|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 GREAT LAKES REGION By CARA EUGENIA JONES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Cara Eugenia Jones
3 To James and Malcolm Out of the frying pan and into the fire.
4 ACKNOWLEDGMENTS This seven year process would never have gotten to this point without the advice, ) to finish a dissertation The chair of my committee, Benjamin Smith, deserve s more gratitude than I can give. Throughout this process he has been an inspiring mentor, a brilliant teacher, and confidant and friend. Improving my subject verb agreement, add ing to my repertoire of Indonesian recipes, and showing me how to be a good scholar, tutelage. The rest of my G oldman also deserve accolades. Ken always had a wise word and funny joke to tell, and influenced me greatly from the beginning of my studies. Will Reno and Zacariah Mampilly read chapters, helped with framing thoughts, and provided hours of entertainment usually at their own expense Tim Longman, Marc Sommers, and Peter Uvin generously shared their extensive knowledge of the Great Lakes and helped me to make many connections both before and during fieldwork. The Department of Political Science and the Center for African Studies at the University of Florida provided me
5 tremendous financial support, camaraderie, office space, a place to present new ideas, and supportive colleagues. The United States Department of Education Fulbright Hays Doctoral Dissertation Research Abroad committee to ok a chance on a project in a country that had not seen a fellow in decades, for which I am eternally grateful and could not have finished the project without. The University of Rochester generously provided me with a year to complete the majority of the w riting and my colleagues there, especially Bethany Lacina and Adam Cohon gave valuabl e insight and drafting advice. introducing me to life of a faculty member. Jennifer Ho lmes has been a friend, mentor and inspiration in the discipline and in life upon our first meeting ten years ago. This project was influenced in a thousand ways by the people I lived and worked with in Burundi. Samson Munahi greeted me and taught me abou t Burundi from my first t rip in 2008. Audifax was an inspi ration in the field and taught me how to eat urorobe like a Burundian. Jean Marie Ngendahayo, Aloys Batungwanayo, Johan DeFlanders, and Patrick Hajayandi gave lots of academic and practical advice a bout research, the topic, and life in Burundi during and after the war. Finally, my anonymous interview subjects, market sellers, bar patrons and students who generously shared their time, their lives, their food and drink, and their spirit with me. While changes everyone f or the better, it certainly did for me. My colleagues in graduate school, especially the notorious Africanist Mafia, earn special thanks here. Wynie Pankani, Ramon Galinanes and Steve Lichty all began graduat e school wit h me, Ann Wainscott and Nic Knowlton a couple of years later They have prov ided companionship, commiseration, and valuable feedback along the way.
6 Kevin Fridy, Aaron Hale, Ashley Leinweber, Hale were encourag ing and helpful, especially with navigating procedures and classes, c onferences, and the job market. Steve Marr deserves high praise for creating science with me, sharing internet videos, and all the Hobbit Christmas cards. My family, including my parents, my siblings, and my aunts, uncles, grandparents and cousins all provided spiritual, mental, and sometimes fina ncial support. Without the Beck Jones Fund for Wayward Scholars, my research would have suffered tremendously. The ability to afford over priced N utella in the fiel d is integral to good research. My mother, Mary, has been especially crucial to the process: She is and has always been a teacher, both as a profession and within our family. I have no doubt that it is this legacy that the work embodies. She also generously came to visit us in the field, bringing an extra suitcase of goodies and a bit of home, financed by and carrying the goodwill of my long suffering father, who not only had to deal with me as a child, but also continues to allow me to k vetch on a weekly basis. The family of my spouse (The Reaves, Prices and Hammetts) also generously accepted me and my desires to take their son to a far flung corner of the world and miss several holidays They provided prayer, material support and a place to crash mentally and physically, when we returned from the field. Finally, I would be remiss if I failed to mention the man behind the curtain. James Reaves has seen this process from idea to print, including the somewhat dubious honor of a year and half in Burundi, followed by a year in Rochester New York and all the temperature changes in between. As a spouse, partner, and friend, he never fails to steer me in the right direction, whether it be in regards to my writing or to my choice in
7 music for the car. The work that he has put into this project cannot be measured. It was truly a labor of love for him to accompany me on this adventure, and although at times the results were unexpected, it has never been anything other than extraordinary Our bea utiful son Malcolm Bennett Reaves gestated the same time I wrot e and finished the dissertation and provided his very own special form of moral support during the defense Although I have never been so sleep deprived, I have also never been more proud of al l of our achievements this year. It is to James and that most of th e credit for this work should go.
8 T ABLE OF CONTENTS ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 13 LIST OF FIGURES ................................ ................................ ................................ ........ 14 LIST OF ABBREVIATIONS ................................ ................................ ........................... 15 ABSTRACT ................................ ................................ ................................ ................... 22 CHAPTER 1 GIVING UP GREAT LAKES ................................ ................................ ................................ ....... 23 Introduction ................................ ................................ ................................ ............. 23 The State in Africa: Background to Rebellion ................................ ......................... 34 Why Rebels? ................................ ................................ ................................ .......... 40 The Argument ................................ ................................ ................................ ......... 43 Layout of the research ................................ ................................ ........................... 51 Tables and Figures ................................ ................................ ................................ 53 Notes ................................ ................................ ................................ ...................... 57 2 A THEORY OF REBEL RE GIME SUCCESS ................................ ......................... 60 Introduction ................................ ................................ ................................ ............. 60 Civil Wars and Rebellions ................................ ................................ ....................... 61 Contextualizing Civil Wars in Africa: Marginalization, Greed, And War: Why do they Rebel? ................................ ................................ ................................ ......... 64 Ethnicity and Conflict ................................ ................................ .............................. 70 Approaches to Civil Wars ................................ ................................ ........................ 75 Quantitative and Qualitative Methodological Tools in the Study of War .................. 78
9 State Formation and Rebel Organization: The Linkages ................................ ........ 80 ................................ .......... 82 The Conduct of War: The Contribution to Peacetime Success ............................... 86 Behaviors of Rebel Groups during Negotiated Settlements: The Keys to Success ................................ ................................ ................................ ............... 92 Theorizing Rebellion Success: After the Conflict Ends ................................ ........... 93 Counter Examples: Failed Groups ................................ ................................ ........ 100 Alternatives to the Theory ................................ ................................ ..................... 100 Conclusion ................................ ................................ ................................ ............ 106 Tables and Figures ................................ ................................ ............................... 107 Notes ................................ ................................ ................................ .................... 111 3 ANTECEDENT STATE CO NDITIONS, ORGANIZATIONAL AND SOCIAL POWER ................................ ................................ ................................ ................ 113 Introduction ................................ ................................ ................................ ........... 113 A Short History of Burundi ................................ ................................ .................... 115 A Political History since the Founding of the Republic ................................ .......... 123 Violence and Ethnic Conflict: Pre Cursors to Rebellion ................................ ........ 130 Independence and Violence ................................ ................................ ........... 131 ................................ ................................ ................... 131 The Interim Period, 1972 1988 ................................ ................................ ....... 132 Lead Up to Civil War 1988 1992 ................................ ................................ .... 133 La Cri se, 1993 ................................ ................................ ................................ 134 Rwanda ................................ ................................ ................................ ................. 136 Dual Colonialism: The Tutsi and Belgium ................................ ....................... 136 Political Changes over the post Independence period ................................ ... 140 Ethnic Conflict in Rwanda leading up to the Genocide, 1990 1994 ................ 142 The Rise of the RPF ................................ ................................ ....................... 146 Rwanda and Burundi: False Twins and Rebel Incubators? ................................ .. 148 Uganda ................................ ................................ ................................ ................. 150 Another Kingdom, Mul tiple Sovereignties ................................ ....................... 15 0 Independence ................................ ................................ ................................ 153 Early Changes and Dictatorship ................................ ................................ ..... 154 The Bush War and the Rising of the NRA ................................ ...................... 154 ................................ .... 156 Ugandan History in Comparison ................................ ................................ ........... 157 The Democratic Republic of the Congo ................................ ................................ 158 ................................ ................................ ..................... 159 Independence: Promise and Failure ................................ ............................... 161 ................................ .............. 164 The West and the East: Two Congos, too Many Problems ............................ 166 Kabila, Rwanda, and the AFDL ................................ ................................ ...... 167 Comparing Congo to the Rest of the Region ................................ ........................ 168 How does State History Matter to Rebellion? ................................ ....................... 169
10 Organizational Power ................................ ................................ ..................... 170 Legacies in Practice ................................ ................................ ....................... 170 The Nature of Social Power ................................ ................................ ............ 171 Grievances ................................ ................................ ................................ ..... 171 The State as an Intervening Force ................................ ................................ 172 Conclusion ................................ ................................ ................................ ............ 173 Notes ................................ ................................ ................................ .................... 174 4 TYPES OF ORGANIZATIONAL POWER AND THE IMPACT ON SUCCESS: HOW REBEL ORGANIZATIONS GAIN AND MAINTAIN INTERNAL COHESION ................................ ................................ ................................ ........... 176 Introduction ................................ ................................ ................................ ........... 176 A Political Hi story of the CNDD: The Founding ................................ .................... 178 Organization and Rebel Behavior by Period ................................ ......................... 183 1993 1996 ................................ ................................ ................................ ...... 184 1996 1998 ................................ ................................ ................................ ...... 187 Tentative Steps to Agreement ................................ ................................ ........ 188 The Break Up, Part One ................................ ................................ ................. 190 1998 2001 ................................ ................................ ................................ ...... 191 The Break Up, Part Two ................................ ................................ ................. 192 2001 2003 ................................ ................................ ................................ ...... 193 2003 2005 ................................ ................................ ................................ ...... 195 How did Organization change the outcome of the war? ................................ ........ 196 The Election 2005: How did Organization Shape the Outcome? .......................... 199 Changes in Leadership since 2005 ................................ ................................ ....... 202 Affaire du Radjabu, 2007 ................................ ................................ ................ 203 Problems with Legitimacy, Corruption, and the Leavers ................................ 205 The Cult of Peter ................................ ................................ ............................ 207 Election 2010 and Consolidation of the Regime ................................ ................... 209 Defining the Variable: Organizational Power ................................ ........................ 213 Leadership, Factions, and Splits ................................ ................................ ........... 213 Communications ................................ ................................ ............................. 214 Funding ................................ ................................ ................................ .......... 216 Tactics of Battle ................................ ................................ .............................. 218 Political Tactics ................................ ................................ ............................... 219 Present Day Organization ................................ ................................ .............. 220 The FNL and the CNDD FDD ................................ ................................ ............... 221 Co nclusion ................................ ................................ ................................ ............ 227 Tables and Figures ................................ ................................ ............................... 228 Notes ................................ ................................ ................................ .................... 229 5 SOCIAL RELATIONS: HOW TO MAKE OR BREAK A REBELLION .................... 234 Introduction ................................ ................................ ................................ ........... 234
11 Social Power: Civilians and State Structures ................................ ........................ 235 The History of Burundi and Civilians: Conflict, Continuity, Change ....................... 238 Buganda ................................ ................................ ................................ ............... 243 Rwanda ................................ ................................ ................................ ................. 245 Eastern Congo ................................ ................................ ................................ ...... 247 The Great Lakes States are Feudal, Too? ................................ ............................ 250 Social Power and Temporal Linkages in Civil Wars ................................ .............. 252 Civilian Support before the War ................................ ................................ ...... 253 A Parallel Movement: The FNL ................................ ................................ ...... 255 Civilian Support during the war (CNDD FDD and FNL) ................................ ........ 258 Civilian Support and Spreading the Message The role of the Mobilitizeur ..... 259 Targeting the population Hutus and Rural Poor ................................ ............ 261 The Abagumyabanga ................................ ................................ ..................... 262 ................................ ................................ .................... 264 The Youth League, Imbonerakure ................................ ................................ .. 266 State Structures and the CNDD FDD ................................ ............................. 267 Parallel Administrations ................................ ................................ .................. 268 Local Councilmen and Co Option of Members of Government ...................... 270 Creation of a Court System Forest Justice ................................ .................... 272 Planning for a Political Future ................................ ................................ ............... 273 Violence against Civilians during the Burundian Civil War ................................ .... 274 Hutu and Tutsi Violence ................................ ................................ ........................ 276 Rape and the CNDD FDD ................................ ................................ .................... 277 Violence against the Palipehutu FNL and its Supporters ................................ ...... 278 The Palipehutu FNL, Civilians, and the War Effort ................................ ............... 279 Civilian Support after the war (CNDD FDD) ................................ .......................... 281 The Building of a Political Party: Disparate Goals and New Cleavages ......... 281 The continuing use of the militia: Imboner akure as Intimidation and Enforcement ................................ ................................ ................................ 281 Why CNDD FDD and not FNL? ................................ ................................ ............ 282 Decentralization and Building Civilian Support in Civil Wars ................................ 284 Conclusion ................................ ................................ ................................ ............ 285 Tables and Figures ................................ ................................ ............................... 286 Notes ................................ ................................ ................................ .................... 289 6 REBELS, WAR, THE GREAT LAKES AND SUB SAHARAN AFRICA ................. 292 Introduction ................................ ................................ ................................ ........... 292 Why the Great Lakes? ................................ ................................ .......................... 293 NRA Rebel Strategies during the Civil War (Ugandan Bush War), 1981 1986. ................................ ................................ ................................ ........... 296 Rwanda ................................ ................................ ................................ ................. 302 Organizational Power in Rwanda ................................ ................................ ... 310 Social Power ................................ ................................ ................................ .. 311 The RPF in Comparative Perspective ................................ ............................ 314
12 1994 Present ................................ ................................ ................................ ........ 315 Comparing the RPF to the CNDD FDD as Post Conflict Governments. ............... 317 The AFDL: Herding Cats in Congo ................................ ................................ ....... 318 The Civil Wa r and the Great Congo War: The Demise of the Kleptocratic State ................................ ................................ ................................ ............ 322 Rebel Organizational Power: the AFDL ................................ .......................... 323 Relationships with Civilians: the AFDL, Rising Ethnic Conflict, and Managing Interests ................................ ................................ ...................... 325 The Dust Settles: The First Governance under AFDL, 1997 2001 ....................... 326 Conclusions about Case Studies ................................ ................................ .......... 329 Testing Rebel to Ruler Transitions Globally ................................ .......................... 330 Robustness Checks ................................ ................................ .............................. 339 Summary Analysis and Possible Data Challenges ................................ ............... 339 Chapter Conclusions ................................ ................................ ............................. 341 Tables and Figures ................................ ................................ ............................... 343 Notes ................................ ................................ ................................ .................... 349 7 CONCLUSION ................................ ................................ ................................ ...... 351 Introduction ................................ ................................ ................................ ........... 351 The Internal Logic of a Rebel Organization: Type of Organizational Power. ........ 355 Civilians, States and War: Social Power and Rebellion ................................ ........ 357 Rebels in Comparative Context ................................ ................................ ............ 358 General Contributions to Academia the Discipline ................................ ................ 360 Contributions to the Specific Civil Wars/Rebel Literature ................................ ...... 361 Conclusion ................................ ................................ ................................ ............ 364 APPENDIX A TIME LINE OF THE BURUNDIAN CIVIL WAR ................................ ..................... 365 B LIST OF ACTORS MENTIONED ................................ ................................ .......... 367 LIST OF REFERENCES ................................ ................................ ............................. 372 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 389
13 LIST OF TABLES Table page 1 1 Reform Rebel Insurgencies in Sub Saharan Africa, 1975 2009 ........................ 56 2 1 Aspects of Rebel Behavior and the Literature by Period ................................ 109 2 2 A New Typology of Rebel Behavior ................................ ................................ 110 6 1 Summary Statis tics of African Rebels Dataset ................................ ................ 343 6 2 Examples of Measures on Major Cases in Study ................................ ............ 343 6 3 Multinomial Regression Results ................................ ................................ ...... 344 6 4 Multinomial Regression Results with Addition of Independent Variable Age of Rebel Movement ................................ ................................ .......................... 344 6 5 Multinomial Regression Results using the Proxy Measure for Organizational Power ................................ ................................ ................................ ............... 345 6 6 Multinomial Regression Results holding Foreign Assistance constant ................................ ................................ ................................ 345 6 7 Multinomial Regression Results, the Effect of Peacekeepers ......................... 346 6 8 Robustness Check on Centralization and Leadership ................................ ..... 346 6 9 Robustness Check on Rebel Strength ................................ ............................. 346 6 10 Robustness Check on Rebel Capability ................................ .......................... 347 6 11 Robustness Check on the Natural Log of Rebel Strength ............................... 347 6 12 Robustness Check on Centralization and Leadership and the Natural Log of Rebel Strength ................................ ................................ ................................ .. 348 6 13 Robustness Check on Rebel Cap ability and the Natural Log of Rebel Strength ................................ ................................ ................................ ............ 348 6 14 Robustness Check on Rebel Capability and Centralization and Leadership ... 349
14 L IST OF FIGURES Figure page 1 1 ................................ ................................ ....... 53 1 2 ................................ .... 53 1 3 Administrative Map of Field Sites, Province and Commune Levels ................... 54 1 4 Administrative Map of Field Sites, Bujumbura City, Burundi .............................. 55 1 5 A Temporal Sequencing of Re bellion and Civil War ................................ .......... 57 2 1 Actors in the Burundian Civil War, 1993 1994 ................................ ................. 107 2 2 Actors in the Burundian Civil War, 1998 ................................ .......................... 107 2 3 Actors in the Burundian Civil War, 2002 ................................ .......................... 108 2 4 A Temporal Timeline of Civil Wars ................................ ................................ .. 109 2 5 Theoretical outline of Civil Wars ................................ ................................ ...... 110 2 6 Temporally Important Behavioral Variables during the scope of a Civil War ... 110 4 1 A Temporal Diagram of the Major Events Burundian Civil War, 1993 2005 .... 228 5 1 Diagram of Relations in Civil Wars ................................ ................................ ... 286 5 2 Temporalities and Interactions with Civilians ................................ ................... 287 5 3 Abagumyabanga membership cards, CNDD FDD ................................ .......... 287 5 4 Examples of Cotisation Receipts, 2001 2004 ................................ .................. 288
15 LIST OF ABBREVIATIONS ABAKO Bakongo Alliance ( Alliance du Bakongo ) ACLED Armed Conflict Locator Events Data set ADC Ikibiri Democratic Change Alliance ( Alliance pour Democratie et Change Ikibiri) ADP Alliance for Democracy and Progress ( Alliance pour Democratie et Progres ) AFDL Allied Forces for Democracy and Liberation, ( Alliance du Forces dmocratique pour la libration du Congo Zare ) AFRC Armed Forces Revolutionary Council, Liberia Al SHABAAB Mujahideen Youth Mov ement ( Harakat al Shabaab al Mujahideen ) AMC African Marine Commandoes, Cameroon APRD People's Army for the Restoration of Dem ocracy ( Populaire pour la restauration de la rpublique et la dmocratie ) AU African Union BAMPERE The Student Movement for Progressives in Burundi (Mouvement des Etudiants Progressistes Burundi) BINUB United Nations Integrated Office in Burundi ( Bureau Integree Nations Unies Burund i) CAR Central African Republic CDR The Committee for the Defense of the Republic ( Comite pour la Defense de la Republique ) CENI National Independent Electoral Commission ( Comision Electorale Nationale Independant) CNDD National Council f or the Defense of Democracy ( Conseil national pour la dfense du la dmocratie )
16 CNDD FDD National Council for the Defense of Democracy Forc es for the Defense of Democracy ( Conseil national pour la dfense du la dmocratie Forces pour la dfense du la dmocratie ) CNDP National Congress for the Defense of People ( Congres Nat ionale pour la dfense du le peuple ) CNRD National Council for the Resistance and Democracy ( Counseil National du Resistance pour la Democratie ) CONAKAT Confederation of the Tribal Association of Katanga ( Confederation des Associations Tribales du Katanga ) CPJP Con vention of Patriots for Justice and Peace ( Convention les patriotes pour la justice et la paix ) DRC Democratic Republic of Congo EAC East African Community EALA East African Legislative Assembly EIJM Eritrean Islamic Jihad Movement EPLF EPRDF/TPLF t/ Tigrayan FAN Northern Armed Forces ( Force Armes du Nord ) FAPC ( Force Armes du Congo ) FAB Burundian Armed Forces ( Force Armees Burundais) FAR Rwandan Armed Forces ( Force Armees Rwandaise ) FARC Revolutionary Armed Forces of Colombia ( Fuerzas Armadas Revolucionarias de Colombia) FAZ Zairian Armed Forces ( Force Armees Zairois) FARS Revolutionary Armed Forces of the Sah ara ( Force Armes Rvolutionnaire du Sahara )
17 FDD Forc es for the Defense of Democracy ( Forces pour la dfense du la dm ocratie ) FDLR/ Interahamwe Democratic Forces for the Liberation of Rwanda ( Forces dmocratiques de libration du Rwanda ) F DPC Democratic Front for the People of Central African Republic ( Front Dmocratique du Peuple du Centrafrique ) FDR Democratic Revival Front ( Front du Dmocratique Rveille ) FIAA Islamic Arab Front of Azawad ( Front Islamique Arabe de l'Azaouad ) FLAA Liberation Arab Front of Azawad (Front Libration Arabe de l'Azaouad) FN/MPCI New Forces/ Ivory Coast Patriotic Movement ( Force ) FNI National and Integrationist Fro nt ( Front des Nationalistes et Intgrationnistes ) FPL A Popular Liberation Front of Azawa d ( Front Populaire de Libration de l'Azawad ) FRPI Patriotic Resistance Forces, Ituri ( Front Resistance Patriotique du Ituri ) FRELIMO Mozambician Front for National Liberation ( Frente de Libertacao de Mocambique) FRODEBU The Front for Democracy in Burundi ( Front pour la Dmocrati au Burund i) FROLINA National Liberation Front ( Front de Libration Nationale) FRONASA Front for the National Salvation of Uganda FRUD Front for the Restoration of Unity and Democracy FUDC United Front for Democratic Change ( Front Unit du Dmocratie et Change )
18 GEDEBU Democratic Generation in Burundi ( Generation Democratique du Burundi ) HDI Human Development Index IMF International Monetary Fund INPFL Independent National Patriotic Front Liberia JRR The Revolutionary Youth of Rwagasore ( Jeunes Revolutionaires du Rwagasore ) KM Kikosi Maalum KY Kabaka Yekka (The King Alone ) LPC Liberia Peace Council MDR Democratic Republican Movement (Mouvement Democratique Republicain) MLC Congo Liberation Movement ( Mo u vement Liberation du Congo ) MLCJ Movement of Centra l A frican Liberators for Justice ( Mouvement des Cent rafrique librateurs pour la justice MNC National Congolese Movement ( Mouvement National Congolais ) MNJ Nigrienne Justice Movement ( Mouvement des Nigriens pour la Justice ) MONUC United Nations Organization Mission in Democratic Republic of The Congo ( Republique democratique du Congo) MPA Azawad People's Movement ( Mouvement Populaire de l'Azaouad ) MPLA Angolan Popular Liberation Movement Movimento Popular de Libertacao de Angola Partido do Trabalho ) MRC Congolese Rally Movement ( Mouvement Rassemblement Congolais )
19 MRLZ Zairian Revolutionanry and Freedom Movement ( Mouvement Revolutionnaire pour la Liberation du Zaire) MRND National Republican Movement For Democracy ( Mouvement Republicain National pour la Democratie ) MRNDD National Republican Movement For Democracy and Development ( Mouvement Republicain National pour la Democratie et Developpement) NPFL National Patriotic Front, Liberia NRM/NRA National Resistance Movement/ Army ONUC United Nations Operation in Congo ( Operation des Nations Unies au Congo ) PALIPEHUTU FNL Nationa l Forces of Liberation, Party for the Libe ration of the Hutu People ( Force nationale du liberation ) PALIR People in Arms for the Free dom of Rwanda ( Peuple en armes pour la liberation du Rwanda ) PARECO Resistant Congolese Patriots ( Patriotes Resistants Congolai s) PARMEHUTU Movement Party for the Emancipation of Hutu People, ( Parti du ) PDC Party of Christian Democrats ( Parti Democratique Chretien ) PRP Popular Revolution Party ( Parti Revolution Populaire ) PSA African Solidarity Party( Parti Solidaire Africain ) RaFD Rally for Democratic Fo rces ( Rassemblement des Forces Dmocratiques ) RANU Rwandese Alliance for National Unity RCD Congo Rally fo r Congolese Democracy ( Rassemblement Congolais pour la Dmocratie )
20 RCD Goma Rally for Congolese Democracy Goma ( Rassemblement Congolais pour la Dmocratie ) RCD Kisangani Rally for Congolese Democracy Kisangani ( Rassemblement Congolais pour la Dmocratie ) RENAMO Mozambique National Resistance Movement ( Resistncia Nacional Moambicana ) RPF( FPR) Rwandan Patriotic Front/Army, Rwanda ( Front Patriotique Rwandais ) RTLM Free Radio and Television of the Thousand Hills ( Radio Television Libre du Mille Collines) RUF Revolutionary Uni ted Front SANDANISTA Frente Sandinista de Liberacion Nacional SLA Sudanese Liberation Army SPLA UBU Labour Party of Burundi UCDP Uppsala Conflict Data Program UDSM University of Dar Es Salaam UFDD Union of Forces for Democracy and Development ( Union des Forces Dmocratiques pour le Dmocratie et Dveloppement ) UFDR Union of Democratic Forces for Unity ( Union des Forces Dmocratiques pour le Rassemblement ) ULIMO J United Liberian Movement for Democracy Johnson ULIMO K United Liberian Movement for Democracy Kromah UNAMIR United Nations Assistance Mission to Rwanda UNAR Rwandan National Union ( Union Nationale Rwanda ise ) UNITA National Union for the Total Independence of Angola ( Unio Nacional para a Independncia Total de Angola )
21 UN LA Ugandan National Liberation Army UPC UPC Union of Congolese Patriots ( Union des Patriotes Congolais ) UPC K Union of Congolese Patriots Kisembo ( Union des Patriotes Congolais ) UPC L Union of Congolese Patriots Lubanga ( Union des Patriotes Congolais ) UPRONA Party of Unity and National Progress ( Unite et Progres National ) UPDS The Union for Democratic and Social Progress ( Union pour la Democratie et le Progres Social ) ZANU Zimbabwe African National Union
22 Abstract of Dissertation Presented to the Graduate S chool of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy GREAT LAKES REGION By Cara Eugenia Jones May 2013 Chair: Benjamin B. Smith Major: Political Science The question of rebel to ruler transitio n is a fascinating line of inquiry connecting conflict behaviors and choices with the reality of post conflict governance. Rebels, eager to assume the mantle of government must often work with former opponents in combat, in addition to regional and int ernational actors, all of whom maintain wildly divergent goals in a fragile context. This study draws upon fieldwork in Centr al Africa, mostly conducted in Burundi, but also Rwanda, Uganda, and the Democratic Republic of the Congo, to assess how rebels become rulers, gaining and maintaining majority political power. It finds that the pathways to power are myriad, and rebel fates in the post conflict period are determined by specific types of power possessed and perfected by the rebel organization during the rebellion. Specifically, that the type of organizational power and source of social power and relations the rebel group purs ues helps to shape post conflict success when the goals of the rebel group shift from attaining power through force and into maintaining it within a democratizing context.
