP) 218,964 16.09% 21.98% -17.35% Bakili Muluzi (UDF) 1,062,336 78.04% 75.62% 30.89% Chakufwa Chihana (AFORD) 71,259 5.23% 12.66% -13.66% Kamlepo Kalua (MDP) 8,709 0.64% 55.74% 0.12%

Total 1,361,268 100.00% 45.69% 0.00%

National Total Hastings Banda (MCP) 996,353 33.44% Bakili Muluzi (UDF) 1,404,754 47.15% Chakufwa Chihana (AFORD) 562,862 18.89% Kamlepo Kalua (MDP) 15,624 0.52%

Total 2,979,593 100.00%

503