23 CHAPTER 1 GIVING UP THE G UN: REBEL TO RULER TRANSITIONS I LAKES Introduction A former CNDD re fighting every day against our enemies. We went hungry and cold because that is the way of war. We fought the soldiers when they found us. We did these actions because we wanted to be able to rule ourselves, and not let 1 After twelve years the Conseil National Pour la Dfense de la Dmocratie Forces pour la Dfense de la Dmocratie (CNDD FDD) gaine d power in Burundi. This interviewee's description of the reasons and behavior of war bring to mind both an old and a newly em erging question in rebel studies : why do some rebel groups succeed at post confli ct governance and others do not ? These questions spring from the disorder and disarray of post colonial politics over the last forty years in sub Saharan Africa. From the tim e when states came into their own in the 1960s, politics in the region became a contest for ethno regionalist or ct that it [the African state] has never had to really fight for its life, neither in order to come into existences, nor to (Kaarsholm 2006: 4) As former c olonies not born of war, African states and their politics were subject to different forces than those driving state politics in the conventional understanding of state development vis vis the history of Western Europe. Given this history, it is unsurprising that a vast number of rebel movements
24 have sprung up around the continent: at least tw o hundred distinct rebel movements have arisen since 1970 (UCDP 2011). These rebellions have translated into the dominant form of civil war since the end of the Cold War (Tull 2005:1) What is surprising is how successful some rebels hav e made the transiti on from non state anti government actors to government leaders and power keepers This is in direct contrast with other seemingly similar rebels cum rulers who have fai led to do so. This work explores this in a process of sequencing from civil war to post conflict state. I concentrate on the region of sub encompasses Rwanda, Burundi, Democratic Republic of Congo, Uganda and Tanzania, because of similarities in pre colonial legacies, post Independence politic s, and ethnic 2 : four of the five governments are former rebel movements that gained power through military victory or negotiated settlement out of a civil war. This chapter serves as an in troduction to my research and proceeds in four sections. Section one defines the research puzzle, the cases under study and the methods of inquiry used in the analysis in short, the basics of the research design. Section two gives the background to rebel lion, an explanation of the state and its role in concerns. Section three offers an introduction to the framework of the causal argument and finally, section four provides a detailed layout for the rest of the work The Research Puzzle In the Great Lakes and sub Saharan Africa more generally, the rebel turned political figure is an increasingly common phenomenon. While these individuals and groups can take on a range of roles following the ending of
25 conflict, I address a specific pathway to the possession of majority state power and political control. Civil wars have victors and lose rs. Therefore, this work examines a range of victor (successful) and loser (unsuccessful) rebel groups to determine what effect rebel behavior has on post reb ellion politics. It does so by focusing in particular on the CNDD FDD in Burundi to highlight the variables at play and the causal linkages in the rebel and post outcomes of violent conflict are (245) but while subsequent studies of civil wars have probed the causes, patterns, behaviors and solutions, post conflict outcomes have yet to receive significant attention. Significantly, the post conflict li terature fails to engage the importance of conflict behaviors upon post conflict decisions made by both rebel organizations and governments, instead focusing on how to further solidify the end of conflict and prevent new conflict emerging. In sub Saharan Africa, numerous rebel groups have emerged to challenge the state seeking an overthrow : some are ideological, some religious or ethnic in character, and some for material or resource gains (Reno 2011: ch. 1) with varying degrees of success or failure in military victor y against state forces To speculate as to how these groups with extreme military prowess came to power seems rather obvious: scholars argue th at the ending of conflict in a total war with a clear victor facilitates the post conflict (Luttwak 1999, Toft 2010, Dereun and Sobek 2004) This is especially true in collapsed or failed states where military power can be the only expression of legitimac y (e.g., Zartman 1995 and Rotberg 2003) But not all po st rebellion governments are th e military victors To reiterate the initial puzzle of this work, w hy and how do some unsuccessful rebel groups (those unable to
26 overthrow incumbent regimes by taking control of the financial and power locales of the state (R eno 2011: ch.1 ) become the post rebellion regime? Furthermore, how do these groups build and maintain power as the post memory? These are the questions central to this research The comparative framework provided in Figure s 1 1 and 1 2 (located at the end of this chapter) conc eptualize the cases discussed and their respective outcomes. The matrix (Figure 1 1) and map (Figure 1 2) illustrate a set of rebel organizations that show the range of outcomes in the Great Lakes regi on of transitions from rebel group fighting a government army to post rebellion regime. In these cases, some rebel groups that were successful at rebellion (meaning that they defeated the government and seized power) were able to transition to successful g overnments that were able to take, build and expand political power. A successful government in this study is one in which the rebels are able to gain political power following the cessation of conflict and then maintain political power through at least on e legitimate post conflict election, maintain political and social legitimacy among citizens, and resists overthrow or seizure of power by other rebel organizations or the military. Other rebel groups were not so fortunate in the struggle, and were unable to win control of the state or were unable to maintain control of the state post conflict. I suggest below that antecedent state conditions, the type of victory, and leaders are not the key to political transformation and longevity: but that the type of or ganization al power and the nature of social power of the group during the rebellion period determine post rebellion success, even when it is unable to take victory militarily Hence, the behaviors of the group during the rebellion are crucial to later succ esses as a government, a necessary but not
27 sufficient condition for successful transitions. The case of the CNDD FDD in Burundi illustrates a fundamental challenge to the traditional premise of rebel victory because it failed to take power decisively throu gh a civil war victory but yet built a stable and successful government. The CNDD FDD built a regime that has lasted for almost a decade and two election cycles, while at the same time resisting credible challenges to its power and ex panding its base of su pport: It i s these outcomes that are most central to the work Analysis of the CNDD ruler transitions that reflects the realities of civil wars in sub Saharan Africa. Moreover, it conf political goals, behaviors or strategies of the groups involved and thusly, suggests that history of the group now holding power matters to the future of the state. Case Selection This proje ct analyzes post rebellion governments in sub Saharan politics on three levels: the specific, the small N case study, and the larger comparative dataset First, I address the research puzzle of how and why a successful post rebellion government has arisen in Burundi under the CNDD FDD following an unsuccessful rebellion. I utilize various methods to illuminate specific features of the CNDD FDD transition that are then used to locate rebel to ruler transitions in two larger comparative contexts. The project focuses the bulk of the fieldwork regarding the various rebel groups operating in Burundi and situates them in broader set of rebel movements in the Great Lakes region. All of the countries in the region are near the elopment Indicator (HDI) ranking list, employment and production are mainly agricultural, and have experienced intensive civil war in the last twenty five years. Although these countries in which rebel movements
28 have become governments are multi ethnic, tw o in the region, Rwanda (under the RPF) (2003:78) wherein one ethnic group makes up at least forty five percent of the population. Ethnicity is not the main thrust of this work, but if as (1998: 9) then it seems appropriate that these blocked aspirations could be over ethnic grievances and changes in post c onflict could reflect this. It also seems appropriate to delineate places in which severe ethnic fractionalization and previous ethnic conflicts might impact post conflict outcomes, as in Rwanda (RPF) and Burundi, and to some extent, Uganda (under the gove rnment of the National Resistance Army (NRA)). A case study is not meant to stand as an expression of a particular phenomenon in one (Gerring 2004: 341) Accordingly, my cas e work on the CNDD FDD and the FNL Palipehutu ( Forces Nationales de Liberation Parti pour la libration du peuple Hutu the other active rebel group in Burundi ) illuminate the features of rebel cum ruler transitions that support successful and failed post rebellion politics. There are three additional countries and their rebel organizations that experienced rebel to ruler transition in the region from which I draw comparison: Rwanda with its formidable Rwandan Patriotic Front (RPF) the Democratic Rep ublic of Congo (DRC) and its ruling Alliance des Forces Democratique pour la liberation du Congo Zaire (AFDL) and Uganda with its National Resistance Movement/Army (NRM/A) Two represent significant improvements in economic development and have been laud ed by the international donor community for doing so
29 President Paul has suffered extreme dysfunction in the post conflict period under the AFDL and its ruler, President Joseph Kabila, and lacks basic capacity and institutionalization across C ausal mechanisms directly related to rebel behavior gleaned from the Burundian ca se study show similarities among the Great Lakes cases because there are fewer structural variations that could provide alternative explana tions. Finally, this research addresses rebel to ruler transitions in a broader manner, testing the implications and findings of the comparative case studies and fieldwork quantitatively. Here I utilize a dataset of all rebel groups in sub Saharan Africa from 1975 2009 and test quantitative models using the main hypotheses. I ntegrating multiple approaches in case select ion and methodology in the research allows for t he exploration of both descriptive and causal inference (King, Keohane, and Verba, 1994:9), generating a new study of post rebellion politics. The case selection proposed strives for internal and external val idity in research design. Internal validity is whether or not the inferences for a given case or set of cases actually apply to those cases (Brady and Collier 2004: 292) in effect, whether the concepts and tudy and can reliably provide causal inference for the question under study. External validity can be likened to can be applied to other settings and studies (ibid: 288) The work here provides for both a specific accounting of the rebel to ruler transition in Burundi, but then also attempts for application beyond these borders, by testing concepts, variables, and theories derived
30 from the case study to other case studies a nd broader cases across the continent. These tests bolster the findings of the work and show the range and depth of the theory of rebel transitions that begin to emerge here. Thus, the work strives for both internal validity in providing specific knowledg e about the CNDD FDD and external validity as it uses the knowledge generated to examine other cases. Method of Inquiry With an eye toward locating the processes that allow for successful rebel to ruler transitions, I apply multiple methods of analysis in the research: quantitative statistical techniques as well as extensive qualitative analysis based upon fieldwork. The case studies rely upon eighte en months of work between the years 2008 2011 in Burundi, Rwanda, the DRC, and Uganda. I conducted t he b ulk of the fieldwork in Burundi, focusing on multiple rebel organizations there the CNDD FDD, the current ruling party, and the FNL a failed rebel group, as well as a potential e merging new group that split from the CNDD FDD I carried out research in twelve primary field sites in five provinces, conducting two hundred and thirteen interviews from private citizens, government officials and politicians and collecting one hundred an d seventy two surveys fro m ex combatants. Provided in Figures 1 3 and 1 4 (at the end of the chapter) are the administrative field site maps illustrating data collection. I used several techniques to elucidate a causal framework accurately describing the history of the movement and particular critical junctures of decisions concerning the future strategy, including participant observation 3 vignette analysis, survey research, semi structured interviewing and questionnaires. Additionally, I analyzed archive s at both national and regional party headquarters of the CNDD FDD and its youth wing, the Imbonerakure, as well as archived newspapers Le Reneoveau Iwacu Burundi, and Arc
31 en Ciel, held in collection at the national university and the archives of the Unit ed Nations Integrated Office in Burundi (BINUB). Interviews were conducted individually or in small groups when it was appropriate, for example, with a husband and wife or other family members who knew each other well with my translator and me Although I spoke and used Kirundi, Kiswahili, and French in the interviewing process, a translator was necessary to interrogate the many cultural ly specific expressions Burundians use to describe life during the war. My translator was for all intents an d purposes Hut u, although of a Congolese father, and was of lower socio economic class and bore no identifying characteristics or distinguishing patterns of speech or lang uage that would mark him as wealthy or a government collaborator, or any other features that might put the data gathered from the interviews in jeopardy. Before conducting interviews, I obtained permission to research in Burundi with both the Interior Ministr y and the Ministry for Higher Education. In every the communal administration to inform them of our desire to do research in their areas and our intent. When av ailable, we also stopped by the office of the chef du zone structured in nature and very few subjects were considered out of scope, although I prepared questionnaire s befo rehand I let the interviewees dictate the length and location of interviews, with most lasting approximately two hours. I conducted follow up interviews where appropriate, especially with those subjects with specialized, historical knowledge of the CNDD F DD or the FNL When choosing interview subjects, we
32 randomly sampled from amongst the communes we visited, walking up and down the collines ( hills) and randomly choosing participants. We strove for a wide variety of ages and ethnicities to gain a full pict ure of life during the war and perspectives of rebellion there We followed this procedure in the urban and rural environments. Interview participants were offered no material compensation for their partici pation. Before and after fieldwork, I also relied on archives at the University of Florida as well as secondary and primary accounts of rebellion in the Great Lakes. I conducted semi structured interviews with leaders and foot soldiers of movements in Rwanda, Uganda and the Democratic R epublic of Congo, including exiles in Kenya and Tanzania, who possessed historical knowledge of these movements and their developments. I conducted work on groups operating in Eastern Congo by utilizing networks of former rebel combatants now living in Bur undi and supplemented these with archival work at the University of Nairobi. Additionally, I collected supporting primary and secondary sour ce material and search ed African news sources to locate key pieces of data about these rebel groups and their trans itions. Finally, I built a q uantitative data set by categorizing and coding according to the typologies developed in the qualitative case work for over sixty active rebel groups from the years 1975 2009 in sub Saharan Africa I determined possible rebe l groups through Encyclopedia 4 This data was used to test the findings of the qualitative work and provide preliminary evidence of a theory of rebel to ruler transitio ns. In the vein of Day, Clapham, Reno and other rebel studies scholars, I devise a set of four ideal types of rebel movements that categorize their behavior by their
33 organizational and social powers during the rebel period used to explain post rebellion po litics. Typology has become a de rigueur methodological tool in the analysis of rebel groups (Beardsley and McQuinn 2009) and more broadly in comparative politics and international relations (Collier et al. 2012 lists over 100) both for its parsimony in explanation and ideal type formation. A new typology was generated in this case to help identify post conflict behaviors, while most other typologies stop at identifying those during conflict. An explana tory typology method is useful for comparing rebels in this case because it allows for the testing of several claims about rebel characteristics and potential post rebellion outcomes in a way that can compound attributes across variables and show variations among in case comparisons as well as cross case co ntexts. Typology is an important tool to understanding rebel behavior because it can provide structure for understanding broad features of rebel movements in an idealized format that creates a standardized reference point. My typology allows for categoriza tion and measurement according to characteristics that have prev iously not been measured in relation to each other, thus combining technique and a new focus of rebel analysis The typology will be f urther elaborated in Chapter 2 but it is important at thi s juncture to est ablish how I coded the research and how to understand th e mechanics of the theory I present. Now that the details of the research have been discussed, this next section explore s the more substantive is sues of my research I explain the bac kground to which rebellion is made more likely in sub Saharan Africa, notably the weakness and malleability of the sta te and politics there. This provide s the
34 context for understanding how rebels emerge, with a further explanation of why rebels become the crucial actors in this research. The State in Africa: Background to Rebellion An examin ation of the state and its impact upon civilians in Africa is necessary to determine the environments in which rebels operate there and the potential impact that the Saharan Africa is often loosely defined and poorly controlled (Herbst 2000 and Boone 2003) by politicians, citizens, and outside observers. S ome argue, there are fewer barriers to reb ellion in weakened states (Rotberg 2003, Reno 2002), although some also argue that the very nature of post colonial African states (weak internally but still juridically sound externally) ensures that rebellion is unlikely (Englebert 2009). Further troubl ing definitional issues of the state in Africa are the embedded relationships between the regime and the state and the newness of democracy that are much more so than in juris In sub Saharan Africa, the lack of definitive authority over large swaths of territory (Herbst 2000:ch. 2) especia lly in the hinterlands, direc tly contributes to rebel movements. This is alluded to in the literature on rebellion in Africa highlighting the split between urban centers (the locus of state power) and the rural highlands where most of the population lives 5 Because of the lack of state control in the hinterlands, rebel groups can operate effectively in these areas without government interference. For example, the CNDD FDD in Burundi was formed in Bubanza, a notoriously under
35 governed p rovince in northwester n Burundi, which was considered a weakly govern ed area in a fairly weak state. The AFDL in Congo coalesced in the perpetual chaos of Eastern Congo where the state had completely failed from the immediate post Independence period According to Herbst, the n ature of state formation in sub Saharan Africa was historically more so about the control of people as opposed to the Western state formation tradition of control of territory (2000:183) It logically follows then that African rebels might strive first for control over populations a s opposed to territorial gains. This implies that rebels will necessarily undertake processes of civilian relationship building to provide more support for their political goals and desired outcomes. The logi c of state formation in Africa wa s arguably fundamentally different from that in Western Europe, in that the boundaries of modern African states were not determined by victories and expansion during war or by voluntary inclusion of kingdoms, but by the Berlin Conference of 188 4 1885, wherein colonial powers divided lands and settled boundaries. Very few changed boundaries or new states have emerged from these divisions, which often set in place arbitrary divisions between co ethnics, divided pre existing traditional political u nits, and united disparate populations under one state. Because of this birth states and their power apparatuses, even in the post Independence period, are often viewed by the population as arbitrary and historically imposed by an alien regime during colonialism. Furthermore, because of illogical border divisions and forced multicultural make ups, loyalties can tend to favor kinship, tribal or regional identifications first and national identities much later. States and their laws, structures, and lead ers floated above societies and often failed to capture the peasantry
36 according to Hyden (1980, 1986) These conditions as described above often lead to disconnect between civilians and governments, contributing to later patterns of rebellion and anti gove rnment sentiment in post colonial political development. While some scholars of African politics may attribute this to all parts of sub Saharan Africa, many variations of these attributes, including strong states with durable power and obedience throughout history, like Rwanda and Ethiopia, exist in the region. The legacy and interaction of the nature of state power versus the ability of the rebel organization to first break and then imitate such power is a key component of this study, and will be discussed further. Now to discuss one of the key tenets in the literature on African politics: the state has been the source of neo patrimonialism, a form of abuse of state resources for private or personal gain. N eo patrimonialism must be discussed i n conjunction with rebellion because many rebel groups cite specifically anti patrimonial goals or seek the destruction of previous networks that privileged one group abo ve another. In Burundi, this government exemplified this by Tutsi Hima domination of one particular region under the military presidents until 1993, a feature that both the FNL and the CNDD FDD mention in anti government propaganda of the time Neo patrimonialism, or the relations with other individuals as a means of achieving goals that seem otherwise (Hyden 2006:186). S tate employees and politicians provide resources (money, material, goods) to their co ethnics, family, friends and others to engende r relationships that exist outside of the state formalized networks. There are two dimensions to this relationship: one, the existence of this network outside of the
37 state, and two, the linkages in these networks based on other forms of association ethnici ty, regional association, religion, or kinship. Given the practices of this form of politics in most of Africa one may suspect that rebel groups may use violence for the purposes of the dismantling of old economies of affection and the creation of new one s. When rebels do take power, c ertainly some rewarding of the spoils of victory according to supporters and group members occurs, and the literature suggest s this is the case in also in Uganda with the positions of power. This evidence does not conclusive ly indentify the aims of the rebellion and its post rebellion success, but deserves further examination non etheless to highlight relationships between rebel groups, civilians and governments. In the work further motivations of profit and personal gain will be explored amongst the rebel groups under study to see the effect these particular motives have on behav ior and outcomes. Further compounding the problems of the state in Africa are porous borders between states and the predation of external rebels across state borders that make politics and civil wars more complex. Lack of security in many border regions al lows civilians to take advantage of this to engage in commerce in multiple countries. These conditions matter in peace as well as war, but in war, external actors often play bigger roles than they would in normal commerce Furthermore, porous borders mean t hat rebel groups can self finance by engaging in predation and looting of resources across borders, or alternatively, can become their own exporters of formerly state controlled resources. Consider for example the case of the AFDL in Congo, which exported minerals out of the Eastern region of the country through Rw anda and Uganda to
38 finance its insurgency, a tactic that was mimicked and later practiced against them by several upstart rebel groups in the second Congo War. Many of the rebe ls listed below in T able 1 1 at the end of this chapter (seventy percent) have received some form of outside aid either from co ethnics across borders or supportive neighboring governments, who might also be ethnically or kin related. Cooperative neighbors (Saleyhan 2007, 20 10) able to shuffle resources diminish the potential costs of warlordism 6 as rebel groups can take advantage of these nearby illicit markets. Previous statistical studies (Fearon and Laitin 2003) noted the role that anocracy 7 plays in contributing to dec isions to rebel. The weakness of the state in Africa in regards to its inability to consolidate power and strengthen institutions also allows for rebel groups to mount successful challenges and takeovers H owever, while states in Africa are among the weake st in the world, almost three fourths of the rebel group challengers provided in figure Four resulted in a failure to gain state power through military victory. Hence, the conditions of are not explanator y for the purposes of rebel to ruler transitions. This essentially means that despite confronting states notoriously understood to project little in terms of power African rebels still typically lose in a struggle against the state If the fundamental weakness of the s tate in Africa was an explanation for post rebellion success, then the continent would see many more internal challenges and potentially, more turnovers through rebellion and civil war in addition to se cessation s The weakened state provid es the environment for a power grab (Wilkinson 2004) but it is not sufficient to result in rebel victory. Furthermore, the operational environment of a weakened state creates just as many challenges for a rebel group turned government as it did advantages when
39 the roles were reversed. A civil war does not wipe the slate clean of the previous Finally, it must be remarked upon that the state in sub Saharan Africa is not u niformly weak across all forty eight countries: certain states, like Rwanda, Burundi, and Ethiopia 8 had complex histories of centralized and highly powerful states extending from the pre colonial period. Interestingly enough, these states are ones that exp erienced successful rebel to ruler transitions. Thus, post colonial government experiences and shapes of the state matter to the questions of how and why rebellion occurs and why outcomes are different. This raises a question that my theory of rebel to rul er transitions also addresses: namely, that a successful post rebellion regime is determined by the capacity of the rebel group to interact with both civilians and previously existing or newly formed state structures. Thus, the behaviors of the rebel group are endogenous to th e prior condition of the state. I elaborate on this concept further in subsequent chapters. S ocieties with certain historical legacies of state society relations can then be co opted by rebel groups to provide support for both the rebel group and the post rebellion regime, ensuring success. For example, outside observers marveled at the speed in subsequent recovery from war (Gourevitch 1998), but scholars su ggest that this was due to the legacy of the Rwandan state and its ability to structure political, economic and social life rather than the actions of the RPF itself (Prunier 2009: ch. 3). This brief statement is not extensive in its explanation as to why some strong states experience rebellion and successful post rebellion regimes, and many weak states do not
40 experience rebellion at all, and hence, will be explored more e xtensively in Chapters 2 and 3 Why Rebels? It is of paramount importance to define the actors under study To that end, how referring to when discussing sub Saharan Africa? 9 A rebel movement, according to de state organization with clear pol itical objectives that contests a (2008:10), although this particular study does not address cases of rebe ls who desire se cessation This definition specifies a particular political objective to the rebel group: it does not assume indivisibility amongst group members or even that all group members are aware of the goals. The terms rebel and insurgent are often used interchangeably by scholars in rebel studies and generally mean anti state violent actors (Mampilly 2011: 9 11) In this work I use the word rebel as opposed to insurgent as a matter of personal taste. Clapham initially typologized rebels in sub Saha ran Africa in the late 1990s, with four ideal types of r ebel movements based on goals: liberation movements, separatist movements, warlord movements and reform insurgencies. Some scholars argue that the motivations and desires of rebel groups are so comple x as to warrant multiple and potentially competing typological identifications (Boas and Dunn 2007: ch. 1) By far in sub Saharan Africa, the most prevalent groups of rebels according to this typology are et postcolonial governments and
41 (Clapham 1998: 7) without express desires of personal gain and personalized power. Others argue th at these particular rebel groups are indicative of a particular time and pattern of politics put in motion by the third wave of democratization (Reno 2011: ch. 4 and Kirschke 2000) and that reform rebels should be treated as a separate category because of these foundations. Because of both their number and their temporally specific political goals I find it appropriate to Furthermore, to return to the issue of competing typologi es, many of the groups who come to be associated with definitions of warlordism (notoriously, the Revolutionary United Front ( RUF ) in Sierra Leone) or separatism ( Army ( SPLA ) in Sudan, for example) began their lives as rebels with reform interests in the states they operate against. Especially in this work, the goals and interests of a group help to determine outcome, and thus necessarily will be studied. Using this typology and definition, I identified sixty such movements in Africa since 1975, provided in Table 1 1. I do so at this juncture to illustrate the variation in geographic location and years of activity of these groups and show the sheer number of groups that have emerged in the modern era. I provide them at the end of this chapter along with the outcome of the civil wars in which they took part. The figur e show s the vast number of groups, differences in their longevity and outcome, and the various countries in whi ch they operate. The groups enumerated are used in the research to build a dataset of rebel groups to test hypotheses about transitions. Reform insurgencies provide a foundation as to the desires of the rebel group which can be expanded on to theorize in t his particular research It follows that reform
42 insurgencies provide a place to s tart looking at groups with explicit ideas and goals about the refo rm of the state, and thus we assume that taking power is of paramount importance to them. I focus on these g roups specifically because of these goals. I also assume something about the groups themselves and their future goals and focus beyond the immediacy of the rebellion and civil war. In this desire, reform rebels may be more likely than other rebel groups to have goals, plans and strategies for not only the rebellion, but the post rebellion period as well. The goal of the rebellion, however, is not of paramount importance to this research : the outcome of the drive to take power from the state is. These goals are often antecedent to the outcomes of rebellion, as illustrated in the matrix, but not all so called reform rebels have succeeded at this goal. The discordance between goal and outcome for reform rebels also raises another interesting issue that will be addressed later in the work that of failed groups and what their patterns of behaviors, dec isions, organizations, and power structure s illustrate about successful transitions. Hence, groups that failed to grab power both in Burundi as well as in other con texts will be examined to contrast with more successful transitions, to further isolate how a nd why social and organizational powers work to enable rebels to become governments. The section following lay s out the framework of the argument of the research detailing the variables under study and providing further explanation to the diagram laid out on page three and four. It also provide s a summary of the argument that will be elabor ated upon further in Chapter 2
43 The Argument Here briefly I define the pa rameters of the argument of my study including its conditions, assumptions, and temporality. I then detail my hypotheses and t he variables under study here. Finally, I conclude with a brief introduction to the argument. Scope c onditions of the a rgument : My argument makes an assumption about the groups under study explicated here: that the rebel group receives some sort of credible political power in the post conflict transition, no matter the outcome of the civil war (i.e., whether it ends by negotiated s ettlement or by outright military victory). I assume this in order to tailor the exact mechanisms at play those that specifically affect post conflict transitions from rebel to ruler. Hence, all groups that have failed at their quest for political p ower ( as illustrated in Table 1 1 in the large numbers of groups that will not be excluded from research however, as they can illustrate features of successful rebel to ruler transitions. My argument is one of political stability and post conflict outcomes. My research examines a temporally limited political space from civil war to transition to the immediate post conflict context (after transitional ceasefires but well before democratization or consolidation). To highlight the period under study, I illustrate the sequencing of rebel lion and civil war in Figure 1 1 Hypotheses: I contend that essential to understand ing the post conflict development of rebel groups are their organizational relations and capacity, and that the interactions of these two variables produce different outcomes in post conflict governance. In this research, I tested two hypotheses about the nature of rebel to ruler
44 transitions. Based on the brief literature review provided above, several factors became apparent about the nature of the African state, pow er, and would be rebel rulers. First, the nature of the state in Africa fundamentally shape s the ways in which rebels plan and carry out their reform insurgencies. Because of a lack of state building both before and after Independence, many opportunities arose for potential African reformists to act as proto state builders during the rebellion. These actions provide long ranging implications for post conflict state takeover for the would be victors, who can use the practices, strategies and power structures put in place during conflict to expand bases of support. Also significantly, would be re bel conquerors depend on the same resources for social power that states themselves rely on: historical state structures, legacies of the pre conflict state, civilian state relationships, and civilians themselves. Thus, I hypothesize the following about w ould be rebel rulers: H 1 : A rebel group with wider ranging organizational power is more likely to transition successfully from rebel to government. H 2 : R elationships to the civilian population and existing state struct ures matter to the successful post r ebellion behavior of a rebel group. Thus, groups with cooper ative, transformative social powers as opposed to exclusions from civilian populations and states are more likely to be successful in their rebel to ruler transitions. T he hypotheses speak to the two variables I find most important to the understanding of rebel to ruler transitions: organization al power and social power I assume that each variable is independent of the other, but allow for potential interaction between the two. To put it plainly either centralization and infrastructure built within the group or social capacity, but find it
45 unnecessary to possess both, and that groups that exhibit both characteristics would be the most able to achieve political order and change in their transitions. Definition of Variables: Here I define the dependent and independent variables under study. As stated above, the dependent variable is the outcome of post rebellion government, which can be defined as Success or No Success 10 While simply winning power can be defined easily enough when a rebel organization takes the capital and defeats the government army, it is not enough to define success. A rebel group must overcome a number of factors to be desc ribed as a successful post conflict government, including shoring up political support to win at least two post conflict elections, resist a return to serious violence, contain potential political and military threats to government success, and gain accept ance from the population as the legitimate, legal, and representative government. This is measured in this study by collecting interview data, perceptions, and international and domestic journalistic accounts of the rebel regime. I elaborate upon this further in Chapter 2 The independent variables are Organizational Power, and Social Power Additionally, there are components of another intervening variable within these two, that of the nature of the pre conflict state. While this last variable is not e xplicitly tested in the hypotheses, it is nonetheless important to the typology and I elaborate fu rther upon this in Chapter 3 Organiza tional Power is categorized as either Infrastru ctural or Cellular (Chapter 4 ), and Social Power as either Incl usive or E xclusive (Chapter 5 ) I provide a brief operationalization below. Organizational Power : By organization al power I refer to the level of centralization between the cadre, the foot soldiers and the high command in warfare in the armed wing of the rebel mov ement, and the relationship between civilians,
46 mobilizers, and political elites in the political structure of the rebel movement 11 I focus on these relationships of control and obedience because they most illustrate the capabilities of the internal infrast ructure of the rebel organization, and the abilities of the group to act in a proto state manner. Similarly, this organization translates correspondingly to the relationship of supporters, the mobilizers and the party vanguard in the post rebellion period. A brief example will help to illustrate this concept The RPF instilled a high level of centralization in the moveme nt from its development in the late 1980s wherein which all decisions would stem from the leader s at the top and all information would pass up and down the chain accordingly. This method of organization even continued after the primary leader Fred Rwigema was killed in action in October 1990 and leadership passed to Paul Kagame. I provide more examples of th is in Chapters 2,4 and 6 but this penetrate civil society and to implement logistically political decisions throughout the finition of infrastructural power and substitute the rebel definition. Making these substitutions affects not the meaning of the definition of infrastructural power, but rather provides an alternative use applicable for non state actors. On the other hand, the AFDL was a highly decentralized movement that sprang from collaboration between Rwanda and four influential community figures from Eastern Congo (Stearns 2011). Bec ause of the very nature of the way the group formed, decisions came from multiple actors and locations, and it was unclear during the war exactly who was in charge either militarily or politically, often leading observers to
47 assume the RPF and Rwanda dicta ted orders to Congolese underlings. These types of decentralized, chaotic relationships often crumble under pressure and face serious challenges both during and after the war. Thus, these examples of organizational power within rebel groups can be classifie headless without a dominant power structure and with equality or unknown positionality between cells and without the complete penetration of both society and territory that other groups possess. While these categorizations suggest that anc ephalous organizational power is more likely a result of mismanagement and poor decision making by the rebel organization itself, it is entirely possible that a rebel group could pursue a cellular organizational power structure in order to minimize capture or information about the network, as expressed in cellular development of the terrorist network al Qaeda in the Middle East. I recognize that some scholars may code military and political organi zation differently and the implications of such difference are that distinct structures and methods of organization existed between different branches of the rebel group. Michael Mann, in his landmark work on states and social power, describes political powers as research I bundle military and political organization under the rubric of organizational power, because although rebel groups claim t o have separate branches, interviews revealed physical separation of combatants, politicians and civilians 12 Furthermore, many former rebel politicians described thes e separations as separation of powers similar to those in American democracy rather than truly separate entities 13 Discussing the political and
48 military organization under the umbrella term of organizational power also allows for more fruitful argumentatio n about infrastructural power within the rebel organization as it relates to participants, the state, and civilians. An example illustrates this: modern rebel movements in the post Cold War period, like that of CNDD FDD in Burundi and SPLA/SPLM in Sudan o ften develop in twofold: an armed wing (FDD and SPLA) and a political wing (CNDD and SPLM). They do so in this manner to both divide and focus energies into segments as well as in preparation for the post war context. The armed wings tend to be more centra lized and cohesive in organizational structure s and relationships, often comprised of rebels with military training, a highly centralized enterprise although, again, I do not find these post conflict desires reason to separate the wings in the discussion, in fact I find this evidence to bolster my claim that we should discuss both under one umbrella term. I measured the level of organization al power by developing political histories of the rebel movement and behavior in the field, and especially the deve lopment of these organizational networks in the transition period to when the rebel movement gained political power. In particular, the research will discuss structures of organization in regard to: leadership, funding, and tactics (both military and polit ical). I coded the level cohesive, centralized, and organized group that projects power over the entirety of its territory held. Cellular organization, on the other hand, denotes a type of organization only able to project power in a small radius around territories held, lacks
49 communications and cohesion between divisions, and fails to hold constant central authority. I elaborate upon these ideal types in Chapter 2. Social Power: Social power refers to the ability of the rebel organization to extract from, dismantle, co opt, and create structures both as a rebel organization during conflict, and later when they have assumed power as a governme nt. It also contains a measure of civilian relations and support. This matters to the reb conduct mobile campaigns with support and to the conduct of voter mobilization campaigns and levera ge of resources post conflict. Social Power can be thought of as either inclusive or exclusive of the civilians and the state. A rebel organization with inclusive social power becomes a cooperative and transformative actor upon the state and civilians, usi ng this source of power to bolster resources both during and after the conflict A group that experience d hig h levels of cooperative social power with civilians was the NRA/NRM in Uganda. During the civil war (1980 1986), civilians provided material suppor t to the rebels while in the bush, and the rebels in turn created civilian run governance councils in territorial areas under their control (Kasfir 2005). These relationships helped to build tremendous support for the NRA in the civil war and post civil wa r periods and a high level of support for Yoweri Museveni, the President of the movement and the country since 1986. Citizens felt strongly that they were part and parcel of the movement even in the early days, and this sense of camaraderie and group parti On the other hand, the RPF, again in Rwanda, had uncooperative and exclusionary relationships with civi lians during the war: using them as human shields, never allowing them to participate in the decision making structures, and some scholars have
50 suggested that the tactics of the RPF actually contributed to the massive number of deaths in the civil war and genocide (Stam and Davenport 2009) This continues to be a problem for the RPF in the post w ar context, as it is accused of only representing Ugandan Tutsi interests, and not those of interests of Tutsi who survived the genocide in Rwanda or the majority Hutu who remain in country An exclusive rebel group either cannot or is unwilling to rely o n social power as a means to provide support for the rebellion. The state and civilians can be ignored as a source of power, as in the case of the highly exclusive RPF in Rwanda, or cannot be tapped into because of the nature of the state or existing polit ics, like the AFDL in Congo. I measured the level of capacity by studying political histories and narratives of development of the rebel organization, as well as the use of voting data, participant observation, and archival research focused on the post con flict period. Of particular importance to the discussion of capacity are state structures and the ability of the rebel group to use, deform, or transform the state or community apparatus to the benefit of the group. Although these are not the same phenomen on, they are both related to the pre existing state (that the rebel organization seeks to replace), and thus can be considered in the same discussion. Also of importance to this variable are civilian relations and support, which is the level of support, f inancial, military or otherwise, provided by civilians to rebels tha t creates exclusive or symbiotic (inclusive) relationships with the group The idealized types of this variable are coded in this mann er, as either inclusive in the source of social power, or exclusive. Argument in Brief: I theorize that post rebellion behavior does not derive from whole cloth following the cessation o f a civil war, but rather, it is a funct ion of power
51 relations from the rebellion period. V ariations in post rebellion succ ess are directly related to variations amongst these sources of power The end of a civil war does not end rebel behavior and magically transform groups into democratic politicians, and thus, rebel behaviors matter to post conflict outcomes. Again, to return to the intervening variable in this research : the legacies of the state in these countries in which the rebel groups operate matter. They matter in that rebels can use pre existing state conditions to continue the disrupted state buildin g process, thus shoring up their own infrastructural power as well as social support. They also matter in that the very nature of the state structures the way civilians respond to potential authority figures, affecting both the source of social power for t he rebel group as well as the potential for penetration by the group into territory under its control. I return to these points and elaborate upon them in Chapter 5 I contend th at social power and organizational power are both sufficient but not jointly n ecessary conditions for successful (although whether they are the same type of success is debatable) post rebellion outcomes. Groups with higher levels of organization with more cohesive internal structures can and do win political power, a nd this organiza tion is vital to explaining successful rebel to ruler transition s Groups that engage in cooperative relationships with civilians and co opt, subvert, or demolish and replace pre existing state structures and relations also win political power, and this ca pacity is paramount to explaining their successful rebel to ruler transition. Layout of the Dissertation The research will proceed in six additional chapters. Chapter 2 offers a theory of post rebellion outcomes, explaining why some rebel groups are more likely to be
52 suc cessful and survive over time. It also situates the expected outcomes and the unexpected variations and further elaborates on the tempora lity of rebellio n. Chapter 3 explores the relationship between rebel characteristics and state and social characteristics, further illuminating the typology and the interplay of the type of social power under study Chapter 4 explores the transformation of the CNDD FDD fo llowing the signing of the peace agreement and the laying down of arms in 2003. It will detail the paradox of regime survivability despite factors that would seem to make this difficult. It will also provide an ethnography of the movement as seen from the perspectives of civilians, participants, and leaders, specifically detailing the CNDD the type of organizational power a rebel movement exhibits and the o utcome the rebel group achieves Chapter 5 dissects the nature of state power and state society structure vis vis the formation and outcome of rebel group organizations. Chapter 6 focuses on the comparative cases: the NRA in Uganda, the RPF in Rwa nda, and the AFDL in the Democratic Republic of Congo. I focus on these cases because the geographic area has remarkable political, social and economic similarities (especially between Rwanda and Burundi) that provide for comparison between the groups and the development of the typology f urther, and also expands the central logic of the argument to a larger group of rebel movements in sub Saharan Africa employing quantitative methods to test the core argument of the research Finally, in Chapter 7 I conclud e with a summary of the claims made and tested a nd a potential for future work.
53 Tables and Figures Figure 1 1. Figu re 1 2.
54 Figure 1 3. Administrative Map of Field Sit es, Province and Commune Levels
55 Figure 1 4. Administrative Map of Field Sites, Bujumbura City, Burundi
56 Table 1 1 Reform Rebel Insurgencies in Sub Saharan Africa, 1975 2009
57 Figure 1 5 A Temporal Sequencing of Rebellion and Civil War Note s 1 Personal Interview, March 21, 2011 2 Meaning an area prone to civil war there are 3 such areas in sub Saharan Africa the Great Lakes, the West A Central African states (Sudan, CAR and Chad).
58 3 I participated in political rallies, elections, and political meetings associated with these in the election period May September 2010. 4 UCDP lists incidences and actors of conflicts, state, non state, and one sided in the world since 1946 5 Although this is changing to reflect a more urban population, especially in larger African countries like Nigeria and Kenya. 6 Of which some scholars accus e a few in this study of doing: see Reno 1999 for discussion of the RUF and NPCL/APRC. 7 Anocracy is the gray space between democracy and authoritarianism as defined by measures on the Polity datasets. 8 All three of which have had recent successful rebel to ruler transitions, despite powerful state apparatuses 9 For clarity, the unit of observation is the rebel organization itself. 10 Refer back to figure 1 1 for an illustration. 11 As will be discussed further in the case study, many reform rebel groups have dual structures in conflict a political wing as well as a separate armed wing. While the lines between the two blur, the rebel group itself is quick to delineate the two. 12 Personal Interview, November 21, 2010
59 13 Personal Interview, February 21, 2011
60 CHAPTER 2 A THEORY OF REBEL REGIME SUCCESS Introduction In 1994, the fledgling Rwandan Patriotic Front (RPF) government inherited offices pockmarked by mortars, state hospitals with virtually no medical supplies, and a completely empty treasury. Its task seemed almost insurmountable to transform this fragile re bel movement turned government with very little in the way of physical infrastructure, to a state capable of ruling over a majority population hostile to RPF interests. To compound this environment, over three million Rwandese citizens living abroad, out o f fear or exile, needed to return to the homeland. However impossible this President Bill Clinton about the strength of resolve in the recovery, both economic and social, fr om the genocide. This chapter provide s a theor y of how groups like the RPF mak e the daunting transition from rebel s to ruler s It focus es i n particular on the structure in which rebels operate in: the civil war, its particularities and immediate aftermath, laying the groundwork for linking rebel behavior with post rebel behavior. It then proceed s from this background to explore the outcomes of rebellion and identifying k ey variables of rebel success. The next section discuss es rebellion temporally, from before the advent of violence to after, illustrating a theory of post rebellion behavior. This show s similarity and change in behavior over time. Finally the chapter explo res failed groups, those that did not gain political power, and potential challenges to the theory.
61 Civil Wars and Rebellions The environment in which a rebel group operates is not always clearly defined until the dust of the war has settled. A rebel grou p does not exist in a vacuum, and soon the boundaries of a recognized sovereign entity between parties subject to a common (Kalyvas 2006:5) Most of the wars of the late twentieth century have been of this kind of war, killing more than sixteen million and raging in more than seventy three countries (Fearon and Laitin 2003, Humphreys and Weinstein 2008: 436) Civil wars are ge nerally fought over longer periods of time than interstate wars (Fearon and Laitin 2003). Moreover, conventional wisdom has held that civil wars are also becoming increasingly international (Salehyan 2010), as neighbors are willing supporters of a rebel gr oup, and international or regional bodies step in to resolve conflict that threatens natural resources, regional stability, or borders. In sub Saharan Africa especially, over twenty countries, or forty percent of the continent, had experienced at least one civil war by 2000 (Elba dawi and Sambanis 2000: 244), and many have experienced recurring civil conflict and war in the decade since. Although it is easier to think of civil wars as wars between unitary actors, it is often not the case: in fact, most civi l wars have multiple actors on opposing sides. A government may fight with a civil defense unit or ruling party militia versus multiple rebel groups who may or may not be working together. It is also worth mentioning that actors can and often do, change ov er time. Thus, groups that are aligned at one point may not at another. Conversely, one group can become many, as the rebels faction or split. With advent of peace and power sharing agreements as an end to civil wars (Zartmann
62 1995, Fortna 2008), actors ca n change alliances or even sides of the conflict as thes e agreements are set in place. A visual example of this concept provided in Figure 2 1 will suffice for clarity and provide a diagrammatic view of the actors in the Burundian Civil War, at the beginni ng of the war in late 1993 to early 1994. However, relationships between actors can change dramatically over time. I provide a diagrammatic view of the actors in the Burundian Civil War at the midpoint of the war in 1998 in Figure 2 2 showing consolidation and conflict amongst rebel groups Finally, at the end of the war, peace agreements and power sharing, amongst other reasons, promoted the disappearance of some groups and the emergence of other fac tions. I illustrate this in Figure 2 3. This brief discu ssion of actors provide s context for the actors of importance in the Burundian Civil War, the civil war which provided the CNDD power discussed in grea ter detail in Chapters 4 and 5 Now that the environment and actors of a civil war has been established, I move on to a discu ssion of how the literature explains civil wars. The scholarly literature on civil wars and rebellion focused on the why and how of the war, especially causality or behaviors during conflict. Further, it has t reated these factors as contained to the scope of conflict, and not extending to post conflict outcomes. On the other hand, the post conflict transitions literature also downplays the importance of previous group behaviors during the war and instead focuse s on how conflict ends and the durability of peace after it does so (e.g. Walter 2004,Roeder and Rothchild 2005, and Toft 2010b) The s e inconsistencies lead to literatu res that compete in some ways, rather than complement
63 Now to address the civil wars lit erature: in recent years civil war scholars addressed questions of rebels and ethnic power (Fearon and Laitin 2003, Cederman et al. 2010) rebellion and financial support structures (Weinstein 2005, 2007) and relations between rebels and civilians (Wood 2 010, Mampilly 2011) why rebels take to arms against the state (Gurr 1970, Clapham 1998) predatory behaviors of rebels during conflict (Reno 1998, Collier et al. 2004, Fearon and Laitin 2003) rebel governance during conflict (Mampilly 2011) and even the (Day 2011) but still lacks when it comes to explaining post conflict outcomes. If, as Tilly (1975:42), then so too can war makers assumptions 1 suggest that when groups mo ve from the bush to the parliament, they attempt to consolidate power, monopolize the legitimate use of force, possibly achieve ethnic changes in leadership, and resist credible threats to their newfound l egitimacy. The literature weakens this position in translating the capacity for state building to rebel themselves, the shape, condition and scope of the prior state power and its apparatuses crystallizes and thus, cruc ial to this examination of the rebel group is the examination of the rebel group and its relations to this state power. The summary of the literature on rebel studies in Table 2 1 shows the comparative lack of studies conducted on these questions and the c ausal directions that lead to more successful and durable post rebellion regimes. As Table 2 1 shows, literature on rebellion has focused mostly on the causes of and behaviors during conflict and less on the rebel group themselves and future goals, to say nothing of the lack of studies pertinent to this conducted after the war ends. In
64 doing so, the literature ignores that rebels are not just creatures of warfare, but changing entities that emerge out of conflict but do not stay at war. Hence, we come to t he current weakness in civil war studies: the assumption that civil wars create a tabula Rosa for the actors, states, and governments that fight them (Bermeo 2003) This assumption brings serious challenges to our understanding of civil wars and rebellion: in doing so, the scholarly, policy and governing communities assigns static roles to these groups divided by time periods in conflict that do not accurately reflect reality. Groups can and often do change behaviors and roles as the war progresses or the d ynamics of vi olence and capabilities change. Yet very few scholars have investigated this discrepancy thoroughly 2 either in the conflict or the post conflict literatures This chapter presents a theory that directly challenges these assumptions. To begin this, however, I lay a foundation about the nature of war, rebels, and the state in conflict. Contextualizing Civil Wars in Africa: Marginalization, Greed, And War: Why do they Rebel? For all the above discussion of the lack of post war study, it is noneth eless still important to discuss why the civil war began in the first place. It is imperative to illustrate civil war motivations because these determine the behaviors, structure and logic behind actions undertaken in war and beyond. They shape who partici pates in the war, and how willing participants are to continue the battle, even when the war seems lost. Examining motivations can be used as a benchmark by which to measure the rebel organizations desired outcomes versus reality. As one scholar expounds, patterns of exit from the war [and thus post war outcomes] reflected the reasons for getting involv Interrogating the causes of the war also
6 5 provides evidence of the logic and actions rebel leaders undertook at critical junctures during the history of the war and rebellion. Scholars like Ted Gurr (1970), Fe aron and Laitin (2000), Reno (2011) and Sobek (2009) all examine the motivations and causes of war. To begin, motivation provides the first key to understand ing rebellio n as a process consisting of continuous nature of conflict and post conflict behaviors. An understanding of motivation shows causal mechanisms and potential ly illuminates post war desires. P ost war changes to governments can be evaluated agains t these motivations to measure success or failure of the rebel organization in changing government structures Lastly, motivation is central to understanding how and why civilians provide support to rebel groups, before, during, and after the conflict. This particularly in sub Saharan Africa, how marginalization leads to conflict. If scholars aspirati of literature contextualizing civil wars can show where and how rebels are marginalized and at what point violent conflict becomes a solution to these failed aspirations 3 Add itionally political and economic marginalization among victimized groups happens equally across the board, to both those that seek violence and tho se that do not. Rebel groups then use these grievance s to convince and coerce participation from otherwise n eutral or reluctant civilian actors. While there are many schools of thought regarding motivations for war, two primary ones in sub
66 of the political economy of war, and the second, gri evances, like those described above relating to marginalization. To begin, I discuss th e greed literature and its insufficiency when it comes to explaining civil wars. Some theorists posit that civil wars in sub (Kald or 1999) in which greed over material resources and not polit ical grievances drives conflict. Notable proponents of these economic conditions of the countries themselves (Collier and Hoeffler 2002: 22). Moreover new developments in trafficking and global resource markets provided expanded economic opportunities and incentives for African rebels to profit from conflict in their political and financial goals (Boas an d Dunn 2007: 10). C ertainly this trade is a facet of wars over the last two decades across the continent diamonds traded out of Liberia and Sierra Leone provided material incentive for warfare there, and the coltan, gold, and the industrial diamond trade in Co ngo is expressed as a primary cause or determinant of the war ( Eichstaedt 2011, Auteresse 2011, Stearns 2011). It is also worth noting the ability of rebels to subvert major trades in larger spheres of influence one only has to look at the influence of th e FARC rebels on the drug trade in the Western Hemisphere for confirmation of this. T he edited volume Greed and Grievance: Economic Agendas in Civil Wars explores the key assumptions emerge underlying the theory, attributed to Collier, of why greed is a st ronger motivation for engaging in civil war than are grievances. The theoretical models for greed based arguments find themselves grounded in rational choice, assuming that actors h ave all information necessary about potential costs and benefits to actions This assumption is interesting given that most of the recent work on greed and rebellion in Africa (Humphreys and
67 Weinstein 2008) deals with actors with low levels of education and interconnectedness with society. Critics often challenge the applicabili ty of this assumption, especially in societies like those in Sub Saharan Africa, which do suffer from information problems among citizens i n even the very best of cases. A greed based approach to motivation in civil war allow s for a more inclusive focus on the players of war, especially at the micro levels. These theoretical approaches are especially useful in this regard, as small actors are overlooked in conflict studies and in the political implications of outcomes of civil wars. N ew work on greed mot iv ations in conflict in Sub Saharan Africa has been especially illuminating in this regard showing the role of material incentive to the those that bring extraordinary wealth they can be as simple a s enough money to feed the would Access to high wealth resources is rare, even in the exceptionally mineral and natural resource wealthy continent of sub Saharan Africa, so an understanding o f a greed motivation must extend to even the lowest level of rebel. I find that natural resource greed is explained differently and applied more often in sub Saharan Africa than the standard explanations for greed motivations in civil war, and thus will discuss it in detail now. The greed thesis pointed to by media representations of war in Sub Saharan Africa and popularized in movies depicts warlords operating diamond mines Blood Diamond ) or controlling oil pipelines (as in the Niger Delta) and inducing often very young, very poor, and very uneducated men to fight for their share of these vast riches, far out of proportion with what the mine owners will receive. In academic literature, the model holds that
68 participants in civil war and viol ence are likely induced to participate by a desire for material wealth and security as opposed to an ideological motivation (Collier and Hoeffler 1998, 2001, 2004, Humphreys and Weinstein 2008). This thesis may be relied on more than it should because of t he frequent abundance of natural resources in the region. Of course c resource rich countries. Gam in many respects what we are seeing in the conflict zones of Africa is the playing out of rivalries for the control of sca In places as diverse as Sierra Leone (diamonds), Congo (diamonds, gold, timber, minerals), and Angola (oil), the presence of natural resources all provided i ncentive to rebellion, as rebel organizations diverted or exported these resources for financial benefit. However, the existence of this behavior in some cases of civil war does not account for whether the presence of natural resource wealth and a correspo nding paucity of that wealth to citizens make civil wars more likely. Indra de Soysa addresses the issue of natural resource wealth and scarcity in the same edited volume. H e finds the abundance of resources tends to increase civil violence, although a scarcity does not. The evidence provided for the rebel groups under study here both confirm and discredit this notion, as some rebellions take place in countries with resources (again, most notably the Democratic Republic of Congo) and others do not like Burundi. involvement, or even fuel conflict progression, greed alone cannot explain behavior before or during conflict and more importantly, the linkages that this research exa mines between behavior and outcome post conflict. If greed was an expla natory variable the
69 rebel group would continue to enrich itself after the conflict ended and it became a government, to no end other than that of increasing its own bounty. To be sure this is an accusation and many times a reality of post conflict politics: that the group in power ut ilizes state resources for its own material benefits (for example, Zaire under Mobutu) But these regimes are rarely post ist in the modern era of reform and international intervention. Olson also theorizes that bandits, whether stationary or roving, evolve towards a happy medium of extraction as they look to future prospects for rule (1991). Finally, as elaborated later, dif ferent rebel organizations pursue various strategies of organizational control and social relations: in doing so, varied patterns of extraction would also emerge (Weinstein 2007, Kalyvas 2006). Furthermore, if greed was substantially explanatory, ci vil war and rebellion would more likely break out in places where natural resource wealth was in abundance for example, Gabon or Botswana (all places where no war has occurred) instead of the most resource poor areas Burundi and Rwanda, where no significant natural resources exist. Finally, I elaborate later, greed may be an explanatory motivation for some of the participants of civi l war, but cannot be a group attribute or mass ideological position. It stands to reason that while there may be some leaders who actively pursue material wealth, the incentives and rewards between participation at the higher and lower levels of the rebellion are often very different. So it may be the case that greed fuels some, but not all decisions to engage in wa rfare. Even the literature on the subject tempered a bit in recent years, with scholars embracing correlations between economic conditions and war rather than direct causations As Paul Richards rightly describes
70 (2005:4) It is with this notion that I understand how rebel organizations pursue their goals of political power. Ethnicity and Conflict Other explanations of violence in Africa center on grievance mo dels, specifically on how ethnic conflict and marginalization drive violence. Ethnicity became a key explanation for violence and conflict in Africa because of a history of fractured ethnicities between many states. The Berlin Conference of 1884 1885 divid ed sub Saharan Africa according to decisions made by the Great Powers of Western Europe, and these divisions not only maintained their powers throughout colonialism but throughout the modern history of Africa as well. In these divisions, ethnic groups beca me divided among colonial boundaries and groups with little in common became one under the colonial power, like the hundreds of ethnic groups that suddenly composed Kenya, for example. African states were often cobbled together because of the needs of Euro pean colonizers, and the many varied groups living within these borders were expected to g et al ong and share control of the state following the end of colonialism, often with no overarching uniting ideology. Ethnicity, as Hyden explicates, can be both a m eans and end in politics. The focus in sub Saharan African studies concentrates on ethnicity as a means to an end, in the Great Lakes this end has typically been the gaining or retaining of state power (2006: 186). In fact, part and parcel to the gaining a nd maintenance of state power there was the answer to the question about the nature of the ethnic group state relationship. This was most frequently determined by contesting and rewriting history, with competing versions and narratives emerging out of poli tical disgruntlements. This presents itself every conflict in the Great
71 Lakes since Independence, and I unravel further in later chapters with particular attention to consociational character of relations between clans and ethnicities and the nebulousness of ethnicity as social order. Understanding the particular grievances at play in civil wars in Africa illuminates the historical incidences and narratives which rebel groups utilize to ga rner participants and support. These grievances often become part of the official founding statement and philosophy of the group, and remain a point of reference these even after the war ends. This merits brief consideration of the literature on ethnic wars. A rich tradition of considering ethnicity as a factor to conflict Ethnic Groups in Conflict in 1985, finding that positionality of ethnic groups within the state can lead to conflict and violence. Thus ethnicity in and of itself, does not cause war, as other scholars agree with (Fearon and Laitin 2003), but heterogeneous societies with pre existing discrimination s may devolve into civil war more easily than others. This is worth considering in terms of theoretical development s in the study of ethnic conflict. The study of ethni city in comparative politics focuses on explaining effects upon political life. Ethnicity is a set of cultural markers that denote common ancestry (kinship), cultural elements and shared historical past (H orowitz 1985: 41), although some only character (Ottoway 1999: 300). Culture includes things like language (Laitin 1992), religion, dress, and r egional affiliation. T wo schools of thought regarding the origins of ethnicity present themselves in the literature : primordialism/essentialism and social
72 (1973: 258) as fixed elementa (1973: 259) and preferences are homogeneous and fixed according to them. Even more crucial to the persistence of ethnicity over time in the literature is the matter of when and how it matters in political life. Instrumentalization of identity by political elites is in sizable part responsible for its persistence. Instrumentalization conceptually assumes by its nature that identities are constructed and fluid and aspects of identity c an be highlighted and downplayed for political gain. Socially constructed identity only becomes a dimension of cleavage when groups come into the sphere of other groups. Hence, instrumentalized ethnicity is constructed by the definition of the th 1969), maintains fluidity across boundaries (Gellner 1983) and is not inherently fixed ( Chandra 2001: 33 7 and Fearon and Laitin 2000: 848). The origins of ethnicity are by themselves somewhat uninteresting because of the lack of link age between the exi stence of ethnicity and conflict. However, in comparative politics and in political life the conflation of origins and persistence that continues to misidentify roots of conflict when discussing ethnically divided societies like the ones in the Great Lakes region. This conflation of definition leads one to question ethnicity as both a means to achieve a politi cal aim and as an end itself to create societies, nations, and states. There is a need to make clear the distinction between eth nicity and ethnic mobi lization Ethnic mobilization is the process by which these groups and their relational boundaries be come salient, sometimes causing violent conflict While mobilization has been used to explain nation building and the shoring up of political support aroun d the nascent state (Geertz 1963 and 1 973, Barth 1969, and Young 1973 ) and political structures, scholars have also explored when ethnic mobilization leads to
73 state disintegration and rebellion (Fearon and Laitin 2003, Lemarchand 1996 ) or state sponsored ethnically based violence (Kaufman 2001, Lemarchand 1996, Straus 2006). 1993 and in the Hutu power movement that caused total state breakdown in Rwanda in 1994. In b oth cases, calls to ethnic population bases by political elites led to forcible removal through murder and exile of another ethnic group from the state, encouraging rebellion and war. C ivil war in Burundi, Rwa nda, Uganda, and the eastern DRC suffers from t he problem of ethnic power relations. This essentially means that even if the stated composition of rebel groups are multi ethnic or non ethnic, there remains distinct desires to shift or change the ethn ic makeup of power to reflect fairer shares amongst e thnic groups. In the Great Lakes region many ethnic groups are divided amongst juridical states further compoun ding political issues Rwanda and Burundi are known variously as Ha in Tanzania, Banyamulenge in th e eastern DRC, an d Hima in southwestern Uganda Hence, there exists potential for cross border networks that could provide external support for rebellion among these groups. Some ha ve theorized about such networks and the ability for rebel groups to gain a nd keep power of their respective states due to outside support (Salehyan 2007, 2010, Wood 2010). Some groups in t he Great Lakes region behave this way, as evidence by support the Tutsi led RPF provided the Banyamulenge in the first and second Congo Wars ( Stearns 2011 Prunier 2009, Reyntjens 2009 ). I discuss ethnicity and ethnic mobilization here because many scholars, observers, and even participants in the Burundian Civil War mention the role ethnicity played in the decision to engage in
74 war 4 encouragin g civilian support, and ethnic voting in the transitional elections of 2005. Ethnicity became a way of identifying potential allies shoring up support, and marking enemies across the Great Lakes It is important to understand ethnicity here as not only a potential cause for the civil war, but also a driving factor in all political, social, and eco nomic life regardless of war. Thus ethnicity may be a cause of civil wars, but may also be a mechanism for peaceful post c onflict governance to be utilized by th e victor to shore up power amongst co ethnics. This brief discussion of ethnicity and its mechan isms in conflict solidifies that civil wa rs are multi dimensional and motivations of actors are only part of a larger process. The explanation of this debate al so provides valuable insight into how these motivations matter to post conflict outcomes. To restate grievance, shows a potential post conflict effect of these motivat ions. This is exemplified scope of positions previously unavailable to ethnic Tutsi, an important, if unvoiced, motivation in its war on the Habyarimana government. Similar ly, the CNDD FDD in Burundi also made ethnic changes in power following the end of the civil war. These examples show how motivations matter to post conflict outcomes, and I return to this point later in greater detail in later substantive chapters. In th e next section, I discuss the approaches to the study of civil war in the discipline, and the differences in these approaches and their applications. I do so to situate my theory within these traditions, and offer an ac count of the CNDD FDD that uses multi ple approaches to promote a more complete vision of the conduct of war and the behaviors of rebel organizations
75 Approaches to Civil Wars Above I made reference to the difference between causation at the micro and macro levels, between low level actors an d elites, and between individual motivations a nd global ones. The question remains as to how the scholarly community understand s how these causations wo rk in the practice of civil war. These are often tested in cross national level studies of global facto rs that make conflict likely, like Collier et al. and also in studies at the micro level, like what propels ordinary citizens and others without specific access to weapons or calls to arm to fight ( Humphreys and Weinstein 2008). In the last decade, schol ars have examined the differences between the macro level explanations of civil war and micro level processes of recruitment and inducement. By studying the micro level of conflict, scholars are able to further clari fy motivations, whether material or not, at the individual level. I mention this as an introduction to a brief review of the literature of causation at multiple levels within a ci vil war. T he evidence I present o n the CNDD FDD in Burundi shows both universally accepted causalities and behaviors and the motivations of individual actors by utilizing top down and bottom up data to provide a more holistic account of the civil war. Macro Level Approaches : These approaches to the study of civil war are the most widely embraced and oldest in the discip line. Traditionally these comprehensive theories were presented within case studies that provided a broad picture of the civi l war and society under study. They tended to focus on the characte ristics of the state (Scott 1977 ) or society (Gurr 1970 Hunting ton 1968) than drivers of conflict at the individual level, and searched for universal motivations among the population. These accounts often highlighted change in political models, at least in the initial study of
76 comparati ve politics in the 1960s and 197 0s. But macro level approaches to the study of war are not a thing of the past, and studies on ideology and war (Palacios 2006 ), religion and war (Stiansen 2004, Prunier 2005), economy and conflict (Collier et al. 2001, Sambanis and Eldawabi 2000, Fearon and Laitin 2003 ) illustrate contemporar y saliency Macro level accounts focus more on structures in and of conflict, not specific actors, and thus are more suited for providing general causality. A variety of methods are employed in macro accounts qualita tive accounts based on fieldwork as well as large N quantitative methods tested using datasets highlighting that macro approa ches do not follow a certain kind of methodological tool. I now discuss t he benefits and drawbacks to these approaches For one, mac ro perspectives on conflict wars into something understandable to non specialists. This is important to both the discipline as well as the policy community, for those who may not hav e time to invest in learning languages, country specific knowledge, and conduct fieldwork. But these benefits are n ot without negatives. M acro level approac hes to civil wars can hide variance amongst individual s and behaviors during civil war. This becomes especially clear when dealing with the aftermath of war, where potential combatants receive equal treatment, despite obvious differences in practice, in both the scholarly literature as well as in the punishment structure. By this I mean that individuals choose to participate in a civil war for a variety of specific motivations, like ideological connections to rebel movements, promise of financial gain, religious connections to the rebel leaders, and pressure from peers(Kilcullen 2009: ch. 1 and 2) 5 Whil e scholars can sort out a general set of motivations that many rebel participants ascribe to in the larger context of
77 the war, it yet remains that among individuals preferences vary. This extends to behaviors engaged during the war: for example, rebel part icipants in CNDD FDD described their varying levels of participation, from acting as information gatherers and early detectors of government army presence 6 to full fledged active combatants 7 The post conflict period is no different, with some rebel membe rs receiving material benefits as part of a reintegration program 8 and others receiving no material rewards and ostracization from former communities for their participation 9 In this thesis, I investigate whether or not there are o verarching themes to w hich interviewees ascribed causation of the civil war in Burundi. I did so in order to ascertain the way in which people think about and conduct rebellion and whether there were By searching for a glob al perspectiv e on the war, I determine what characteristics of the rebel group appealed to people most, and how that message was effectively spread (or in some cases, was not) among varied groups in the population. I highlight this more thoroughly i n Chapters 4 and 5 Micro foundational Studies of Civil War : Now I turn to the contrasting micro level studies of war, because this particular study utilizes both macro level accounts of the Burundian civil war as well as micro level interview data focusing on individual log ic of rebellion. By studying the micro level of conflic t, scholars clarify motivations at the individual level and are able to further parse out data on a smaller level These are often carried out by conducting massive survey collection projects or employ ing many semi structured interviews in the form of questionnaires to see variability Micro foundational studies have special applicability to the research I carried out in Burundi because they show variation amongst individuals and across temporal dimensi ons concerning
78 behaviors and perceptions of the war. These differences are crucial, because many Burundi specialists discovered global causes and uniform motivation to the war, and employ ed these in their analyses although not without caveats (Lemarchand 1970, 2009, Daley 2007, Krueger 2009). T he data presented in Chapters 4 and 5 provides additional information about the consistency of global assumptions for the war and extends arguments about the civil war to the post conflict period. Quantitative and Qualitative Methodological Tools in the Study of War In the study of civil wars, much has been made about the differences between quantitative, large N data based studies and more specific qualitative accounts of wars. Although each approach has its staun ches t proponents, the newest innovativ e work about civil wars uses qualitative accounts in conjunction with larger tests of theory across contexts, to show similarities and differences among civil wars (Kalyvas 2006, Weinstein 2007, Mampilly 2011). I give a brief introduction to these methodological positions in civil war and conflict studies now, in order to better situate my theory and evidence among previous work and preview the usage of both in this research I used both types of evidentiary support i n order to more completely satisfy questions about rebellion and post rebel governance in Burundi as well as across sub Saharan Africa. Qualitative Approaches : Qualitative approaches are those that use tools like case studies, participant observation and o ther forms of analyses to develop a rich contextual and causal narrative. Qualitative evidence is typically gathered by the researcher conducting fieldwork, living and working amongst the population under study, requiring back ground knowledge and potential new language skills. For many years, the case study method of qualitative analyses was de rigueur in the study of
79 comparative politics because it required the researcher to use a vast array of tools developed through detailed and intensi ve training in gr aduate schools to produce specific knowledge about particular events and places. These approaches understand particulariti es of human events and provide exact data, especially in understudied populations and peoples where other forms of data may be unreliable or unavailable. They lack in the abi lity to generate quantified, mathematically driven data across many cases. My work here comprises a large portio n of qualitative field studies conducted in four countrie s in the Great Lakes, providing deta iled political histories of active rebel mo vements there. Quantitative Approaches : Quantitative approaches are those that use mathematically oriented data a nd tool s to provide models show ing relationships between variables. Scholars utilize them by gathering data and then compile it into sets for further analyses. Sometimes researchers gather data by conducting surveys in the field: Other times, they are compiled by coding variables and events according to rigorously developed schemas. Quantitative analyses provid e incredible breadth of data when used to measure the impacts of characteristics of the conduct of war duration, number of troops, material support and the like. They are also beneficial for testing hypotheses and relationships among a large number of data points. Where these approaches are not sufficient is in targeting outliers and understanding contextual specificities that may not be apparent in large N regression equations. These approaches are complementary to the qualitative data I gathered because they illustrate relationships between variables and rebel group success on a larger scale, where I test
80 social and organizational power among sixty rebel gr oups across sub Saharan Africa from 1970 2009, confirming my qualitative work in the Great Lakes. The theory I present about how rebel groups become governments attempts to fill some of the gaps exposed by this review of the substantive and methodological literature of civil wars. It provides a new understanding of rebel to ruler transformations as multilayered processes, not discrete events, and using manifold methodological tools is necessary. Thus, the causes and behaviors of war at both at the elite an d foot soldier level are important factors that shape the outcome of post rebel gov ernance and stability. I present a capture of the timelines of civil wars in my theory in Figures 2 6 and 2 7 I use the s e diagram s to underscore a crucial part of this res earch that scholars should not view civil wars as separated by tempor al categories, but rather as linked events in a chain that follow and affect one another. Further, behaviors found in one temporal aspect continue to other temporal aspects of a civil wa r and post conflict time. Social and Organizational Power matter at all times of the war, but are highlighted at certain critical junctures to ensure success I examine this in greater detail later, but for now I provide a conceptualization of the temporal ly important variables in Figure 2 8 State Formation and Rebel Organization: The Linkages This section demonstrate s the historical processes of war making as state makin g to show the legacies of war as a form of dominance to produce a new state. The rebels in question in this research seek this, although it is not the creation of a new state, but rather the dismantlement of previous governments and the beginning of a new political order.
81 With the question of banditry and greed should arise the simila rities between the state scholars believe that a rebel group can operate more effectively when they are able to claim terr itory and extract rents from it (Mampilly 2011) I prov ided a discussion of the state in s ub Saharan Africa in Chapter 1 as a background to the larger discussion of the historical relationship between rebellion and process es of state formation. Scholars have long posited the European state development was a process of stationary banditry (Olson 1993: 567) wherein which rulers took to settling and taxing the population while providing security creatin g feudal structures that d eveloped into states. Mann and Tilly, among others, built theories of state formation based upon the logic of the Weberian state, that states sought a monopoly on the legitimate use of force and violence, so gained by first proving on the battleground the ability to organize violence. According to Tilly, European states were shaped by war making, extraction from the population to support war making ventures, and capital accumulation to fight future conflicts (Tilly 1985:172) The process finds familiarity i n the actions of modern day rebel groups in sub Saharan Africa, who, through conflict are able to extract resources from populations or banditry of state resources to further their capabilities. But rebels not only accumulate capital for the purposes of en larging the war chest modern day rebels also concern themselves with sustaining their enterprises in the post conflict period. For example, the CNDD FDD engaged in a campaign of fund raising for the purposes of raising an electoral war chest as early as 20 01 10 long before the actual elections and while violence was still ongoing. The purpose of these funds was clearly articulated to future constituents and receipts were even provided to citizens to legitimize the process. While
82 not all rebel groups engage i n stationary banditry or state formation in the Western European tradition, enough of these behaviors are conducted to make scholars pay attention to these kinds of transit ions and the potentiality for new states (Weinstein 2007, Reno 2000). This is import ant when considering how these potential goals of the rebel organization map on its behaviors and strategies during the various time periods of the civil war which I explore in the following section. T his section seeks an understanding of rebel behavior during the actual conduct o f the civil war In t his section, I identify literature on how rebels win the civil war and control of the state or barring that, the path to political victory. I n this litera ture review, I address the benefits and challenges to each type of victory either through military means or through negotiat ed settlement. Finally, I identify the key variables of rebel behavior under study here social and organizational power and link these to the process of rebel to ruler transformations. What aspects of rebel behavior are critical to conducti ng a war? The social sciences studied many aspects of these in regards to the sociological, political, economic, religious, and cultural aspec ts of civil war, as demonstrated earlier in this chapter. This review focus es these studies in hopes of identifying the primary and ancillary. In some sense all behaviors during this period are prim ary to the conduct of warfare and t his is not to be debate resolve to embark on the high risk strategy of seizing power through mobilization and violence. The venture has to be planned. Fighters have to be trained, the weapons obtained. Tactics have to be devise 2005: 4).
83 This section explicate s how this occurs, and the various types of victories that are possible for rebel groups. Winning through Military Victory : T he statement above suggests the seizing the capital and organs of power by force. This position find s root in historical understandings of war and state building, which see violence as a means of reinforcing power and hierarchal relationships that established dominion over lands and people, at least in so far as the process was under stood in the building of c ivilization in Western Europe. Scholars argue that historically this was not the case of state development in Africa (Thie s 2009, Herbst 2000, Boone 2003), and this posits some explanation as to the number of failed rebel takeovers borders were no longer in dispute post Independence (Thies 2009: 629) made it so that all challenges to sovereignty would be of an internal nature, thus drastically cutting the can governments to fight wars, while also cutting the incentives for rebel groups to wage war. That is not to say that rebe l groups in Africa are unsuccessful at winning military victories to refer to Figure Four provided in Chapter 1 seventy three percent of the rebel groups that came to power did so through military victory, and the so the National Resistance Army( NRA in Ethio pia, and the RPF in Rwanda claim ed power through total military victory (R eno 2011: 119: 128). The civil wars literature is somewhat divided as to what ty pe of victory is more common Fearon and Laitin (2008) state that military victory is the most common form of civil war termination and Walter (1997) pointed out that wars rarely
84 ended in settlement before in the 1940 1990 period, while the vast majority o f scholars that study post Cold War conflicts find negotiated settlement to be the most common (Mason et al. 2011, Johnston 2007). Using the UCDP datas et also yields results favoring more negotiated settlements than military victories to the termination o f conflict. Of 288 outcomes in civil wars worldwide from 1945 present, only 48 were ended in victory by the rebel army. Similarly, in the d ataset presented in Chapter 6 only 11 of 60 rebel groups won power by military victory. T here are far more immediat e benefits for the rebel group in question if they take power militarily rather than through negotiated settlement. For one, groups that win the war militarily have a clear mandate for post war agendas, as all other credible players are necessarily trounce d (Luttwak 1999, Duffy Toft 2010a, Duffy Toft 2010b:. 41) With a clear military victory, there can be no indecisiveness about who will be in power post conflict (Weinstein 2007, Duffy Toft 2010b:ch. 3) With decis ive victories, the rebel group can be see n as stronger victor in eyes of international community, and thus may be more likely to receive financial and poli tical support that can be used to further solidify power. In the long term the benefits are better still, as the historical record shows. The NRA, in power since 1986 in Uganda, has successfully consolidated its regime and grip on Ugandan political life (Kasfir 2005). Similarly, the EPLF in Ethiopia consolidated power for a successful solidification of its regime (Reno 2011: ch. 3). Both of th ese regimes have benefitted significantl y from being in power, in access to the international community and its financial and political resources. Internally and p erhaps more significantly, by winning the w ar militarily and these groups have been able to r ewrite
85 the history of their struggle and their countries. In no place is this more apparent than under the RPF government in Rwanda, in power sinc e 1994. The RPF have rewritten pre colonial relationships, the processes of colonizat ion, and Independence, wh olly changing the narrative of struggle and ethnicity in the tiny country that lost three quarters of its Tutsi population to the genocid e. The RPF narrative positions itself the dominant story of the genocide as exhibited and taught by the Gisozi Kigali M emorial Centre, the official national genocide memorial. This is an extremely powerful position for the RPF gover nment, as it allows it to manipulate domestic and international policies and age ndas in favor of its position of power and to write the defini tive national narrative for future generations (Lemarchand 2009). D espite benefits to winning the civil war militarily, there are some challenges, post conflict the rebel victor must face. To begin, winning a war militarily is incredibly costly in terms of time, money, and human capital. A lengthy civil war expends not only the lives of combatants, but also can result in negative civilian perceptions as the war drags on (Samii 2010: 12). A victorious rebel group must combat all of these losses and build l egitimacy amongst those who did not support them during conflict. A victor rebel group often suffers from the problems of asymmetric information in warfare, wherein the government may be far better informed, trained, and prepared. This becomes very apparen t in the post conflict period, where rebel governors face a lack of training in government practices and transitioning into new roles may be difficult. For example, much was made about the supposed lack of governing ability of the CNDD former rebel c ommander of the western region, Manasse Nzonbimpa, who was thought to by party officials to be a much better commander than governor 11 This led to a new
86 factioning in the post rebel CNDD FDD government, wherein which Nzonbimpa was expelled from the party and leadership roles while still maintaining immense support among the populace, a splitting to b e discussed further in Chapter 5 In short, the rebel faces an uphill battle in conflict and post conflict governance. Although a military victor may be viewed by the international community as the decisive victor, the nature of the ci vil war itself may increase its reputation as vicious, tyrannical, or bold beyond measure. This is especially problematic for future relationships b etween the rebel group and international actors, who hold the power to create embargoes and extend much needed loans to these new governments. The Conduct of War: The Contribution to Peacetime Success What behaviors do rebels exhibit during the civil war that shapes the outcomes of these wars? There are behaviors that some have specified lead to the winning of the war itself: strong leadership, tactical accomplishments, strategic battle decisions, and the maintenance of a territorial base. All these are i mportant to victory in the war, but also speak to a broader characteristic of a rebel organi zation, that of the ability to a ct quickly, coherently, and without internal conflict or dissent. These characteristics echo (1993:5). In practice, he defines this as:
87 people and resour While obvious differences between a state and a rebel organization preclude the foundations are still quite applicable, and furthermore, show the a bility of rebel organizations to act as proto states and their state building capabilities. I term this more generally the level of control and ability to project power expressed by the group that allows them to maintain order, giv e direction to followers, and accomplish goals This organizational power of the rebel group affects the soldiers actually follow commands given (obedience and loyalty), addressing the divisio n of labors and communications described in obedience can effectively plan strategy and tactics without dis sent, but also influence the relationships between civilians and combatants. They can do so by prov iding a united front, easy and timely communication and planning, as well as providing a strong N ot all rebel groups seek a positive relationship with the civil ians in their area of influe nce: some actively commit violence against civilians, whether randomly, like the Revolutionary United Front ( RUF ) in Sierra Leone, or at the behest of their leaders, like the RPF in Eastern Congo. These decisions may not be static over time: a rebel organi zation may initially pursue a strategy of cooperation with civilians and then later result to intimidation through violent means. Some groups even shift dramatically during the course of the civil war. Rebel groups must also provide management of civilian areas and protection of territory they
88 intend to hold. For example, the NRA provided both civilian protection, when possible, and organized local civilian leadership councils while fighting the Obote Government in the Ugandan Civil War from 1980 1986 (Kasf ir 2005: 275). Again, organizational power play s an important role in how these behaviors are expressed. Rebel Preferences : Do groups that take up arms against their governments have a preferred strategy for ending the civil war? This question is central to understanding the type of strategy pursued during the war. As illustrated, different ways of winning are more amenable to different strategies. In some ways, rebel preferences depend on the context of the civil war itself. If the war becomes protracted and a clear war, some scholars argue that the goals and preferences of the rebel group changed over time as the likelihood of a clean military victory diminished, as early as 1997 (Nindoera 2010: 4). The strategy of the group moved from military victory against the government army to a political victory through negotiated settlement and transitional elections. Once leaders chose to pursue this communications between the military and politica l commands reflected new tensions eventually leading to a split in the movement between the more war oriented members who believed military victory to be the only victory (under Commander Jean Bosco Ndayikengurukiye) and those who were willing to accept smaller political victories (under the uncle of Jean Bosco and original founder Leonard Nyangoma). On the other hand, other groups like the RPF would accept nothing but total military victory, even at the expense of Tutsi civilians whom they could have protected (Stam 2009). They did so because a total military victory would e nsure
89 absolute power post conflict and allow the RPF to provide for Tutsi interests within a potentially ethnically hostile social environment. Winning through Negotiated Settlement : As conflict changed in th e post World War II era, so did the ways in whi ch governments addressed it. As scholars have sharing arrangements between embattled incumbents and insurgents preferred instrument of peace Mehler 2005: 376). These agreements are often cheaper overall than total wars, reduce necessities of third party military interventions, and provide protection for human and material resources. Furt hermore, t hese agreements act as a foreign polic y strategy for world powers, and serve a s a domestic strategy to allow governments to maintain some power while providing con cessation for rebel demands and effectively (at leas t in theory) stop violence. Th ese agreements are not new form of discourse s, but rather find World Politics, which argued that pluralism was good for the health of democracy and could be used in a manner to miti gate conflict and promote governance Lemarchand notes the role consociationalism has played in the potential durability of the settlement of conflicts conflict strategy seemed to be the most l ikeliest durable post conflict country in the region (2006: 423 425). Benefits : While the benefits of winning power t hrough negotiated settlement seem less immediate and less apparent than those of military victories, they can be nonetheless enticing to a rebel movement seeking political power. In the immediate aftermath of a civil war, a negotiated settlement can be an attra ctive option because of guarantees embedded in the settlement. Often times, these settlements contain
90 provisions for re integration of combatants into civilian life or govern ment armed forces, seen by everyday rank and file soldiers as a form of protection from elite negotiators for international peacekeeping f orce, which reduces the burden of a rebel group having to provide security services to its civilian population, as well as ensuring (at least As Walter notes, one of the most signi ficant factors for rebel group disarmament is the fear or 27). Peacekeeping troops or an international for ce can allay these fears. Finally, negotiated peace agreements can end protracted conflicts that seem unbreakable, especially when the population and combatants suffer 12). weary populations are likely to p 29), even if one side must necessarily negotiate away some political power. Thus, rebel organizations are sometimes pressured by citizens and civil society figures to end the conflict in a peaceful manne r. In the long term, the benefits of negotiated settlements become somewha t cloudy, m uch of this due to the vagaries of settlements. Because of the relatively contemporary nature of these agreements, it is yet unclear if these have sunsets and can be guara nteed for longer term prospects (Roeder and Rothchild 2005:14) In some cases, peace agreements are to be supplanted by new agreements or constitutions, as in the case in the agreements in the wake of the 2007 post election violence in Kenya. In other case s, as in civil wars in Burundi and the DRC, the limits of the agreements are
91 unclear. For example, in Burundi transitional elections and even transitions to a Hutu leader (during the transition period) were pushed back several times because of FDD viewed as stalling for time for the Buyoya led (Tutsi) government. This vagary can open up potentially troubling political minefields, for now, opposition parties can use the strategy of n egotiation to sta ll politics and push for new agreement s or changes as in the case following the 2010 elections in Burundi. In Burundi, elections occur in a five round cycle, with the order consisting of Communal, Presidential, Senatorial, Parliamentary, and Colline level. In May 2010, allegations of fraud launched soon after reports of the CNDD FDD winning the communal elections. This led to a boycott of the Presidential and subsequent elections by major political parties, and a call by the major oppositi on coalition ADC Ikibiri ( Alliance pour Democratie et Change Ikibiri) to return to the drawing board and change the 2005 Constitution (the result of post conflict settlement). Some also argue that because of the very nature of give and take found in a neg otiated settlement, transitions are delayed and strong institution building is hindered (Duffy Toft 2010b: 40 41), thus problems like the one described above could be avoided by quicker consolidation of constitutional processes. Challenges: Although a pol itical victory through negotiate d settlement is no less as meaningful to a rebel group, challenges remain to building legitimacy and post conflict power and stabilit y with this type of win. First the victory is not as decisive as a military one, and usually results in a power sharing arrangement for the transitional and immediate post conflict governments. Rebel groups and governments may be unwilling to share power with former enemies (Duffy Toft 2010b:ch. 3, Walter 2002: ch. 1) A
92 degree of politica l skill and finesse is necessary in these types of governments, and many rebel groups are inexperienced in these type s of ventures. The problem also exists of trying to build political support and power in a co nsensus government with defined parameters tha t may work against the group to limit these powers (Licklider 1995) For example, the Burundian constitution delineates that the vice presidents must be composed of one Hutu and one Tutsi. This often leads to one vice president coming from the CNDD FDD and another from a majority Tutsi party, thus limiting the power of CNDD FDD in this role by constraining who can hold office. Finally, it must be noted that many scholars have found that a war that ends in negotiated settlement is more likely to return to c onflict (Fearon and Laitin 200 8, Walter 2004, Mason et al. 2011 ). Behaviors of Rebel Groups during Negotiated Settlements: The Keys to Success What behaviors of a rebel group lead to successful post conflict government and political power even when the group itself did not win power in the war outrigh t? A key factor contributing to rebel success in this case is the capacity of the group to use civilian support during the negot iations and in the transition and the scope of this support. Rebels then translate this support into political support as a party if and when elections come post conflict. becomes incredibly important at this time. The success of the rebel group during negotiations transform state structures to suit its needs. This can include things like co opting state leaders at the local levels, li ke the CNDD FDD did in Burundi as the war wound down in the early 2000s replacing or supplanting state structures, or rewriting constitutions and redesigning institutions. I provide a few examples here to illustrate how this supplanting
93 occurred. It did t his by creating its own ubashingatahe the tra ditional court of elders, to dole out justice and enforce the rebel law in rebel controlled areas 12 The CNDD FDD also co opted nyumba cumi and intang ible benefits alike to join the movement and spread the message amongst their households. These political strategies are part and parcel of a larger group of strategies employed during both the civil war and the transitional post conflict period to gain su pport for the rebel organization as it becomes a political movement. Th eorizing Rebellion Success : After the Conflict Ends This section focus es on the post conflict time period, to provide a two pronged approach that accounts for both historical and proxi mate factors that shape the post rebellion performance of rebel groups While in Chapter 1 I specified that success is defined as whether a rebel organization gains and maintains power post conflict, I take a moment here to more thoroughly define success here. Success in post conflict governance is more than just gaining power through an election or series of elections: it carries a degree of regime and state cohesiveness that allows the former rebel o rganization to effectively govern the state with all the rules and rights (authority, tax collection, monopoly of force) with no serious violent entrepreneurs that could spark additional rebellions. Success also entails the ability of the former rebel grou p to maintain enough internal cohesion so that the rebel group, despite some potential fragmentation, maintains enough consistent logic and group identity as to be the same group before, during, and after conflict. I illustrate this in later chapters throu gh the rise and maintenance of power of the CNDD FDD in Burundi. Based on this definition of success, I theorize that a rebel group does not
94 arrive at this without significant reliance of the behaviors, conditions, and actions of the rebel org anization during the civil war. These factors are part and parcel of the rebel organizations source and reliance on social power, and type of organizational power developed as a rebel group and continued onto the post conflict temporal period. I measure success in c ompar ative perspective in Chapter 6 by using conventionally accepted measures of state and success. Why these characteristics matter As I discussed above, the two characteristics of rebel groups that maintain importance even over the temporal dimensions of a civil w social power with civilians and the state. The preconditions of the state itself also play an important role in the abilities of the rebel organization, d iscussed further in Chapte r 5 But these characteristics capture broader dimensions of the study of warfare specifically, that organization speaks to a broader focus of scholars on contemporary and modern dimensions of warfa re, while social power sp eaks to historical models of the development of states and societies. The t ype of o rganization al p ower All rebel groups have some form of organization, but what do rebel scholars mean in the use of the term and how do they objectively measure levels of organization in th e rebel movem ent? Scholars define organization along a number of dimensions: sociological, political, and economic. My definition captures multiple dimensions, but most prominently views organization as a degree of internal cohesion, power, and congruency that allows t he movement to express unified political will and dominate others This centralization captures more than roles within the military stru cture and leadership itself, also capturing these relationships
95 as they move beyond these realms By organization, I re fer to the internal logic, cohesion, and order of the rebel organization I measure the type of organization in the substantive chapters by investigating the development of the rebel group in the battlefield, especially development of organizational netw orks and capacities from the field to the transition period to when the rebel movement gained political power. From these accounts, a coding sche ma develops that specifies how a rebel movement has organized itsel f and indicates changes over time 13 I also analyzed the level of organizational power quantitatively according to this schema for a robust measure in a larger test of the data explicated in Chapter 6 Organizational Power can be classified into ideal types falling in two categorie s: infrastructural and cellular These classifications capture directionality, projection, and cohesion of power within the rebel organization I choose to use the terms infrastructural and cellular because they accurately describe the two ends of a continuu m of the organizational power a rebel movement possesses. As explicated above, infrastructural organizational power denotes a group with a substantial internal frame, one in which orders are given and followed, communications are without problems, and tact ics, strategies and decisions are made at the head of the rebel organization and carried out, in hierarchal fashion, by those underneath. On the opposite end of the spectrum, a group with cellular organizational power exhibits very few of these characteris tics, and instead is limited in its range and capabilities in internal and external affairs. A few examples of what this look s like in practice illustrate these: A completely centralized (infrastructural) relationship would mean that all battle decisions c ome from one place of authority and are disseminated downward through the ranks, from high command to foot soldier. This
96 1986) and the 1994), where military orders were issued directly from one leader in each case (Yoweri Museveni and Paul Kagame, respectively.) This is in direct comparison with the decentralized style of command utilized by the CNDD FDD re able to issue parallel, sometimes conflicting orders, leading to a serious splitting of resources. The downstream effect of these is thus: a group that exhibits strong and unified infrastructural power is more likely to continue to do so even after a tr ansition from rebel to ruler, and can take these internal networks and divisions into the government, securing resources and power along the way. A group with cellular organizational power will have an uphill battle in its transition, as the very diffuse n ature of the rebel organization may prevent the significant holding of government power at the center of chapters of the continuum of organizational power. The Nature of Social Power: Inclusivity C apacity is usually a discrete terial ability to carry out war (Wood 2010) and scholars use capacity and capability interchangeably to signify the relative material strength of a rebel m ovement while conducting a war. These definitions often do not speak to conditions of the state or interactions with civilians, and thus are confined to the scope of the civil war itself. In my definition of social power is in a two pronged manner: one as approach to civilians and communities, and two, the rebel group pre existing state structures under reb el control. These comprise of the ability to use pre exis ting structures, or to transform them This
97 definition ca tches both the source and extent of the capacity of the rebel group as it applies to civilians and the state. The degree of social power refers to the ability of the rebel organization to extract from and create new state and civilian structures of suppo rt both as a rebel organization during conflict, and later, when they ha ve assumed power ability to conduct guerilla mobile campaigns with support and to the conduct of voter mobiliz ation campaigns and leveraging resources post conflict. I measure the source of social power in later chapters by employing post hoc 14 narratives of development of the rebel organization given to me by participants. I also relied on post conflict voting data, participant observation in the 2010 elect ions, and archival research of relevant political documentation, showing the outcome of the rebel rebellion period. I classify the nature of social power here as inclusive or exclusive. I do so in thinking of this type of power along a continuum, wherein which one side would be inclusive and the other, exclusive. Rebel organizations vary in their degrees of willingness or ability to interact with citizens and the state, and a continuum captures thi s variation. It is then translated into a typology by employing ideal types and categorizing rebel groups by which side of the spectrum they fall along most closely. The Typology In the tradition of other scholars in contemp orary rebel studies, I use the se variables to categorize the post rebe llion behavior of rebels in a ty pology. The typology contains four types of rebels as measured by the type of organization al power and nature of social power and the interactions of these two variables. I created a typology by listing descriptive qualities on an x and y axes of a matrix, creating a two by
98 two table (Collier et al. 2012:3, Elman 2005: 296 297). The interaction of these dependent variable the outcome of post rebellion politics. According to Elman, this type of model concepts (Elman 2005: 296) allowing for the incorporation of otherwise mes sy [it] is primarily a complement to deductive approaches, because filling in the cells requires working through the logical implications of the theory: given its posited causal relationships what particular outcomes are associated with different combinations of values of the theory's (ibid: 298) Thus, the outcome of the rebel group CNDD FDD in Burundi will be illuminating to filing in the outcomes in the rest of the matrix shown in Figure 2 9, because they illustrate a particular set of variable interactions that contrast with others. An ex planatory typology method proves most for comparing rebels in this case because it allows for the testing of several claims about rebel charac teristics and potential post rebellion outcomes in a way that can compound attributes across variables and show variations in these variables in in case comparisons as well as cross case contexts. The intervening variable present in both the source of soc ial power and the type of organizational power are the preconditions of the state, a variable disc ussed in detail in Chapter 3 Type A CNDD FDD These r ebel types exhibit cellular organizational relations and inclusive social power as a rebel group and as a government can be classified as a Type A rebel group They ga in political strength from its abilities to revamp state
99 structures and also benefit from popular s upport. However, it is unable to consolidate decisions under one hierarchy and frequently exp erience problems with discipline inside the group and leading to lack of cohesion The downstream implications for this are that strength of the rebel group over time may diminish, especially as internal splitting and divisiveness may weaken the base as s upporters leave following leaders. This could result in both a problem of suc cessation with the leadership structure of the rebel organization in addition to overall lose of strength and numbers. Type B AFDL II. These r ebel types exhibit cellular organi z ational power and exclusive social power as a rebel group and as a government can be classified as Type B rebels. They are the type of the four wit h the most to overcome, as its structure provides for no central direction and they are unable or unwilling t o rely on civilians or the ability to transform state structures for additional power or stability. These groups are likeliest to experience direct threats and challenges to their holds on political order and stability post conflict. Type C NRA These reb el types are characterized by infrastructural organizational power and inclusive social power as a rebel group that leads them to have the best of both worlds. They shape state structures and have the benefit of popular support, while also possessing a command structure that discourages dissent and encourages a solid face against the opposition. They draw upon these pre existing characteristics when they gain power to consolidat e and expand political gains and further stabilize the post conflict context.
100 Type D RPF These rebel types are characterized by infrastructural organizational power and exclusive social power as a rebel group w hich leads them to become internally strong but externally challenged governments. These groups are able to centralize decision making capabilities and exhibit strong leadership, but are unable or unwilling to rely on popular support as a mechanism for consolidating power. The governments that resul t from these rebel organizations can maintain political power but often face accusations of authoritarianism and despotism from the civilians they rule. Counter Examples: Failed Groups While not included in the typolo gy itself, failed rebel groups ( those that do not gain any sort of political power either through military victory or through negotiated settlements) provide another method of comparison by which to examine successful rebel to ruler transitions. An examination of these groups show s how differ ences in strategies and conduct of war, especially in organization and capacity, matter to the success of a rebel group, both before, during, and after conflict. Furthermore, comparing groups that operate within the same civil war and among the same popula tion, like the CNDD FD D and its rivals, the Palipehutu FNL ( Parti pour la libration d u peuple H utu Forces nationales de libration ), highlight these differences. These will be dis cussed further in Chapters 4 and 5 in conjunction with more detailed dis cussion of organization al power, social power and the data collected. Alternatives to the Theory Wh ile Chapters 3, 4, and 5 discuss the key variables of the theory presented here in greater detail, nonetheless I addr ess potential challenges to my theoriz ing here These alternatives can roughly be categorized into characteristics of the rebel group
101 and characteristics of the war and the general environment. It could be that successful rebel to ruler transitions are less a function of the characteristics an d type of organizational power and source of social power of the rebel group itself, and more closely related to other characteristics of the group. For example, it could be that the ideology of the movement matters to post conflict success. There are Marx ist and Socialist groups that have succeeded in transitioning from rebels to political power, like the Frente de Libertao de Moambique (FRELIMO) in Mozambique and the Frente Sandinista de Liberacin Naciona l ( Sandinistas) in Nicaragua. While ideology induce s followers to participate in conflict, as yet there are no accounts or evidence for how ideology affects post conflict success. Even the EPLF in Ethiopia, while nominally considered Marxist, is awarded far more credit in its successful rebel to rule r transition for its organization and discipline (Pool 2001:ch.1). Similarly, some ethnically based movements that have succeeded in gaining post conflict power like the RPF in Rwanda and the CNDD FDD in Burundi. Although these groups claim to be multi et hnic and do in fact contain members of multiple ethnic groups, eth nicity alone does not drive their success. For one, in Rwanda, the minority ethnic group gained power, not the majority, and thus civilian support is not a contributing factor. In Burundi, a lthough the CNDD FDD is a majority Hutu movement that came to power in a majority Hutu government after years of Tutsi rule, very few Hutu claim to support CNDD FDD purely because it is conflict period and the party does not o utwardly rely on ethnicity during election time 15 While ethnicity may induce some to join, it certainly cannot explain the participation of all, whether or not the war is ongoing. Finally, it could ed for longer, succeed better at
102 attaining post conflict power than others. This would be due to the practice and experience of the rebel organization that allows for decisive and cohesive decision making command structures, and logistical training earned over a long period of time, characteristics that serve them well when achieving power. While this argument cannot be easily dismissed, the same factors that support it also support an argument favoring rebel groups with prior military experience. Further more, groups may weaken, and not strengthen, over time, as shown by the gradual lack of popular support, demonstrated through protests, editorials, and citizen sentiment, especially by urban Ugandans, of the NRA in Uganda as the movement ages (Crisis Group 2012). The second set of rival explanations center around characteristics of the civil war from which the rebels emerge. Some argue that the way in which the war is won determines the shape of the post conflict government and political order more so than war political l andscape. Others may argue that the duration of the war affects post conflict outcomes conflict tran sition into peace depended on the fatigue of a decades long war and a t hirst for peace rather than other factors (Uvin 2009). While respondents spoke to me of this but we are tired of the war), combatants expressed that they were willing to continue the fighting should they be so asked 16 and civilians responded that while war conditions were disheartening, life in general in Burundi was generally the same regardless of war
103 or peace 17 Finally, much scholarly debate exists over the role o f ne ighbors, international peace keepers and the presence of outsiders as a factor that contributes to post war political order. Neighbors can provide sanctuary (Salehyan 2007, 2010), financial support (Schultz 2010), or even logistical and military support ( Stearns 2011, Prunier 2009, and Reyntjens 2009) to a rebel organization, crucial to attaining its political goals. For example, much scholarly discussion centers on the role of Rwanda in the overthrow of the Mobutu g overnment by the Alliance des Forces Dm ocratiques pour la Libration du Congo Zare (AFDL) Rwanda actively gave weapons, military and logistical support including troops, as well as financial support to a group that was formed in Kigali, Rwanda during a meeting of all potential rebel leaders active in the Eastern Congo, organized by the RPF (Stearns 2011: 23). Clearly, the role of neighbors in conflict can be understated, as they can be responsible for much of the early victory in conflict. But the support of neighbors does not explain how and why rebel groups can transition to post conflict power 18 as the support of the external nation can only feasibly continue for so long, and governance is often not the goal of the external government (as in the case of the RPF in Congo, whose interests lie more towards border security and financial prosperity). Other external actors also play important roles in civil wars. Much as the United States acted to restore order to Germany and Japan in the post World War II era, so have South Africa, the United Nat ions Department of Peace Keeping, and various other international governing agencies intervened to restore order and promote political development in Burundi, the DRC, and other post civil war settings around the world. United Nations Peacekeeping troops d eployed in MONUC, the United Nations Organization Mission in Democratic Republic of Congo (known by its French
104 name Mission de l'Organisation des Nations Unies en Rpublique dmocratique du Congo ) under chapter VII (active peacekeeping with the ability to use force) guidelines to enforce the 1999 Lusaka Agreement there. In other post conflict contexts in Africa, an African Union mission is often deployed as a peacekeeping force: recent examples While there may be some eff ect these international forces have on conflict, the far more likely international involvement is t hat of neglect, as in the case of the genocide in Rwanda in 1994, where the United Nations Assistance Mission to Rwanda pulled all but a skeletal force out i n the earliest days of the killings. Further, it would be incredibly difficult for any sort of outsi de group to significantly affect politics at this point in history when total war, like World War II, is a rare occurrence. Finally, as Rothchild and Roeder not diminish the underlying need for reliable institutional protections for vulnerable building activities must occur wi thin the country to stabilize a post conflict government. The final set of rival explanations center around the characteristics of the environment in which the re bels operate. I t may be that post conflict political power is more dependent on the nature of the state, the geography, or the pre existing political legacies of the country of the civil war. I find these arguments significant to my own. To projection of power over people and not territory, than the ways in which governments achieve stability will have less to do with war and more to do with civilian support (e.g., Weinstein 2007, Mampilly 2011, Metelis 2009) This dovetails my argument about a acity to interact with and draw from the civilian population. The prior
105 conditions of the state not only affect the way in which a war is conducted and the background barriers or conditions, but more importantly, affects how a rebel group is able to manage and conduct itself during and after conflict. The pre existing state may structure relationships as to make groups more reliant upon their own infrastructural power, or conversely, may do so in a way that prevents rebel organizations from fully developing this kind of power. Additionally, the prior institutions of the state matter a great deal as to how the rebel organization is able to depend and use civilian resources. feel more supported during the civil war, the source of social power available to a rebel group may be curtailed. Hence, prior institutions matter a great deal to this analysis, but are an intervening independent variable, that further elaborates the relat ionships between post conflict rebel success and organizational and social power of the rebel group itself. But even Herbst hi mself accepts that this was not the case in all countries in Africa (2000:183) and the Great Lakes, with its history of territori al accumulation through war and pre colonial expansion of kingdoms, may be an exception. While measuring the prior conditions of the state is not a directly quantifiable property (although Freedom House, Polity IV scores or the Failed States Index may be i lluminating), furt her elaboration in Chapter 3 shed more light on this concept. While qualitative accounts of causation of successful rebel to ruler transitions vary a nd cannot be fully examined in this research a preliminary test exists a way to quantit ativ ely test the power of these rival explanations, carried out here in Chapter 6 By proxying variables for these explanations and testing them amongst a population of rebel groups in Sub Saharan Africa, the power of such explanations can be examined
106 and compared with that of my hypotheses. While this may not fully provide explanation for all transitions, at the very least I can test my hy potheses vis vis these rival explanations to further develop a preliminary model. Conclusion This chapter briefly reviewed the literature theorizing rebellion over the various temporal dimensions of a civil war, from the pre conflict stage to the post co nflict one. In doing so, it highlighted the relationship between motivation and causation, conduct of the war, and o utcome of the war, further underlining the notion that civil wars and post c onflict stability are a process and not divided spaces. The chapter also presented a new theory of rebel behavior that carries into the post conflict period, a theory that focuses on the effects of the type of organizational power and source of social power of the rebel movement as they apply across temporal periods. I expand this theory in detailing the rise of various rebel groups in the following chapters.
107 Tables and Figures Fi gure 2 1. Actors in the Burundian Civil War, 1993 1994 Figure 2 2 Actors in the Burundian Civil War, 1998
108 Figure 2 3 Actors in the Burundian Civil War, 2002
109 Table 2 1. Aspects of Rebel Behavior and the Literature by Period Figure 2 4. A Tempor al Timeline of Civil Wars
110 Figure 2 5. Theoretical o utline of Civil Wars Figure 2 6. Temporally Important B ehavioral Variables during the s cope of a Civil War Table 2 2. A New Typology of Rebel Behavior
111 Notes 1 2 One notable exception is Bermeo 2003. 3 whether from material or intangible goods 4 Personal Interview, March 21, 2011 5 Personal Interviews, March 2, 2011, January 21, 2011 6 Personal Interview, February 23, 2011 7 Personal Interview, April 15, 2011 8 Personal Interview March 18, 2011 9 Personal Interview, April 4, 2011 10 Personal Interview January 21, 2011 11 Intervie w, Bujumbura, November 21, 2010 12 Personal Interview February 21, 2011 13 To not be immediately discussed here, but will be important later.
112 14 I recognize that using post ho c stat ements is somewhat problematic. I attempted to also use statements made during the conflict when possible, and looked to in country sources of media for this, 15 Personal Interviews, March 21, 2011 and March 25, 2011 16 Personal Interview, November 12, 2010 17 Personal Interview, December 3, 2010 18 Although this may be a factor in the continuing decline of the AFDL in Congo following the removal of RPF support after 2006.
113 CHAPTER 3 ANTECEDENT STATE CONDITIONS, ORGANIZATIONAL AND SOCIAL POWER Introduction The state in Africa, described briefly in Chapter 1 is thought to shape all social and political order, a momentous task when considering fragmented ethnic groups, slow economic growth, and the persistence of the power of personalities and patronage within African societies. Even when taxed in this way, however, the state in Africa provides structure and coherence in ways far beyond those of citizen and subject, ext ending to non state actors and groups, like the rebel organizations this work studies. The chapter lays out how antecedent state conditions affect and interact with the two variables under study, the type of organizational power and the nature of social po wer. In doing so, it brings to bear a study of the African state and how pre colonial, colonial, post region, tailoring the patterns and responses of rebel organizations ther e. As one scholar 259) and thus, in this historical delving an understanding comes of the compa rison between these histories that shaped rebel organizations that sprang to power in the twentieth century. This lays the foundation for explaining how the preconditions of the state act as an intervening variable upon the two independent variables, the t ype of organizational power and the nature of social power to rebel to ruler dynamics. If, as tate formation
114 and many analysts views civil conflict as competitive state how the state emerged and developed proves useful to analyzing rebel organization practices (2012: 246). The chapter focuses in particular on two impor tant Burundian pre colonial and colonial factors ethnic structures and the degree of state centralization or coherence, and two post Independence factors: political instability and the late 1980s transitions as a result of international restructuring of a id. These factors directly contributed to the rise of rebel movements in Burundi and find similar parallels in the histories of the rebel organizations discussed in the work in Rwanda (the Rwandan Patriotic Front or RPF), Uganda (the National Resistance Ar my or NRA), and the Democratic Republic of Congo 1 ( Alliance des Forces Dmocratiques pour la Libration du Congo Zare or AFDL). In this chapter, I discuss the histories of the Great Lakes to elicit an understanding of how these histories shaped their res pective rebel rulers and the differences in state conditions that shape post conflict governance. I then examine how the relationship between the prior conditions and structures of the state and the development of organizational power within the various re bel organizations under discussion. This leads to the impact that these have on the development of the sources of social power the rebel organizations depend on Finally, I explicate the relationship between pre existing state factors and the variables und er study, and why preconditions are an intervening but not wholly separate actor upon the source of social power and type of organizational power the rebel organization exhibits. I provide the history of Burundi and its pre colonial, colonial, and post colonial developments to show the nature of the African state and its impact there. Referring back to Chapter 1 and the brief discussion of the state as a palpable environment for
115 rebellion and power seizures, this history shows how the Burundian state eme rged as a strong regulator of social and political behavior there, especially in, as Newbury suggests above, cementing the politics of difference. These historical factors were used by the Conseil National Pour la Dfense de la Dmocratie Forces pour la D fense de la Dmocratie (CNDD FDD) to build their rebel movement internally and legitimate support from civilians. They were also utilized by its competing rebel movement, the Part i pour la libration du peuple H utu Forces nationales de libration ( Palipehutu FNL ) I also provide accounts of other rebel movements: the RPF in Rwanda, the NRA in Uganda, and the AFDL in the Democratic Republic of Congo, to further illustrate how history shapes outcomes. A Short History of Burundi Pre Colonial Kingdom The kingdom of Burundi prior to colonialism was divided in a hierarchal form of monarchy under a Mwami (king), with labels of Hutu and Tutsi more suggestive of class than actual cultural or racial differences and steeped in historical lore more so than a ctual conditions. Scholars generally accept that Hutu indicates a longer historical association with the land (since the 6 th century) and Tutsi indicates a migrant one, with waves of presumably Nilotic Cushtic immigrants arriving in the region about the 14 th century (Newbury, D. 2001, Watt 2008, Chretien 2003: ch. 1). Early regional explorers admittedly only hypothesized about the notion of a pre colonial migration (Newbury, D 2001: 273), and the existence of flexible oral traditions (Lemarchand 1996: ch. 1 ) that manipulated history make the veracity of such claims suspect. In any case, Burundians themselves tend to accept the pre colonial existence of classes and labels, which later defined modern ethnic conflicts. The pre colo nial
116 kingdom bore the addition Hutu nor Tutsi, but above both as an extension of the royal family. Governance in Burundi was oligarchical in nature with four clans the Batare, the Bezi, the Bambutsa and the Bataga all Ganw a, competing for power and sharing cycles of kingship (Eller 1998:209). The king was also prevented from gaining authoritarian control as in Rwanda because the princes engaged in local conflicts amongst themselves and their supporters, and thus no one king was able to rule with an iron fist completely. The system also had a feature of power sharing among the four clans, with rotating cycles. This consociational but fragmented system led to structural instability and weakness, as power was easily captured an d transferred between various Ganwa leaders. These relationships also created a more decentralized system of governance than existed in similar pre colonial kingdoms. Power structures were parallel at some levels, and centralization, although static, was m uch less defined than in the Buganda or Rwandan kingdoms (Chretien 2003: 71 79). This system had a positive effect on ethnic relations, however: because power was so insecure and fluid, the higher social classes (Tutsi and Ganwa) adopted a more generous and less contentious stance towards the peasant class (Lemarchand 1970:24) and thus ethnic relations were less structured than in Rwanda (Eller 1998: they did exist before colonialism, were intra ethnic in nature rather than interethnic (Newbury, D. 1998: 76) an d clan, social, and regional cleavages provided more directions for conflict than simply between Hutu and Tutsi.
117 Consolidation and the Pre Colonial State pre colonial history is somewhat anomalous on the continent because it is one of the few places where pre colonial state development mirrored that of development in Western Europe. By this I mean that consolidation amongst the kingdom occurred through conq uest by small local wars and centralization, uniting various clans that previo usly existed side by side. Scholars generally accept that the kingdom form ing the basis for modern Burundi was established around 1680 by Mwami Ntare Rushatsi of the Batare clan, and generally remained the same size until an expansion project to bring in neighbors and revenue began in the 19 th century, stabilizing to current map dimensions around 1850 (Watt 2008: 23). Mwami Ntare Rugamba took part s of the conquered Bugesera vassa l dom a minor entity between the Rwandan and Burundian kingdoms, enlarging and fusing the state (Watt 2008: 26 27) although borders between the two kingdoms were respected with little overt threats to sovereignty. Tensions between the Rwandan pre colonia l state and the Burundian one ran rather low overall because of the presence of local conflicts over projections of power between various princes of the Burundian royal family (Watt 2008: 27) and the cohesion of the kingdoms internally (Newbury, D. 2001: 2 59). By 1850, Mwezi Gisabo, the youngest son of Ntare, gained power, sparking a battle between elder (Batare) and younger sons (Bezi) (all members of the Ganwa Batare struggle did no t rule out subsidiary conflicts within the groups; each princely claimant sought to conflict for the kingship dominated power struggles, internal struggles also remained,
118 did turn towards Hutu Burundian oral tradition, lore of anti Roi(King) ran rampant at this time, with characters like Kilima, an d the powerful, not just the arrival, the kingdom was not split among Hutu Tutsi lines, but rather among peasant Ganwa] as chiefs and princes, rather than against any specific socio circumstances allowed for colonialism to fundamentally alter the social context of power and reshape relationships along different ethnic dimensi ons suited to the style of rule imposed after viewing Burundian society from the outside. It did so by allowing the colonizers to determine which social classes would be utilized in governance and alliances, and thus could create ethnic conflict by the inc lusion or exclusion of Hutu and Tutsi in these alliances. Colonialism the Berlin Conference in 1884 continent amongst themselves for colonialism and exploration. Early exploration of the region by missionaries and adventuresome types began in earnest by 1890. In 1897 the colony of Ruanda Urundi was founded by the German government, who administered the territory, which collapsed the tw o kingdoms into one administrative unit, through indirect rule (through the Mwami) until losing all of their colonial territory following their
11 9 defeat in World War I. Possession of the joint colony was given in the form of a trusteeship to the government o f Belgium after the war in 1921, which instituted policies that eventually played a large role in shaping subsequent episodes of political violence. The Belgian colonizers crystallized ethnicity by instituting identity cards in the late 1920s. Conflict dur ing this period of colonialism became less intra ethnic and more macro (and inter of the workers, and greater visibility of class differentiation, all combined to intensify ethnic con Ganwa and Tutsi cla ss to maintain order, giving the dual effect of amalgamating the Ganwa and Tutsi into one s uperior ethnic group, while serving the purpose of intensifying differen ces in status across al l ethnic groups (Lemarchand 1996 :10). This solidified conflicts already emerging at the threshold of colonialism described above, and instilled the beginnings of power based on Tutsi hegemony. Although the Belgians were able to chang e some dimensions of identity during the early colonial period Burundians still primarily referred to themselves by clan identification, a social marker that allowed them to build relationships, find friendly faces while traveling, and provide marriage ch oice (Chretien 2003: 88). By the late colonial period, though, Belgian made identities and policies became more d ominant markers and carried significant political and social weight. By 1933, about two thirds of the chiefdoms were consolidated, from 133 to 46, and finall y to 35 in 1945 (Lemarchand 1996 :43), with none of them Hutu (Gahama 1983: 104) Independence Although the League of Nations mandates over colonial possessions transferred to the United Nations, little changed in the ways of life for the
120 A political sector continued, with the majority of leaders coming from the Bataga and Bambutsa clans (a historical shift from the Batare and Bezi leaders of the pre col onial chiefs (Chretien 2003: 271). T he newly established United Nations post war environment, and emerging liberation movements across the world combined to make an es pecially palpable environment for colonies and colonial powers alike Movements towards independence across sub Saharan Africa created a dom ino effect, but Burundi an claims for political independence did not begin with gusto until the late 1950s. Surrounded by such powerful figures in African politics as Patrice Lumumba of Congo (most famous for heartening Congolese liberation with Soviet assi stance and murdered under mysterio us circumstances with CIA involvement) and encouraged by contrived with Belgium to gain independence with traditional rule of the monar chy still in place. In November of 1959 the Belgium government agreed to political reform of the local governments, with the chefferies (chiefdoms) to become provinces and the sous chefferies to become communes with appointed burgomasters and elected commu nal councils. A Counseil du Pays would also be elected from the councils and given legislative powers, with the Mwami to retain powers as a constitutional monarch (ibid: 51). This was largely due to the disastrous decolonization experience in neighboring R nd
121 conflict between the Belgian powers the Catholic Church, and inter ethnically between the new Rwandan citizens. The newly emerging parties came into existence before this declaration, and relied extensively on networks of Ganwa leaders and supporters to put into place the political support needed to win upcoming elections. The first political party, UPRONA ( Unite et Progres National ), was formally established by the son of Mwami Mwambutsa, Prince Louis Rwagasore, in 1959 and formally registered in January of 1960 (Daniels 1992: xxi). This party aimed to reflect the democratic ideals put forth by neighbor Julius Nyerere (the Tanzanian nationalist and leader of the independenc e movement there), lessen ethnic cleavages, and provide a plan for Burundi Prince Rwagasore was almost uniformly admired for his desires to democratize politics in Burundi, his refusal to play ethnic politics and his reputation as the favored s on of the Mwami. However, that did not help UPRONA to prevail in the first communal elections in 1960, which were won by the PDC ( Parti Democratique Chretien, the leading competitor and especially popular amongst Bujumbura city Tutsi and the Belgian coloni al authorities). A string of ar rests, including Rwagasore, carried out by the Belgian colonial administration to el iminate competition for the PDC ensured the vision of Prince Rwagasore and the UPRONA as nationalist leaders and patriots. This ensured that legislative elections in 1961 were easily won by minister. Mere months later, the presumptive prime minister was assassinated by a Greek agent of the rival PDC on October 13, 196 1, with support again coming from the Belgian colonial administration who felt UPRONA was too communist leaning and antithetical to Belgian post colonial desires This act set the tone for future politics in
122 Burundi, making assassination a formal accepted expression of political discont ent, a claim Burundians would often make over post Independence period. Belgium hoped to avoid the bloodshed and chaos of the transition in Rwanda in Bu rundi, but their actions supporting the PDC and these events that transp ired under their direction destroyed that hope The loss of the dynamic leader destabilized the UPRONA party, stifling further expansions of the democratic process and resulting in the overthrow of the monarchy and emergence of a military state Independen ce was finally granted by a reluctant Belgium on July 1, 1962, but from 1961 until the takeover by the army in 1966, the political sphere in Burundi rem ained mired in chaos, bleeding over into the public sphere and causing ethnic massacres and political as sassinations. The former ethnically united UPRONA party divided in to two factions the Monrovia (Hutu) and the Casablanca (Tutsi) both named after supposed international factions and philosophies they adhered to These factions jointly divided Parliament and all prime ministers from Independence until the appointment of Pierre Ngendandumwe, the former ally and friend of Rwagasore and a Hutu member of the Monrovia faction, in early 1965. Tragically, the same day of his announcement of cabinet positions and less than a week after taking office, Ngendandumwe was assassinated by a Rwandan Tutsi refugee on January 15, 1965 (Watt 2008: 31). Modern Burundians now memorialize the great Hros de L'indpendance in media and culture, and politicians use them for cal ls of ethnic unity and the strength of Burundi without foreign intervention. Again, assassination was seen as a viable political option for all players whether Hutu or Tutsi. Hutu led coup attempts against the monarchy in May and October 1965 culminated in the death of thousands of Hutu and Tutsi civilians (in response to
123 government and exile attacks), as well as political assassination s of Hutu leadership in UPRONA, including another charismatic leader of the Independence movement, former Rwagasore conf idant and Hutu leader Paul Mirerekano. early post Independence histo ry reflected the position of scholars of the country that violence became the default mechanism to resolving conflict, ethnic or otherwise (Watt 2008: 32) Even though ethnic co distinct ethnicity [had] been a facet of life in Burundi and Rwanda, both at the level of etimes follow these patterns, scholars generally acknowledge violence as a political tool (Lemarchand 1996: ch. 2). S tate repression in the form of a military takeover of government eventually quelled the violence, led by Michel Micombero of the Burundian Armed Forces, which had gradually removed all but token Hutu presence. Until July 1966, Micombero served as a captain in the Burundian National Army, after which he became State Secretary of Defense as part of a new UPRONA government (Chretien 2003: 314). Four months later, he overthrew the Mwami Ntare V (who himself had just deposed his father Mwami Mwambutsa) and proclaimed Burundi a republic, with himself as President and now Colonel in the Armed Forces, and anti Hutu political reprisals emerged as Tutsi hegemony of government and social life there coalesced. A Political History since the Founding of the Republic From 1966 until the first democratic elections in 1993, Burundi remained under military rule despite successive coups. Of the three leaders d u ring this time period, all were military commanders of Tutsi Hima lineage from the commune of Rutovu in the
124 southern province of Bururi Micombero, Bagaza, and Buyoya all sought to keep this hierarchy of Bururi and Tutsi supremacy in both the armed forces and the civil government. This system of governance instituted de facto ethno regional segregation and discrimination at all levels, including civil service, education, and the armed forces. The Regime of Micombero. Mic ombero, a Tutsi of the Hima substructure fro m Bururi, became president of Burundi in 1966. Under his reign, politics became very tense and ethnic repression was high, especially in the army and government, but otherwise free from major acts of ethnic violence until the watershed even ts of 1972, an episode described by the pre eminent Great Lakes scholar Rene Lemarchand as a What begin with a n attempted overthrow of the government by Hutu army members in 1969 culminated in the killing of approximately one hundred and fifty thousand Hutu intellectuals, school boys, skilled workers and any other perceived threat to the hegemonic control of political power in the hands of the Tutsi Hima elite three years later (Weinstein 1976). The minor revolt of 1969 led to other m inor revolts and signs of political strain, including a plot in 1971 by excluded Tutsi from Muramvya (the traditional seat of power under the Bezi Batare dynasty). The defining revolt, however, occurred on the 29 th ols of Ugandan government of Idi Amin forcibly handed over the deposed Mwami, Ntare V, back to the Burundian government, where he was almost immediately executed. Amon g the victims of the violence were many fathers, uncles, and brothers of those who would become the CNDD Eustache Ngabisha (ibid: 37). Over 300,000 Hutus fled immediately into exile in
125 Western Tanzania, creating a cauldron of political tension and grudges against the Tutsi government. The events barely even registered internationally, and no interventionary force stepped in to stop the bloodletting. The Micombero government explained in the events that it sought the elimination of the threat to the Tutsi ethnic group. The document described in detail the events of 1972 and the logic behind the government response. The genocide and removal of any potential Hutu political challengers shored up the pol itical power of Tutsis and the Bururi clan in specific The Regime of Bagaza General Micombero was deposed in a bloodless coup while out of the country in August 1976 by his cousi n and fellow Rutovu bred Lt Colone l Jean Baptiste Bagaza. Bagaza rode under the banner of the elimination of cor ruption and a return to order and prosperity. Bagaza deemed any mention of et hnicity illegal (Lemarchand 1996: 10) as way of disguising the growing Tutsi and Bururi centric repres sio n and discrimination, which flourished as the military increas ed its grip on the state. This wa s remarkably similar to post genocide Rwanda under the R wandan P atriotic F ront which has repressed both Hutu and Tutsi genocide survivors alike. ime also sought the elimination of the power of the churches over Burundian politics, banning newspapers and radio stations from Catholic and Pro testant alike (Watt 2008: 40). making Kirundi the on ly language of instruction in schools (eliminating future job opportunities for Hutu who could not afford private schooling or tutoring in French, still the language of government and commerce), villagization 2 among the peasant classes, and the restoration of oral tradition as a form of historical
126 discourse (Lemarchand 1996: 107). Under the Second Republic the new military junta (The Supreme Military Council, of which only 3 of 30 members were Hutu) wrote a new const itution and elected a new parliamen t in 1981, bu t much like the regime before rumors of corruption and anti democratic sentiment quickly surfaced. Peasants were leery of the forced villagization programs and suspected the government attempts were rooted m ore in control rather than promoting self sufficiency (Kay 1987: 8). Tracts change in the distribution of wealth, resources and reduction of corruption (Lemarchand 1996 : 115). The Regime of Buyoya In September 1987, Major Pierre Buyoya seized power and suspended the constitution (Lemarchand 1996: 118) under the guise of the necessity of regime change while Bagaza attended the Francophonie summit in Quebec, Canada. Buy oya, also a Tutsi from the Hima clan and of Rutovu, Bururi, established a ruling council similar to ones established under other regimes called the Military Committee of National Salvation, which dismissed the constitution and dissolved the Parliament (ibi d: 117). Many relationships that suffered under the Second Republic were now restored, including those between the church and state. Buyoya even late 1987 (ibid: 119 ). But 1988 ushered in waves of political and social repression, including the expulsion of school children based on ethnicity and intra regional tensions. These events were brought on by tracts distributed in rural areas depicting the Hutu hopes of opening political space, despite the increasingly mixed signals given by the Third
127 Republic. Acute ethnic tensions rose, and the harassment of Hutu communities c ontinued. Then the events of Ntega and Marangara, coupled with the assassinations of The Palipehutu FNL is long thought to be behind the massacre of Tutsi civilians in Nt ega and Marangara, two communes in Ngozi and Kirundo provinces respectively, which previously experienced little ethnic violence, although were the site of the previously described tracts. Many external and internal factors contributed to the events, inclu ding a fall in coffee prices in a region where coffee cultivation and smuggling is common, the proximity to the Rwandan border and memories of the 1959 revolution, and poor infrastructure within the provinces themselves (Lemarchand 1996: 122 123). As witn ess statements provided, the events began when Tutsi communal leaders in both communes used force to arrest so called local self defense groups of Hutu. In Marangara, some of these groups had burned bridges and barricades erected after the gendarmie came o at night plotting Tutsi destruction (ibid: 124). The violence shifted to Ntega around the 14 th of August, where similar events unfolded. Tutsi civilians became the targets of the roving b ands of disgruntled Hutu self defense groups, with up to 300 murdered (ibid: 125). These actions unleashed a wave of repression and massacre by the army against Hutu civilians, with deaths between five and twenty thousand (Horowitz 2002: 41). Violence cont inued to be the dominant mode of political discourse throughout the third wave of democratization in Burundi, and the newly minted National Commission to Study the Question of Unity, put forth by Buyoya in 1989 as an answer to critics, would surely fail in its long term goals As the Third Republic and Buyoya demonstrated
128 before, although willing to engage in politicking about ethnicity, they were more than willing to continue repression and ethnic violence despite these platitudes and promises. Because of pressure from the international community to open political space and restructure politics in order to receive donor funds, the Buyoya regime moved to create a new government, and the charter of unity was drafted by the ethnically mixed commission and rati fied by referendum in 1991. A new constitution was established in 1992, introducing multi par ty democracy (Chretien 2003: 318 320). This repression also bred succe ss in the form of a new type of politician: Melchior Ndada ye, considered central in the stru ggle for democracy and ethnic harmony in the early 1990s and a hero and martyr to Burundi ans, Hutu and Tutsi alike Ndadaye emerged as the leader and presidential candidate of the FRODEBU ( Front pour la Dmocratie au Burund i ) political party in the early 1 990s, writing political expositions about the nature of justice and democracy, explicitly multiethnic in character (Lemarchand 1996: epilogue). Like many other future politicians in Burundi, Ndadaye fled the country as a result of the 1972 genocide and fi nished his education in education and banking at the National University in neighboring Rwanda. While in exile, he co founded the Mouvement des Etudiants P rogressistes Burundi (Bampere), chairing the organization until the merger of a number of organizations led to the creation of the Labo r Party of Burundi (UBU) in 1979 (Reyntjens 1993: 30). He returned to Burundi in 1983, working the banking sector there, and secretly organization FRODEBU around 1986 under President Ba gaza. FRODEBU was organized explicitly as a political organization, and not a rebel one, in direct contrast to the other active Hutu organization, the Palipehutu (Palipehutu FNL).
129 st the short lived Ndadaye government is a fairly common phenomenon among emerging democracies, especially at the time in Central Africa (Mansfield and Snyder 2005: 5 6). Some literatures suggests that democratization and concurrent economic liberalization causes civil wars (Tirone and Savun 2011, Mann 1999, 2005, Mansfield and Snyder 2007, 2009, Lake and Rothchild 1996) in that instilling (and in some cases, the international community demanding) such processes without strong internal institution building creates weak regimes and can facilitate nationalist violence, which can erupt in ethnic conflict (Snyder 2000). In Burundi, a lack of such institution building directly contributed to the assassination of Ndadaye by an unchecked military regime, which then steamrolled the Parliament, National Assembly, and subsequent government leadership to take power for them and return Buyoya to the Presidency in the early years of the civil war. Thus, the pre existing centralization structure of the military was able to survive democratic change while institutions were not 3 This is an important characteristic of the Burundian state, especially in light of how rebel organizational development mimicked this p attern discussed in Chapter 4 The Election and Promise of Ndad aye : a New Regime With backdrop of increasing ethnic violence and tension international donors, mostly Belgium, France and the United States, began to push for new political space an d new actors in concordance with liberalization policies more globally The first democratic elections were to be held in June of 1993, in accordance with principles laid out in the new constitu tion established in 1992 Ndadaye was seen as a fair, earnest, and wise leader, who immediately installed a technocratic gov ernment with Tutsi well represented
130 (about a third) a month after winning the elections. But his regime suffered from almost immediate attempts by Tutsi military officers to regain power: the first such attempt occurred on July 2, 1993, when forty troops a Second Commando Battalion tried to seize power (Lemarchand 1996: 178). This early unwillingness to accept civilian or Hutu leadership. T he October 1993 successful coup was the event that started the rebellion and civil war and dramatically shifted the dialogue of political violence in Burundi into full scale war. Now, I move to a discussion of the history of ethnic violence in the post In dependence period in Burundi. While politics and violence there are uniquely intertwined, the timeline of political change does not always follow the timeline of ethnic violence, such as the 1972 genocide occurring in the middle of an otherwise stable Mico mbero regime. Thus, an examination of these phenomena separately allows for greater understanding of the causes and consequences of both political action and more directly ethnic ones. I also discuss ethnic violence to illustrate how the state previously u nderstood and dispatched this type of violence against its civilians, showing the preconditions of the Burundian state that led to the formation of rebel organizations there. Violence and Ethnic Conflict: Pre Cursors to Rebellion Ethnic tensions were nev er far from the surface of these regimes and erupted periodically into mass violence. Scholars generally agree that there have been four periods of civil war since Independence 1965 66, 1972, 1988 and 1993 2005(N garuko and Nkurunziza 2000:372); although wi thin these periods there have been various other
131 types of political violence. Ethnic massacres without war occurred in the early 1960s and 1991 1993 and political assassinations took place throughout post Independence history. Independence and Violence independence was n o less tumultuous, and the two influenced each other reflected in administration had designs on the two colonies remaining a joint enterprise following Independence; but ethnic relations in Rwanda were far too tense for that goal to be realizable The Hutu Revolution in 1959 sparked three years of interethnic violence there, culminati ng in the deaths of tens of thousands of Rwandan Tutsis, thousands of Rwandan Hutus, and the expulsion of several hundred thousand Rwandan Tutsi to neighboring countries, including Burundi. These incidents understandably worried Burundian Hutu and Tutsi al ike, and scholars have noted the degree to which politics were influenced by ethnicity in the first days of Independence (Chretien 2003: 311 313, Lemarchand 1996: ch.3). While ethnic massacres perpetrated by the government of Burun di had occurred historically before the genocide of 1972, none were as violent or as centrally planned and organized as the one that occurred then. This wide scale attack, accompanied by political repression on Hutu peasants by the Tutsi dominated army cau sed serious problems for the region at large. Massive influxes of refugees from Burundi into neighboring Tanzania and Rwanda created long term destabilization (in Rwanda) and
132 raised major questions as to the care, maintenance, and potential political right s of created in Tanzanian exile that served to further separate and define ethnicities in terms of victim and oppressor(1995:20). This event was also very personal to a great number of future Hutu militants and politicians, and the extent to which it mattered to the founding of the CNDD FDD and its philosophy was huge. In the early days of the and soldiers had lost family members to the genocide (Nonyzima 2004: 368 ). The Interim Period, 1972 1988 Violence in Burundi tended to be ethnically driven and episod ic in nature, history, it was not an everyday occurrence and for long stretches of time ethnic violence ceased. The period after the 1972 genocide until the events of Ntega and Marangara can be described as such a period. While political changes occurred, including the overthrowing of Micombero by Bagaza in 1976, they did so with a minimum of bloodshed, and because the change of power was intra ethnic in nature, propensities towards ethnic conflict diminished. Furthermore, demographic shifts made control of civilians easier: between 200,000 and 300,000 Hutus were driven into exile in Rwanda, Tanzania, and the Democratic Republic of the Congo as a result of the genocide, and mo st of the political and social leadership, intellectual leaders, and potential challengers were murdered, leaving a internal population with no choice other than to support Tutsi politicians and hope for lessened discrimination. The one potential violent e ntrepreneur
133 was the Palipehutu FNL, created in exile at the Tabara refugee camp in western acted on an aid committee at the camp (Watt 2008: 85). He wrote political tracts and boo ks describing the ideology and pre Burundian state, as one of invasion and conquest by the Tutsi of the Hutu population. last Lead Up to Civil War 1988 1992 The period of time beginning with August of 1988 ushered in n ew political violence, repression, and fear amongst the population in Burundi. The Palipehutu FNL ( Parti du Liberation du Peuple Hutu Front Nationale du Liberation ), created in exile in a Tanzanian refugee camp in 1980, attacked and killed hundreds of Tutsi civilians in two communes, Marangara and Ntega, near the Rwandan border in August 1988. The military responded in kind with repressi on and violence and as a result as many as fifty thousand of Hutu died or fled to R wanda (Lemarchand 1996: 126). Conflict continued to remain just below the surface, and outside events in Rwanda, like the RPF invasion in October of 1990 and subsequent mas sacres of Tutsi civilians in 1991 and 1992 intensified ethnic sentiment and fear. Coupled with crippling economic circumstances due to the falling coffee and tea prices in the worldwide re cessation of the late 1980s, Burundi found itself a political powder keg, with the politico military rulers struggling to (unity).
134 La Crise, 1993 In the political vocabulary of Burundi nearl the crisis probably because the gover nment army at the time tried to downplay the dimensions and directionality of the war. For this reason, the war and coup is rather, in a series of events that unfolded slow ly over several years. In 1993, the tide seemed to be changing for Burundians the Constitution had been written and accepted, government was seen as full of promise and hope for Bu rundi to avoid the ethnic clashes plaguing R wanda. Discussions commence d regarding the return of refugees from the 1972 genocide However, this vision remained short lived : mere months after the election, Ndadaye was assassinated by a group of Tutsi army officers fearful of violent reprisals against Tutsi by the new Hutu government and intent on another military coup. Hutu civilians responded in turn with their own expressions of violence. 4 was the common sen timent expressed before commencing pogroms against their fellow Tutsi civilians. Scholars estim ate that over fifty thousand Tutsi died at this time (Sch errerr 2002: 48). H ow many Hutu died over the span of the war remains unclear but massacres of Hutu als o happened in concurrence with the massacres of Tutsi at this time. The Tutsi dominated army would not stand for the destruction of their kinsmen, and responded with violence and repression as they had before against Hutu civilians. Estimates range from on e hundred thousand to tw o hundred thousand Hutu civilians dead, and almost two million citizens
135 dramatization of the [ethnic] cleavage, with its all or nothing s takes, has been carried tien 2003: 313) A nother decade and much more violence would pass before Burundi knew some measure of peace and resolution to the problem of ethnic violence animosity mattered a great deal to the outbreak of civil war in 1993 and subsequently, to the developmen t and framing of the CNDD FDD. This history, especially of violence and mistreatment, affected Hutu, political and ordinary alike, in a number of ways. For one, this history provided a way for the CNDD FDD to forage links to the civilian population, il, and fighting for equality and rights. The rebel movement also had much to learn about state management and the diffuse nature of political power and institutions in Burundi, which would later shape the internal structure and organization of the group. A fter this description of the prior conditions of the Burundian state leading up to the civil war, I now move to the comparative histories of other states in accordance with thes e four factors that I outlined above. Again, in particular I focus on the pr e Independence conditions: ethnic structures and the degree of state centralization or coherence, and two post Independence factors: political violence and change and the transitions of the late 1980s and early 1990s. It is with these factors in mind that I demonstrate later how these histories affected the development of rebel groups throughout the region, to be elaborated in Chapter 6
136 Rwanda It is perhaps most central to the examination of state development in Burundi to move to state development in Rwan da, for they shared many of the same attributes, kingdom structures, and a colonial power. These characteristics lead many to compare and contrast various points in their histories (Lemarchand 1970, Uvin 1999, Lemarchand 2006, Mansfield and Snyder 2005: 6) Much like Burundi, pre colonial Rwanda was characterized by a kingship structure to which all subjects fell under its rule, also called the Mwami. But in Rwanda the Ganwa class never existed and thus the pre colonial power struggle firmly rooted itself i n dichotomous issues of Tutsi leadership and oppression against a less powerful Hutu farming community. For many years the history of Rwanda was couched like that of Burundi, with the caste system supposedly minor in its discriminations changed through the advent of colonialism (Macquet 1961). Later scholars determined through oral histories that historical Rwanda created (Newbury, C. 1988 and Newbury, D. 2001). Rwanda and B urundi shared a colonial power and scholars are not surprised at the similarities in colonial governance across the two, however, many note the degree to which post Independence events, especially ethnic conflict and genocide, mirror each other. Dual Colo nialism: The Tutsi and Belgium As expressed above, the colony of Ruanda Urundi first became a colonial possession in the 1890s as part of German expansion into sub Saharan Africa. The World War I, under a very liberal trusteeship (that allowed for almost no oversight) from
137 the League of Nations, transferred to the United Nations after World War II. The kingdom coalesced under Mwami Rwabugiri from 1865 1895 (Eller 1998: 206). The Tutsi in Rwanda w ere co opted by the Belgian colonial authority in an effort to govern the colony through indirect rule by the Mwami and the traditional court, a motion cemented by the tales of the official court historian (and Catholic priest) of the colonial period, Alex is Kagame, who wrote of the mythical origins of the three brothers Gahutu, Gatutsi and Gatwa, all accorded responsibilities according to their social capacities. In the traditional myths, the Twa were insignificant, the Hutu lazy and irresponsible, and the Tutsi responsible and trustworthy (Kagame 1952 and 1959). Burundi, on the other hand, never had folktales of this nature underpinning ethnic relations. While similar to rule in Burundi, Rwandan society experienced far more authoritarian tendencies, becaus e of the lack of the intermediary class of Ganwa that promoted consociationalism and local conflict. In pre colonial Rwanda, traditional authority was reinforced in two ways: by the collection of taxes and the grouping of territories into vertical organiz ational structures of neighborhoods, hills and districts (Macquet 1961:100 102). These colonial Rwanda, rent the Belgian colonial authority, all power and governance decisions rested with the Mwami, who was Tutsi and favored ethnic ties. The decisions of the Belgians in Rwanda were also far starker than those in Burundi: competing identities surfaced as important facets of conflict in Burundi, whereas in Rwanda the colonial authorities reified an existing order, that Tutsi were of higher social caste than Hutu, and thus m ore
138 deserving of education, employment, and other benefits. Furthermore, Rwandan society was more homogeneous in its categorizations than Burundi: while intermarriage existed, it was far less common than in Burundi, and traditional clans tended to be of on e ethnic group. This is not to say that these differences in region and clan did not exist, but their impact mattered less in Rwanda than in Burundi. It is then very surprising that Rwanda colonial patterns with Tutsi leadersh ip continuing, despite historical dominance. Important in shaping post colonial politics in Rwanda was the Catholic Church. The Church in Rwanda became an important player during colonialism, with the Pere Blancs (White Fathers) establishing missions and s chools there as early as 1903, although they did not gain wider influence until the reign of Yuhi V Musinga ended, as he was hostile to religious influence (Longman 1997: 8 and Longman 2001:168). In the 1950s, church and Rwandan state relationships begun to weaken, mostly due to an influx of priests and missionaries influenced by social democratic philosophies of the post war period (Longman 1997: 12).These priests saw a majority populace under the yoke of minority political control, and actively sought to encourage Hutu consciousness of the political situation, a far cry from previous indirect political actions. The events of the 1950s in Rwanda would have devastating consequences for the region, to say nothing of the deleterious effect that it would have on its neighbor. They concern both the new relationships forged in the church and the legacy of inter ethnic relationships in Belgium, for it is these forces that shaped the behaviors of international actors that in turn pushed, prodded, and poked the woul d be independent country. Most of the priests who arrived in Rwanda post World War II were of Flemish (Dutch speaking) origin rather than Walloon (French speaking), who
139 experienced first hand political discrimination and tension in their native Belgium. Th ese priests took their experiences and used them to promote new democratic change in Rwanda, turning away from historical Tutsi leadership and encouraging a Hutu takeover a mass uprising by the peasant class against not only the Tutsi led monarchy, but also ordinary Tutsi civilians. In November of 1959 pogroms began, lasting two years and causing hundreds of thousands of Tutsis (as many as 300,000) to flee abroad, includin g the future leader of the RPF, Paul Kagame (Kinzer 2008:ch. 1 and Longman 1997: 12). The strength of these attacks combined with support from the Church convinced the Belgian colonial government to abruptly switch allegiances, and by independence in July 1962, the government consisted of an almost entirely Hutu presence, removing local Tutsi authority (Straus 2006: 21). It was this change that forever altered the nature of ethnic power in Rwanda, and set in place a chain of events that fell, like dominoes, between state power, political violence, and ethnic rebellion there. By 1959, four political parties existed in Rwanda, the two most vocal of each ethnicity being the PARMEHUTU ( Parti du Mouvement de l'Emancipation Hut u ) and the UNAR ( Union Nationale Rwan daise ), which represented the Tutsi monarchists. The violence initially began when UNAR activists roughed up a Hutu politician, setting in place a pattern of attack and counter attack, known as jacquerie, rural riots (Straus 2006: 178 and Lemarchand 1970:1 68). Armed opposition groups under the UNAR formed in exile during 1959 1962 for cross inyenzi during the genocide to describe all Tutsi. In March 1962, for reasons unknown, the raids caus ed a response in the form of serious ethnic massacres, claiming at least one
140 thousand Tutsi deaths (Straus 2006: 184). By late 1963, a new round of violence erupted. The new Kayibanda government perceived imminent threat from the one hundred and fifty thou number was exaggerated) (Straus 2006:185 and Lemarchand 1970: 198 206). Both the UNAR and the PARMEHUTU were factionalized, in addition to the decline of coffee production, the major cash expo rt for the small nation. In December 1963, a band of rebels invaded from Burundi, killing government soldiers, seizing vehicles, and caches of weapons, garnering more than one thousand to seven thousand men by the time they marched on Kigali (Straus 2006:1 85). Government soldiers eventually turned back the attackers, and the Kayibanda regime arrested and executed Tutsi politicians in UNAR, in addition to organizing local civilian led self defense committees. In a foreshadowing of the eerily similar circumst ances of the 1994 genocide, many thousands of Tutsi civilians were massacred in concentrated attacks organized at the commune and prefecture levels. Political Changes over the post Independence period Rwanda differed from Burundi in terms of who contr olled state power immediately from Independence. Straus notes two significant political trends of post The first republic under President Gregoire Kayibanda explicitly rewrote Rwandan lings were commonplace during the early years (until about 1964 and also, significantly, in 1973) of the regime.
141 Kayibanda downplayed mentions of ethnicity, but enacted far more discriminatory policies than his successor (ibid: 23). Kayibanda also solidifi ed southern rule in the country, as he favored government and military officers from his home region of Gitarama. Like Burundi, Rwanda also experienced a bloodless change in political power. In 1973, on the heels of the genocide in Burundi, northerner Juv enal Habyarimana replaced Kayibanda, stressing peace through the balance of ethnic and meant a strict quota system that purposely undercounted the percentage of Tutsi in the population (from fourteen to nine percent) and applied these to all government positions and services, including the provision of secondary and higher education. This was seen by the Hutu population as a way to counteract negative effects wrought by c olonialism, instill democracy through ethnic parity, and make up for lost opportunity (Uvin 2002: province of Gisenyi. This change in power happened around the same time as the genocide in Burundi caused a massive influx of Burundian Hutu refugees into Rwanda, promoting anger and insecurity, especially along political elites (Straus 2006: 189). In Jan uary 1973, Hutu civilians embarked on a purge of secondary schools and colleges to rid them of Tutsi students, which then spread to public and private employers. These Habyarimana staged a coup and arrested Kayibanda and leaders from his regime, violence, the regime enjoyed a quiet, if completely autocratic rule after five years of
142 military transition (Chretien 2003: 307). Habyarimana instituted a completely one party state under his ruling party MRND ( Mouvement rpublicain national pour la dmocratie ), where every adult, Hutu or Tutsi, became a member (Prunier 1995). The party sy stem closely resembled the levels already existing in the state, with party organizations at the cellule, commune, and province levels. This built a one party state, focused on ryday conduct and lives of civilians, ensuring that Tutsi were kept in positions of Church an up until the civil war. Ethnic Conflict in Rwanda leading up to the Genocide, 1990 1994 Unl ike neighboring Burundi, Rwanda experienced remarkably little ethnic conflict following the events of the immediate post Independence period, and for many years, rule, t he opening up of the country to international business, and great strides in development. Periods of violence tended to be intense but contained, occurring in the Independence period (1959 1964), again in 1973, and then not again until the civil war and le ad years of Rwandan history preceding the genocide, ethnic peace or an absence of ethnic the e nd of one party rule and the invasion of the RPF on October 1, 1990. While political
143 liberalization imposed from the international community did not start the war, it directly contributed to its outcome and the subsequent genocide. Drastic changes in econo mic and political conditions both at home and abroad reached a crescendo in 1989 (Chretien 2003: 320). The RPF, as discussed la ter on here and in Chapters 4, 5, and 6 grew out of a cadre of Rwandan exiles born and raised in refugee camps in the southern r guerilla army, the National Resistance Army (NRA) and took up arms against Obote in the Ugandan Bush War (1981 1986), hoping for more tolerance and acceptance in Uganda. Previously existi ng refugee committees in the camps, most notably the Rwandese Alliance for National Unity (RANU), had for a decade discussed how to gain rights of return from the Rwandan government. But violent rebellion against the Habyarimana regime was never explicit u ntil the future leaders of the RPF, closely allied the RPF in 1987, an inside organization with the NRA. I delineate the specifics of this birth in subsequent chapters. The RPF itself became an important character in inkotanyi armories. The fighters were quickly rep ulsed by the Force Armees Rwandaise (FAR), supported by French, Belgian, and Zairian fighters and weaponry (Chretien 2003: 321). Habyarimana also quickly acted to stir up anti Tutsi sentiment in the population, even staging a fake attack on the capital, Ki gali, on October 4 (ibid: 321). The RPF simultaneously recruited additional troops from exiles living in Burundi, Zaire, and Tanzania (Kinzer 2008: ch. 2). Owing to international pressures, press freedoms and
144 constitutional reforms legalizing multiple part ies became law in June 199. New parties, like the MDR ( Mouvement dmocratique rpublicain ), the supposed heir to PARMEHUTU, emerged as well as the renamed MRND under the name MRNDD ( Mouvement rpublicain national pour la dmocratie et le dveloppement )(Chr etien 2003: 322). But significantly, a new network of Hutu power activists and groups exploded onto the political scene, supported by the akazu (little house) of the First ous Kangura (The Awakening) newspaper, which would become the leading voice in the call to genocide and RTLM ( Radio Television Libre des Mille Collines) the Hutu hate radio station. Pogroms were organized at the communal level by the bourgomasters after every opening of Rwandan political space, resulting in widespread massacres across the country in places like Bagogwe (March 1992), Bugesera(August 1992), and Kibuye (1992 January 1993). (ibid: 325 and Straus 2006: ch. 7). In a foreshadowing of the peace process installed later in Burundi, negotiations began between the rebel RPF and the Habyarimana government in Arusha, Tanzania, in 1992. These negotiations, at least from the outside, opened up political space to opposition parties and were to provide a right of return for Tutsi refugees living abroad. International actors like Julius Nyerere celebrated the intricacy and success of the accords, but Arusha ultimately was ly (Jones 2001: 69). The process culminated in the signing of the comprehensive agreement after a two year struggle on August 4, 1993. Concurrent with the process a dangerous trajectory of ethnic polarization, wherein which extremist media, political
145 parties like the CDR ( Comit pour la Dfense de la Rpublique a group closely linked with the akazu ) and even ethnically based militia like the Interahamwe, the notorious group of genocidaires, rose to action. The peace brought by the accords, however, was to be short lived. For one, the Habyarimana regime never accepted these agreements continuing to incite ethnic conflict and fear of the Tutsi among the population, commit massacres, and arrest opposition leaders. Secondly, the RPF was also a wary partner in the agreements, and showed a willingness to return to war should circumstances wa rrant it (Kinzer 2008: ch. 2). The RPF resumed hostilities against the government army in October 1993, coinciding with the apparent plausibility to the notion of a Bahima (pan Tutsi) conspiracy to re conquer th e entire region and re impose the old feudal ord 278) These circumstances created a volatile political environment with potentially disastrous consequences for all of Central Africa, one that would have future reverberations e ventually engulfing the entire region The event that sparked this chain of events was the downing of the plane carrying both Rwandan president Habyarimana and Burundian president (and a Hutu) Cyprien Ntaryamira, ironically ret urning from the latest round of Arusha Accords on April 6, 1994. What happened next shocked even the most seasoned of all Great Lakes observers, as the next hundred days saw close to one million Tutsi and moderate Hutu murdered at the hands of the Interaha mwe, armed civilians, and the Rwandan army. It became clear for the killing, documenting future victims and distributing weapons in the months
146 beforehand to eliminate th e potential threat any Tutsi could inflict on the hegemony of Hutu political power. The genocide not only caused the immediate problem of violence and a return to civil war in Rwanda (the RPF immediately launched a counter offensive, taking Kigali in July of 1994 and effectively stopping the ethnic massacres), but more importantly, influenced events region wide. In Burundi, the genocide incited the Tutsi military and government to attack Hutu civilians with even more voraciousness, spurring a simmering low intensity conflict into an outright civil war. In Uganda and Tanzania, hundreds of thousands of Rwandan refugees streamed across the borders and caused strains on already poor economies. The Ugandan government under the NRA also faced additional issues of the level of support it would provide to former rebel ally in arms the RPF, and how willing they were to show their level of involvement to international donors. Finally, the coming to power of the RPF government and fear of retaliation resulting from the Rwandan genocide unleashed approximately two million refugees, most of them Hutu and many of them former Interahamwe and Rwandan government and army leaders, into the Kivu regions of Eastern Congo. This led to imminent human catastrophe, and the nascent R PF government now had serious security concerns in Eastern Zaire (Congo), an already unstable region in a crumbling country under Mobutu Sese Seko (Stearns 2011, Auteresse 2011, Lemarchand 2009). I explicate on the intricacies of the impacts of the genocid e and rebellion in t he region further in Chapter 6 The Rise of the RPF Here I briefly summarize the conditions of the state and the civil war that led to the creation of the RPF, providing context for a further discussion of the founding in
147 more detail in later chapters. This section details the conditions of the state during the immediacy of the civil war begun in October 1990 to the adv ent of genocide in April 1994. The Rwandan Civil War was so much shorter than the Burundian one because although they occurred at roughly the same time, more international intervention and pressure pushed the actors in the Rwandan conflict into negotiation early. The war in followed quickl y by periods of relative calm, and never did the rebel organization CNDD FDD ever reach the kind of military power and threat that the RPF did. Clearly, state power played a vital role in the conduct of the genocide; it was the strong organization and cent ralization of the state that allowed for lists of victims and weapons to be dispersed, and orders to be quickly given and then carried out. Furthermore, while some aka zu and the Habyarimana regime still exhibited not only political dominance, but social dominance as well. Even though elites cleverly manipulated the population in sum game, the Hu tu power elites were only able to do so by instilling the message through all levels of a highly organized state system, historically determined but perfected under the Habyarimana regime and consolidation of his one party state (Storey 2001, Prunier 1995) The strength of the state was such that many genocide theorists suggest that it thus killing became de facto state policy (Straus 2006: 308). Another factor worth consideration in the rise of the RPF and its subsequent
148 While the genocide left Rwanda penniless, pockmarked from war, and with far less inhabitants t han before, within four years a recovery was well under way, with state enterprises functioning at a higher level than most contemporaries with little to no corruption and new industries booming. Undoubtedly, the fast recovery owed some of its success to the strength of the RPF and the leadership. But even more crucial was the ability of the Rwandan state to revert to what it had always been: a highly organized, centralized political and social network with incredible sway over the civilian population, alb eit now one cemented largely on Tutsi holding positions of power. The RPF as a governing institution in the post genocide era fostered and encouraged even more centralization, reducing the number of potential points of power (the number of provinces went f rom 10 to 4 in 2003) and centralizing business and education concerns. Rwanda and Burundi: False Twins and Rebel Incubators? I now turn to a comparison of the histories of the states of Rwanda and Burundi across the four dimensions discussed in the introd uction: ethnic structures and the degree of state centralization in the pre Independence period, and political violence and coups and the late 1980s transitions as a result of international restructuring of aid in the post Independence era. In the pre Inde pendence period, Rwanda and Burundi seemed to mirror each other first Germany and then Belgium colonized kingdoms that looked on the surface remarkably similar, with a heterogeneous population ruled under a Tutsi monarch. The colonial powers even chose to jointly administer the two, which led to uniform policies towards ethnicity and indirect rule. But the 1950s and political agitation from the Catholic Church in Rwanda forever altered how ethnicity shaped politics, both in Rwanda but also in Burundi: thi s marks the first serious divergence in
149 the colonial histories of the two. From the Hutu Revolution of 1959, a new consciousness was born amongst both ethnicities, but it took two different routes to genocide and violence. In Rwanda, the turnover was immed iate, and ethnic violence perpetrated by both the would be Hutu government and the Tutsi inkotanyi fighters polarized post Independence politics. In Burundi, although violence occurred, the Tutsi monarchy remained staunchly supported by the Belgian colonia l administration, and thus e thnic conflict was downplayed. But the two countries experienced a difference in ethnic conflict due to not only colonial factors, but also the historical legacies of ethnic structures existing before colonialism. As described a consociational rule between royals and an intermediary social class focused political conflict on the internal: inter clan but intra ethnic. Rwanda on the other hand slowly consolidated an oppressive political and social structure based on caste and ethnicity over time that marginalized Hutu even before the colonial powers did. Hand in hand with the ethnic structure of the state was the level of coherence the pre colonial state exhibited. In Burundi, the pre colonial kingdom was a much more fluid enterprise pursued a more aggressive campaign of territorial expansion through conquest and war, one that also helped in pursuit of centralizing existing subjects and installing tax struct ures to support these efforts. In the post Independence period, coups and political violence in Rwanda were kept to a minimum for three reasons. First, the strong grip the Kayibanda and Habyarimana regimes maintained on the state b y centralizing and consolidating political power. Secondly and in line with the consolidation of power by the regimes was the concentration of political resources and power at a very small level
150 among a very small number of people that severely limited the abilities of would be overthrowers. Finally, past episodes of ethnic violence directed and controlled by political actors ensured a well founded fear among the population about the violent consequences of political action. In fact, political action within Rwanda effectively did not exist until the October 1990 invasion and subsequent negotiations with the RPF coupled with international pressures opened up the system to a multitude of parties representing all ethnicities and all political ideologies. The in ternational pressures at work included not only the international actors present at the Arusha Accords negotiating table, but also large international agencies like the International Monetary Fund (IMF) and the United Nations (and the countries who followe d their lead), who began to impose political and social conditionalities on aid funds, upon which the Habyarimana historically depended on, even more so after the bottoming out of world coffee and tea prices in 1988. In some ways, both Rwanda and Burundi b enefitted from international interventions like conditionality because of a political class otherwise reluctant to add to their numbers. But overall, as later explicated, very few transitions of political power occurred in the Great Lakes region, whether d ue to international actors or internal struggles. Uganda Another Kingdom, Multiple Sovereignties In Uganda, pre colonial power also took royal form, as in Rwanda and Burundi, but in Uganda, multiple kingdoms existed over multiple ethnically differenti ated groups, although conventional observers sometimes downplay the smaller kingdoms in favor of the larger and more elaborately chronicled Buganda kingdom (Chretien 2003: 140). The
151 introduction of modern humans to the area happened around the first centur y A.D., and like Rwanda and Burundi, was a mixture of Bantu agriculturalists and Nilotic pastoralists. The Bunyoro kingdom, with its powerful reputation became the repository of the rituals and rites of the Cwezi religion around the 16 th century (ibid:148) and controlled a loosely autonomous peripheral kingdom, although it lacked the centralization and dominance of other kingdoms in the region. The kingdom declined in the eighteenth century, to the benefit of other competing kingdoms in the area, like the Toro in southwest Uganda, and the Nkore dynasty, around Lakes Edward and George. Other smaller kingdoms included the Buzimba, Buhweju, and Kitakwenda, many of which were under the protectorate of the Bunyoro. At the end of the eighteenth century, the Nkore rose to power in place of the Bunyoro, establishing protectorates and attempting to hone in on salt works in Toro (ibid:150). In the southern regions of Uganda, the Giseka and Bugesera clans interacted with the kingdom in Rwanda, sharing alliances and ma rriages (ibid: 151) and the Buha and Buzinza controlled salt and iron enterprises. Thus pre colonial Ugandan history was more decentralized than other kingdoms, and in general, dynastic clans existed side by side with little inter kingdom violence or conqu est through war. The Buganda kingdom began to expand through territorial conquest in the seventeenth century, consolidating power and replacing Bunyoro by the beginning of the nineteenth century (ibid: 156). Buganda and the other clans differed greatly in their abilities of institutional capacity and change. In Buganda, a centralization process under the kabaka (king) began in the late seventeenth century that included the elimination of potential rivals of the kabaka often his brothers. Colonial explorati on by foreign powers of Uganda began in the eighteen
152 colonial Uganda resembled a patchwork of various clans, chiefs, dynasties and fiefdoms rather than one centralized kingdom. Explorations led by the Royal Geographic Society and British Government began in earnest in the 1850s in search of the source of the Nile, systematically exploring the many lakes an d waterways. By the 1870s, only two kingdoms of any importance existed: the Bunyoro, under Kabarega, and the Buganda under Mutesa. Because Uganda initially was explored by the English, missionaries and religious explorers were Protestants and not Catholics like those in Rwanda and Burundi, where natives could be called to the priesthood. Britain proclaimed Uganda a protectorate in 1894 and the 1910 Brussels Conference solidified borders between 214). True consolidation of the smaller kingdoms under the Buganda finally occurred in 1896, with the Nkor e, Toro, and Bunyoro subsumed. The British colonizers took advantage of this, with administrators regulating and validating co urt processes, which would eventually become the Parliament in the immediate pre Independence era (ibid: 225). The 1900 Agreement between the British and Uganda modernized the feudalism of the Buganda, keeping the king, the justice system, and the Parliame nt, cementing indirect rule. However, real self government did not occur until the 1950s. Colonial Uganda was largely free of the ethnic conditions of comparable Rwanda and Burundi, although Bugandans were given greater benefits. Because of changes in edu cation and throughout the early twentieth century (ibid: 266). The new consciousness encouraged
153 nationalism and civic association, both factors that would gain more impo rtance during the Independence movement. Violent uprisings occurred against both the colonizers and their Ugandan counterparts, the Baganda, in 1945 and 1949, sparking the nationalist struggle. The 1950s brought a crisis in would be Ugandan identity, as th e heterogeneity among people and concurrent modernization led Ugandans to a debate about unity. Because of this, many parties developed throughout the decade, including north, pushed for political control outside of Buganda hands. Independence Independence in Uganda was far less violent than that in Rwanda and Burundi, the birth of India in 19 independence in October 9, 1962. Throughout the process, the old demons of competing regions and kingdoms re emerged. The Buganda kingdom, which until this point had been awarded great power and autonomy from the British, demanded their own sovereignty, including the continuing of their own parliamentary body. This position was reversed when they refused to participate in the 1961 general elections, thus ensuring a Catholic prime minister, someth ing the protestant Buganda greatly detested. still demanding an autonomous council (Chretien 2003:295). The March 1962 constitution unified the five strongest kingdoms and doled the position of the federal presidency to the Buganda, the only autonomous one among them, who created an official party, the Kabaka Yekka (KY) ( The King Alone in Luganda). Although KY won
154 the general elections in Buganda territory, overall the UPC g ained majority political power in Uganda. Cleavages among religious, ethnic, and regional lines continued to abound in the early independence period, leading to the rise of the Obote dictatorship. Uganda was subject to dictatorship because it was thought that the only way to govern was through absolute authority, in order to keep ethnic and regional loyalties from causing conflict or violence. Early Changes and Dictatorship By 1964, Obote and the UPC Obote and the UPC gained an absolute majority over political and military power, including placing northerners in positions of power in the army. In 1966 Obote suspended the constitution and proclaimed himself President (ibid: 296), and the following May f orced the kabaka (king) Mutesa into exile, and abolishing all monarchies within Uganda. Life continued without much incident under Obote until an overthrow by Idi Amin, the chief of staff of the President, in January 1971. Almost immediately, violence erup ted and the economy tumbled. Asians, who played large parts in the Ugandan business community, were expelled in 1972, followed by massacres, torture, and execution. It would take another eight years for internal rebellion to the regime coupled w ith Tanzani an military intervention in 1978 1979 to overthrow Amin, and return Obote to power, successfully staging elections in December of 1980 (ibid: 297). The Bush War and the Rising of the NRA The chaos of the Amin regime, the war with Tanzania, and the overthro w of Amin all led to a lack of development in Uganda and the stifling of political expression and opportunity. In a foreshadowing of patterns to come, many new political opponents
155 were created when Obote forced former government members into exile, like Y oweri Museveni, a University of Dar Es Salaam (UDSM) graduate who served as minister of defense under Obote until forced to take refuge in Tanzania. Museveni was one of a new class of political elite in sub Saharan Africa, educated at UDSM when the univers ity of African and Outside Scholars, including Tanzanian Julius Nyerere and fellow Sudanese understanding of African democracy and political change that would reverberate across Frente de Libertao de Moambique ( FRELIMO). Post the development of Front for the National Salvation of Uganda (FRONASA), modeled after FRELIMO and operating in Tanzania against the Amin government (Kasfir 2005: 278). FRONASA also include d several Rwandans, like Paul Kagame, who would become instrumental in the Rwandan Patriotic Front (RPF). Although developed in exile in the 1970s, the movement participated heavily in the Uganda Tanzania War in the later part of the decade to remove Amin, sometimes coming into direct conflict over faction, Kikosi Maalum (KM), both of which operated within the Ugandan Na tional Liberation Army (UNLA). After the assistance of Tanzanian troops, Obote and Museveni returned to Uganda during the transition from war to peace, including the December 1980 elections. It was widely believed that these elections were manipulated by the UPC (ibid: 279), which helped frame the would be National Resistance Army (NRA) rebellion developing with Museveni at the head. The
156 rebellion officially took up arms against the Obote regime in February 1981, with few weapons at their disposal (Museveni 1997: 122). For the next five years, the civil war ebbed and flowed, with the NRA utilizing civilian support and resources to build popular legitimacy while winning military victories against the government army, although it took until 1985 for the rebel organization to gain enough strength to defeat the UNLA (th e government army) (ibid: 281). The war fina lly ended with the seizure of Kampala, the capital city, and the pushing back of the UNLA to Sudan in early 1986. The effect the NRA had upon rebel organizations not only in the Great Lakes region b ut across sub Saharan Africa cannot be underestimated. Indeed, without this rebel organization, the RPF, Palipehutu FNL, CNDD FDD, and AFDL would never have 1986, the year organization that focused less on the leader or liberation from white rule and more on liberation from bad African leaders who plagued the continent from east to west. Not only did the success of the NRA i n Uganda illuminate a path to victory for other rebel organizations to follow by example, but the group actively supported other rebel organizations materially, including most notoriously the RPF (which grew out of the Liberation Army (SPLA) in Sudan, the Alliance des Forces Dmocratiques pour la Libration du Congo Zare (AFDL) in Zaire/Congo,
157 and others. The level of involvement in new reform rebellions allowed Uganda to benefit from economic development (selling weapo ns and exporting resources in civil war zones) in addition to becoming a major regional power in the East African Community (EAC) and the African Union (AU). The NRA government also provided a post conflict model for would be rebel governors, lasting more than two decades and four Presidential elections (two of which were multi party). While I elaborate in more detail on the rebel organization in later chapters, the effect of the NRA on political dissent and expression in Africa must be highlighted here. Ugandan History in Comparison Now I focus on how Ugandan history compares to the history of Burundi and Rwanda in regards to the four previously mentioned factors: the degree of state centralization and ethnic relations pre colonially and during the colon ial era, political violence in the post Independence era, and the transitions period of the late 1980s. The last factor, that of political transitions imposed by the outside in the late 1980s did not have the same effect in Uganda as it did on fomenting re bellion in Rwanda and Burundi, because the NRA regime was establishing itself as the post conflict political power and thus cannot be compared here. All three countries experienced bloody periods post Independence, although Rwanda and Burundi had the dubio us honor of hosting ethnically driven conflict between Hutu and Tutsi, while in Uganda political violence extended to political enemies of the Amin regime. The regionalist tendencies of conflict in Uganda tended to downplay ethnic violence in favor of poli tical discrimination by colonial ethnic structures reflected greater
158 This meant that emerging Independent Uganda would also be composed of a myriad of ethnicities, both Bantu and Nilotic. No one ethnic group in Uganda controls more than 16.9% of the population, the largest of which is the Baganda, according to the 2002 census. (CIA World Factbook), implying that political dominanc e by one singular ethnic group would be incredibly difficult and probably impossible without coalition. Finally, comparing the kingdoms across Uganda, Rwanda and Burundi finds some similarities, although less than previous colonial explorers stated. In all three countries, a centralized form of kingship existed; however, in Uganda the diversity of ethnic groups also meant a plethora of kingdoms. The largest and most organized, the Buganda, resembles Rwanda and Burundi most closely. But the resemblance does not become apparent until the Buganda made great strides towards centralization in the 1800s. The Buganda kingdom was not nearly as centralized and structured as the mwamiships in Burundi and Rwanda, and chiefs and sub chiefs in Buganda had far more autono my than their counterparts across borders. The Democratic Republic of the Congo Discussing the preconditions of the state in the Democratic Republic of the Congo, formerly known as Zaire under the Mobutu regime, is perhaps the most difficult task in rela tion to other states in the region for a number of reasons: the lack of a kingdom like structure in the pre colonial period to structure subsequent relations, brutal colonialism by an individual rather than a country, the sheer size, and its utterly chaoti c post Independence history. People populated pre colonial Congo far earlier than other areas, for more than 80,000 years. Modern settlement by Bantus began in the 7 th century AD. Although modern day Congo covers an enormous swath of territory, the
159 namesak e kingdom, that of the Kongo people, established itself in the 14 th century and dominated power politics there. The Kongo were notorious for selling subjects of smaller dominated kingdoms into slavery, especially to early Portuguese explorers. These explo rers also brought Roman Catholicism to the kingdom, beginning conversions as early as the 15 th century (Nzongola Ntalaja 2002: 14). A clear difference between the pre colonial history of the Democratic Republic of the Congo and other countries in the regio n is the early engagement between the peoples of the DRC and outside influences. The Kongo kingdom was highly militarized, and interstate as well as civil war broke out continuously during its history, although civil war became the more common practice af ter 1600. Colonialism took its toll on the Kongo and other kingdoms in the country, ola Ntalaja 2002: 13). The country today reflects the multitude of kingdoms present in the pre colonial era, with over 250 separate ethnic groups, the overwhelming majority of which speak Bantu languages. intrigued European monarchs and would be financiers, including King Leopold II of Belgium. The Brussels International Geographical Conference of 1876 telegraphed these intentions to the rest of the world under the guise of establishing an association of entrepreneurs, journalists and doctors to counter the slave trade (ibid:15). By 1885, of territor y to be recognized as the Congo Free State (Chretien 2003: 215). Other
160 plans for Africa were benign and in any case were happy that Belgium stayed out of the melee of colon ial claims at the time. The Congo Free State practiced monopolies over ivory and rubber, both of which caused decades of pain for Congolese forced to work under them. The Congo Free State government seized lands for public or crown use, greatly changing th e nature of agriculture in addition to reaping profits from rubber and other products grown on them (Nzongola Ntalaja 2002: 22 23). As many as ten million Congolese died under his tenure, as a result of murder, starvation or exposure, and disease (Hochschi ld 1997: 225 234). Leopold proved his willingness to fight against other colonial powers in skirmishes that happened along the eastern Congolese border with Rwanda, Uganda, and Burundi in the 1890s. Eventually, compromises were signed in Brussels in April 1900, and again in 1907 and 1912, firmly establishing Congolese borders. In 1908, the Congo Free State was dissolved and possession returned (although it had never really left) to Belgium. Leopold died in 1909, although very little changed in terms of gove rnance and economic structures already established according to Free State principles. Belgian colonialism in Congo followed patterns of trade control and economic dominion rather than the indirect models that used traditional mon archies in Burundi and Rwa nda. Because of the size of the territory, trading posts were established in key areas to connect trade routes back to important ports and centers of commerce, like the one in Leopoldville (now Kinshasa). State rule was established at the local level using chiefs and traditional rulers in the 1920s and 1930s, and the 1933 reforms set the structure of provinces and territories that was used for the rest of the twentieth century (ibid: 36). After World War II, Belgium took small steps towards
161 incorporating th e few educated and socially connected Congolese into the government structure and society in eventual preparation for Independence. These evolues (civilized Africans) were to be the leaders of Congo after Independence, thus linking the old colonial power t o the new country. Independence: Promise and Failure Resistance to colonial rule in Congo took place at the same structures the colonial government devised: in the tax, labor recruitment, and political autonomy systems. Small rebellions had become commo n during the early part of the twentieth century, in Shi territory (1900 1916) and in Luba Katanga (1907 1917) (ibid: 41). Additional strikes occurred among workers in the mining and farming sectors throughout the Belgian colonial era, especially in the W orld War II era when Africans resented the impositions the colonizers placed on the colonized. These actions paved the way for the January 1959 rebellion in Kinshasa, a major event in the decision to grant Independence. In Congo, all sectors of society ba nded together against the colonial powers: workers, traditional chiefs, and even the evolues in direct contrast to elite driven rebellions in other parts of Belgian Africa. The Alliance du Bakongo (ABAKO), a group of evolues created in 1950 originally for the purpose of social integration and cultural emancipation of the Kongo people, dealt the first blow to the Belgian colonists in August 1956 with a public meeting designed to agitate Congolese against the Van Bilsen 5 30 year plan for Congolese independen ce and instead call for immediate independence. Elections held the following year cemented the vision of ABAKO and defined the nature of Congolese independence protest over the following two and half years. Political reforms of the period installed partie s to become the most important
162 leaders in the revolutionary struggle: the ABAKO, the Parti Solidaire Africain (PSA), the Mouvement National Congolais (MNC) of Patrice Lumumba, Balubakat of Jason Sendwe, and the Confederation des Associations Tribales du Ka tanga (CONAKAT) of Moise Tshombe. All of the parties except for the MNC were ethnically or regionally based, and many, including the MNC, split over the course of the drive to Independence. Lumumba, Kasai born and self educated, surfaced in Kinshasa workin g as a publicity agent for a brewery, learning organizing principles and engaging in civic associations in his previous car eer as a postman in Kisangani. He became so well known a political actor that by 1958 he was met by other African independence leader s Tom Mboya and A.R. Mohamed Babu to join the All in Accra, Ghana. In Accra, Lumumba interacted with important pan Africanist leaders of the time, like Kwame Nkrumah (Ghana), Frantz Fanon (Algeria), Gamal Abdul N asser (Egypt), a nd many others. profoundly, and he returned to Congo with a new vision of inclusiveness among Congolese, Pan African cooperation, and economic independence from colonial powers, a vision he shared at a m ass pol itical rally in December 1958. Lumumba continued to agitate for freedom from Belgian rule, and encouraged others to do so. On January 4, 1959, another rally was held by the ABAKO party, although party leaders were denied the permit to hold the rally by the Belgian mayor of Kinshasa. The crowd that gathered refused to disperse, throwing rocks and attacking motorists, setting off three days of riots which other Congolese joined. Official estimates place the number of Congolese dead at 49, injured at 11 6, and European injuries at 15, although other estimates number closer to 300 dead (ibid: 86). This violence sparked other protests
163 across Congo, including Lower Congo and Kisangani. The Belgians eventually assented to demands for independence immediate w ith a convention held in early 1960 deciding to grant independence on the 30 th of June, 1960. The fragile congress of people who fought for Independence, however, would soon fall apart. Independence ushered in ethnically driven conflict, as rivalries and even small civil wars (between the Hutu Tutsi in the East and Lulua Baluba in Kasai) broke out, in addition to a violent call to national unity in Kinshasa, spearheaded by now prime minis ter Patrice Lumumba. Less than a month after the inauguration of Lumumba and the granting of independence, a rebellion erupted in Katanga, backed by Western The Unite d Nations organized the Operation des Nations Unies au Congo (ONUC) to prevent direct confrontation between the West and the East carried out in the new state, including 20,000 troops and a nation building brigade (Ngonzola Njalala 2002: 94). Lumumba trie d to calm initial disquiet in the armed forces, who believed that change between the colonial power and the new state power was not happening quick enough. He did so by installing new leaders (one of which, a colonel designated chief of staff of the new Co ngolese army, Joseph Mobutu) and promising reform. Lumumba ignored ONUC and Belgium by maintaining relationships with Soviet Russia. This was absolutely unacceptable to Brussels, New York, and Washington, and effectively sealed civilian coup that sparked a constitutional crisis on September 5 1960, quickly followed by a military coup spearheade d by Mobutu on September 14th. At the same time active
164 rebellion sparked in the south of the country in the mineral rich region of Katanga. The Shaba rebellion was declared July 11, 1960. Katanga had become a settler region during Belgian colonialism and s ettlers remaining after colonialism found allies in Moise would later occur in Zimbabwe. The se cessation was ended by UN military force in January 1963. Over the next several years, chaos enveloped the Congo, as the small rebellions emerged, including one led by Laurent Desire Kabila (leader of the AFDL who would take the capital from Mobutu in 1996) in the east and supported by the Cuban revolutionary, Che Guevara. ocket: Kleptocracy and the Regime President Mobutu declared himself head of state on November 24, 1965, after five years of politicking in Kinshasa amongst the post Lumumbist political elite. The new to bring forth a Congo free from strife and conflict and ready to embrace development. Although supported by Western forces, Mobutu employed personal rule as the dominant political strategy, making all political actors dependent on him for spoils and powe r, and using the state as his personal treasury. He changed the name of the country to Zaire to Africanize the regime enterprises operating in the country slowed the growing economy in 1973, with declini ng growth throughout the 1980s. By 1978, even the major Bretton Woods economic agencies stepped in to provide stabilization, training, and adjustment (ibid: 151), with no luck reversing the trend. By 1990, most foreign aid fle d the country and most Zairians struggled to attain a minimum level of poverty in their standard of living. The state
165 became a shell of what it was under colonialism, unable to provide even basic services or governance, and theft, neglect and decline ruine d the social and economic fabric of life there. Internal opposition was fairly non existent, although external opposition grew, especially under former Mobutu loyalists who were forced out the regime. The 1980s saw some political change, including the deve lopment of illegal opposition parties, like the Union pour la Democratie et le Progres Social (UPDS) led by former Mobutu crony Etienne Tsisekedi that defied the ban and organized pro democracy rallies in Kinshasa in 1988. International donors in conjuncti on with internal pressures forced the lifting of the ban on multi party politics in April 1990, roughly the same time as reforms imposed on o ther countries in the region. time, repositioning themselv es for greater power in creation of the Union Sacree (Sacred Union), a conglomeration of opposition parties, in July 1991. With 204 opposition parties emerging, Mobutu strategized as to how to maintain power despite clear popular sentiment against him: he did so by appointing Etienne Tshisekedi prime minister, although thi s only lasted until October 1991. Complicating the political transitions during this period were ongoing battles between Mobutu, the Army, and citizens, as the value of the Zairian franc p lummeted and soldiers were no longer able to even into the twilight of the regime: in late 1994 and 1995 as the former Rwandan Armed Forces and government fled across allowed the genocidaires to set up residence among refugees in the East, and refused to collect contraband brought across by them (Stearns 2011: ch. 1 and 2). This immediately soured any potential good relations with the fledgling RPF regime, but also
166 ensured that Congolese citiz ens in the East would be trapped between the genocidaires in the camps and the RPF forcefully pushing to have them returned and threatening to invade Congo to do so. The West and the East: Two Congos, t oo Many Problems aintaining his kleptocratic regime and autocratic political power was to ensure that Congo stayed divided, ethnically and physically. Security forces controlled exclusively by Mobutu gave priority in membership to his ethnic group, the Ngbandi and co Equat eur (a central province) ethnics, the Ngbaka. Public resources were not used for education reform or building roads, especially the further one traveled from Kinshasa, but instead prioritized according to the patronage principles of Mobutu. Ethnicities and ethnic relations differed tremendously across the provinces, and some (like those in the East) experienced greater int er ethnic tension than others. In some ways Congo was simply too big, and historically controlling the outer regions was incredibly diffi cu lt no matter who was in charge. The proximity of ethnic groups in the East to co ethnic and co language group members across borders also slowed any ideas of Congolese nationalism. In the East, the languages of ethnic groups varied: Shi, Kinyarwanda/Kiru ndi, Kihumu, and many others, but the lingua franca became a dialect of Swahili widely spoken across the Eastern regions. Swahili was brought by slave traders from East Africa in the 1800s, but in the modern era many people use it as the language of commer ce, social life, and even religion. In the West, the lingua franca became Lingala, the language of the capital tempered with strong French and Portuguese influences. The linguistic divide helped ensure that the East would never hold political power in the West, although Goma and Kisangani were
167 historically important ports and centers of higher education. Furthermore, the development of refugee camps along the Rwandan and Burundian borders in the mid 1990s following their respective civil wars antagonized Co ngolese citizens who again saw the interests of Kinshasa and Mobutu overtaking those of the citizens of the East. But disconnect in the Western and Eastern halves of DRC extended further back. In the East, issues of land conflict and reform, ethnic tension Banyamulenge/Tutsi and the native autochthones (who were closer to Hutu, although varied in their ethnic groups) existed in ways that rarely found traction in the West, although Mobutu meddled from time to time, such as the 1989 identification rulings passed in the Eastern regions only (Stearns 2011: 65). This shows highly fragmented ever projecting much power. Kabila, Rwanda, and t he AFDL This section describes how Laurent Desire Kabila, one and the mentioned above during the rebellions of the 1960s, came to rise up against and eventually overthrow the Mobutu regime during the mid 1990s. It must be stated of course, that of all the rebel organizations in the region, the one that benefitted the most from outside influences and an invasion force by the Rwandan (RPF) army (Stearns 2011: 43), with support from former Ugandan rebel Museveni. Museveni had met Laurent Kabila in the 1980s, when Kabila was based in Dar E s Salaam, Tanzania (ibid: 53). Kabila, the former Marxist revolutionary instrumental in the Katanga revolution of the 1960s, settled int o an easy life in Dar Es Salaam with a diplomatic passport from Tanzania, occasionally writing
168 treatises on Marxist revolutions and traveling to Communist China, but otherwise quietly fading into obscurity. Along with three other Congolese leaders, Deo Bug era, an architect from North Kivu, Andre Kisase Ngandu, an aging rebel operating in the Ruwenzori Mountains near the Ugandan border, and Anselme Masasu, the youngest would be rebel leader by two decades who currently served in the RPF army (ibid: 87), Kabi la and company settled into Kigali in the summer of 1996 along with the Rwandans to plan the attack. They drafted a founding document and discussed questions of the spoke sperson of the movement. Although both Ngandu and Masasu were able to security forces, weak as they may be at the time, proved difficult. Thus Rwanda and the RPF army st epped into provide military assistance, including troops, material, and weaponry. The first battle occurred in late August of 1996, as Banyamulenge soldiers attempted to cross into Rwanda to attend training camp and were intercepted by the way. This sparked the first Congo war, which many Congolese involvement. Comparing Congo to the Rest of the Region To return to a point made earlier, the Democratic R epublic of the Congo is the context hardest to compare to the other countries in the region for several reasons given above. In regards to pre colonial Congo, ethnic groups were so numerous as to preclude any serious form of ethnic domination, besides tha t resulting from sheer size of the kingdom (as in the case of the Kongo), occurring as in Rwanda and Burundi. There
169 were almost no cases of historical kingdoms dominated by one ethnic group socially placed ahead of another. For the most part, the pre colon ial ethnic structures of many groups with their respective traditional chief governance still exist today in the DRC. As to the degree of state centralization or coherence existi ng before colonialism, there was none. The Kongo Kingdom controlled large swaths of territories of what is now DRC, but the formation of Congo did not occur until colonialism knit all the would be provinces together. Even now, many would argue that Congo s uffers from a serious unity problem. As for the post Independence factors discussed in the other countries, political violence and coups were non 1990s. Kleptocracy and strict control of the security apparatus of the state ensured no credible challengers could emerge, and the disastrous conditions of life wrought by most foreign aid money and even required international intervention in the late 1970s to restructure aid and governance, the late 1980s transitions as a result of international restructuring of aid that occurred in other countries across the region did not occur in Congo until later in the 1990s. Even then, the transition in Congo was the longest in time frame and the least violent, with civil war only occurring when outside forces (Rwanda and Uganda) became involved. How does State History Matter to Rebell ion? In this section, I explore how the histories of these states impacted the development and outcome of the successful rebel to ruler transitions, linking state conditions back to the crux of the thesis, rebel movements and their development. I find
170 that while the preconditions of the state before the rebellion, including historical political and social development greatly affect the behavior of a rebel organization, these patterns and legacies are by no means the only variable that do so. Furthermore, th e preconditions of the state have an interaction effect in combination with the rebel Organizational Power The nature and antecedent conditions of the state matter a great deal to the w ay in which a rebel organization develops. As shown in the histories of the various countries and their corresponding rebel rulers the CNDD FDD (Burundi), RPF (Rwanda), NRA (Uganda) and AFDL (Congo), a legacy of centralized relations and a history of monar chial relations existed in three of the countries: Rwanda, Burundi, and Uganda. Scholars state that it is difficult to argue that these pre colonial histories had no effect on modern relations ( Herbst 2000 Boone 2003, Hyden 2006), and some argue that the y conditioned social behavior in such a way as to make genocide possible in Rwanda (discussed in Straus 2006: 149 150). In the DRC, the rebel group AFDL that developed did so in a chaotic fashion similar to that of the DRC itself, with many factions and co mpeting interests. The nature of the previous state governments provides conditioning for which social behavior and political protest (of which rebellion is certainly a form) models itself upon. Legacies in Practice The antecedent conditions of the state can provide a model for the rebel group to use in its development. But not all rebel groups follow the same structure that the state did. The question then arises of how much the rebel movement fighting against the
171 state chooses to resemble (or not) the s tate. Further, how much do these strategies matter to the overall development and trajectory of the rebel movement in their quest to change the political power of the state? It seems easy enough to say that the more hierarchal a pre existing state, and the longer these hierarchal relationships have existed in society, the more likely it is that the rebel group would emulate these behaviors. In no place is this example more apparent than in the case of the RPF in Rwanda. Unsurprisingly, Rwanda is the countr y in the region with the most severe of ethnic structures conditioned by a highly centralized pre colonial kingdom that became a highly ce ntralized post colonial state. The RPF government is also the former rebel movement in the region most accused of beco ming authoritarian in their post rebel governance (Seay 2012, Longman 2012). The Nature of Social Power Social relations, in some way are conditioned by state political practices, also matter a great deal to the development and outcome of a rebel movemen government takeover. The interactions of society conditioned by the antecedent nature of the state can temper these relationships and provide either a hostile or welcoming social environment upon which would be rebels can act. This is espec ially crucial when 1971) and explaining rebellion as a process of grievance against the standing government regime. Grievances The preconditions of the state before th e civil war erupts can provide a legacy of relationships which rebels may draw upon. For example, l egacies of the r elationships
172 between king and subject during pre colonial times or citizen and ruler in the modern era can have lasting effects. Social grie vances as a result of relative deprivation in cause for war, and this is certainly the case among rebel groups in the Great Lakes region. In Uganda, torture and ethnic m assacres at the hands of Idi Amin ensured strong feelings against the regime sparking the first rebellion, led by Milton Obote and populated by future NRA members. It was this sentiment against the government that m minded rebels unafraid to challenge a government that deprived citizens of rights (Museveni 1997). The grievances against government) is infrequently used among local populati ons and even less so in government and official usage. Similarly, the RPF also focused on grievance based motivations for rebellion stemming from the ethnic massacres and discrimination of the past suffered under Hutu govern ments against Tutsi civilians. T he CNDD FDD in Burundi likewise held enormous historically driven grievances against the Tutsi government there, for ethnic violence and discrimination, especially the 1972 genocide. And although the AFDL in Congo grew mighty under foreign direction, it was no less disastrous policies toward the East, to say nothing of his general disregard for the well being of those under his government. The State as an Intervening Forc e The histories of the various states in the Great Lakes region matter a great deal to the way in which rebel groups there developed and operate, both as rebel
173 organizations and as governments. That being said, however, the antecedent conditions of the st ate are not a separate independent variable in this particular analysis. Although countries with similar histories and conditions abound in the region, similar outcomes do not. For example, both Rwanda and Burundi share remarkable similar pre colonial and colonial histories, as well as strikingly analogous post Independence features. Even though these countries resemble each other c outcomes, including the ones this study is most concerned with, are very different. In Rwanda a rebel group that closely mirrored the organizational relations of the state came to power through a military takeover after a devastating genocide. In Burundi, a rebel group that looked very unlike the state came to power after negotiations with strong civilian support after a decades long civil war of varying intensities. T hus the state definitely matters to the way in which rebel organizations in the Great Lakes region and interacts with the source of social power and the nature of organization of the groups, but is not a wholly separate phenomenon Conclusion The chapter provided an examination of the histories of the states in the Great Lakes region to illustrate how the antecedent conditions of these places conditioned, primed, and influenced the rebel groups that would eventually rise up to take power there. In particular it used Burundian state history and four factors there ethnic structures and state cohesiveness in the pre Independence period and violence and coups and internationally imposed transitions in the post Independence period as a lens to comp are the antecedent conditions of other states in the region Rwanda, Uganda, and the Democratic Republic of the Congo. Rwanda and Burundi share the
174 most similar pre rebellion conditions, although discussed in later chapters, especially Chapter 6 the devel opment and outcomes of their respective rebel organizations differ colonial and post Independence history provides strong patterns and incentive for a rebel organization there to centralize their organizational structure, as the count ry was used to centralized leadership, if dictatorial in practice. However, both Uganda and the DRC differ wildly from Rwanda and Burundi in terms of ethnic structures and how those structures affected rebellion. Finally, it becomes apparent that comparing Congolese history to those of the others in the region is most difficult, as the legacy of hundreds of ethnic groups and minor kingdoms and a completely dysfunctional kleptocratic post Independence regime makes any analysis of rebel organizations there co mplicated. The legacies of these histories in regards to how rebel organizations developed in the respective countries becomes more apparent in the following chapters, but this chapter serves as an introduction to the environment in which rebel organizatio ns emerged from, and the states against which they went to war. Notes 1 I use Zaire (1970 1997), Congo and DRC interchangeably to signify t he same countr y. 2 The forcible removal of villagers from one part of the country to another in order to break regional loyalties and further control the population. 3 power in Africa (Herbst 2004).
175 4 Personal Interview, November 10, 2010 5 Van Bilsen was a Belgian college professor who in 1955 wrote a pamphlet on the gradual transit ion of Congolese independence over a 30 year period, endorsed by Belgian groups in Congo connected to the Catholic Church.
176 CHAPTER 4 TYPES OF ORGANIZATIONAL POWER AND THE IMPACT ON SUCCESS: HOW REBEL ORGANIZATIONS GAIN AND MAINTAIN INTERNAL COHESION Introduction In the late 1980s and 1990s, r ebel groups like the Rwandan Patriotic Front ( RPF ) National Resistance Army ( NRA ), the Movimento Popular de Libertao de Angola Partido do Trabalho ( MPLA ), and Resistncia Nacional Moambicana ( RENAMO ) succeeded in changing the African political landsc ape, bringing change to post 1992: 38 41) sought changes in regime structure and the nature of political power that reforms that occurred in the late 1980s and early 199 0s as described in Chapter 3 These groups sought the end of the one party state, economic liberalization and the freeing of industries from under the th umb of government or parastatal enterprises controlled by cronies, and less ethnic or regional discrimination in public provisions. Many of these rebel groups emerged in civil wars with roots in rapid economic and political changes brought about by liberalization directed by i nternational donors and government agencies. One such g roup and the subject of this chapter is the CNDD FDD ( Conseil National Pour la Dfense de la Dmocratie Forcs pour la Dfense de la Dmocratie ) in Burundi The CNDD FDD emerged out of a civil war brou ght upon by liberalization of a military regime. The war began as a result of the assassination of the first democratically elected President in October 1993. Melchior itself was due to ongoing pressures from outside for Burundi to quick ly democratize and
177 engage in multip arty politics since the 1990s. Despite all efforts to seize the capita l and force a victory, the CNDD FDD was unable to do so, and consequently, sought political power by negotiation when stalemate seemed inevitable. This chapter explore s how the type of organizational power employed by the rebel gro up prevented it from achieving dominance on the battlefield, stalled the negotiations process, and caused major splits and factionalization among the group while it was still i n the bush Although the CNDD FDD was unable t o take power militarily, it achieve d political power through a negotiated settl ement that ended the war. The chapter also elaborate s on t he organizational type and strategies employed after the group left the f orest and went to parliam ent. It proceed s from here exploring in depth the transformation of the CNDD FDD over its lifetime First, I introduce the rebel organization, detail ing the paradox of regime success despite characteristics of the group and its org anizational strategies that would seem to make governing difficult. Further, I provide examples for the theory of rebel to ruler transition described in chapter two and detail a new analysis of the movement as seen from the perspectives of civilians, parti cipants, and leaders. I also focus on how the conditions of the state impacted the development of the CNDD FDD and its social and political relations. Finally, I introduce another rebel group also active in Burundi alongside the CNDD FDD, the Part i pour la libration du peuple H utu Forces nationales de libration ( Palipehutu FNL 1 ) and provide a comparison betwe en the types of organizational power of the CNDD FDD and FNL to elucidate specific characteristics of rebellion in Burundi that apply to groups there more generally.
178 A Political History of the CNDD: The Founding Rebel movemen ts did not play major roles in Burundian politics until the wave of democratization in the late 1980s and early 1990s opened up political space, forcing liberalization although the factors that scholars associate with rebel movements (exile populations, government repression and deprivation) existed before these political changes. The absence of rebel org anization for so long a period of post Independence history largely owes to the historical occurrence of hiera rchal state structures that sustained themselves from the early consolidation of the kingdom preventing the (Tilly 1978:192) While the centralization of the state was more fluid th an elsewhere in the Great Lakes regio n, as discussed in Chapter 3 the Burundi an state remained a powerful entity with direct local authority and control even at the community level. Thus the state apparatus remained both strong enough to crush opposition when necessary, and also exercised social control over the population that allowed for some Hutu to gain positions of power, like the father of Pierre Nkurunziza gaining a seat in Par liament R outes to power were somewhat open although tightly controlled, at least until the 1980s, and because conflict was multi directional (intra and inter ethnic) would be challengers could find some inroads. In 1993, the terms of the scale of viol ence used to communicate political goals increased dramatically assassinations, roving urban death squads and civilian massacres all became common practices af ter the death of Ndadaye in October 1993. Again, while this violence had similarities to other waves of violence, this was a new and surprising occurrence compared to the relative peace and stability of the 1970s and early 1980s. This peace and stability had come at the cost of mass repression against
179 civilians, but still outright state sp onsored violence had been minimal. Rebel groups began to emerge following the assassination of Melchior Ndadaye. S poradic civil violence began in earnest at this time for several reasons: the activity level of rebel groups in the region expanded dramatical ly, especially in neighboring Rwanda, the imposition of multi party democracy by international donors opened up the once severely restricted political space, and increased civil conflict in neighboring countries Uganda had already seen successful refor mist rebels come to power in the mid 1980s, and rumblings of a Tutsi rebel force and civil war in Rwanda (which actively recruited in Burundi) further intensified the political context This created a dangerous environment and changed the ways in which act ions were interpreted by various political players. Nascent groups, whether developed internally or abroad, would no longer be tolerated, as the Palipehutu FNL was tacitly so in the 1980s, and would be met with swift justice Following the assassination of Ndadaye and his entire successive line (Sullivan 2005:75) in early 1994 small bands of insurgents began to attack Tutsi civilians, especially in the countryside where mili tary presence was less concentrated(Ngaruko and Nkurunziza 2000:376 ) Thi s caused counter attacks by civilians and the military, plunging the country into a maelstrom of violence that generated many more insurg ent and rebel movements of Hutu and met with m ilitia and pro government forces of Tutsi The nature of Burundian society history, and previous conflicts made collective group action like rebellion a fractious enterprise regional differences, multiple claims of representation of a single ethnic id entity, and corruption and neo patrimonialism all
180 caused deep divides I discuss later how some groups, despite this history, collaborated at different points during the civil war. Dissenters created t he CNDD FDD out of dissatisfaction with anti reformist attitudes of the Tutsi army and Tutsi government and the murder of Ndadaye. The group also laid claim to a number of historical grievances of oppression and political violen ce as described in Chapter 3 particularly the events of the 1972 genocide. The pol itical arm of the movement was created in exile in Europe in February 1994 from a secret plan of the leaders of the political party FRODEBU ( Front pour la Dmocratie au Burundi the Hutu victors of the 1993 elections formerly headed by Ndadaye ) to develop an armed militia called the Democratic Generation in Burundi ( Gnration Dmocratique du Burundi GEDEBU). While this original plan to supply GEDEBU failed, some leaders of FRODEBU were able to shift the resources behind GEDEBU to a new emerging movement The political movement was officially created and christened the Conseil national pour la dfense de la democratie (CNDD) on November 24, 1994, by Leonard Nyangoma, the former minister of interior and public security under Ndadaye, who became the president of the organization. The Forces pour la dfense de la democratie ( FDD ) not abroad, and many discrepancies exist about the exact location and time of its founding. Officially, the FDD was cr eated during the summer of 1994 in the northern province of Cibitoke (Ninodera 2008: 106 and Party Records, collected November 2010 ) although members lay c laim to founding the movement in Kamenge (led by the a northern and l argely Hutu part of Bujumbura 2 Kinama, another northern Hutu part of Bujumbura city 3 Musigati, Bubanza province (the national
181 forest, Kibira) 4 and Bukavu, Eastern Congo (Vorrath 2009 ). 5 The reasoning behind the multiple points of creation may be that v and fought against the Tutsi army until a more cohesive strategy (and name) was devised. Th e CNDD FDD intertwined with the goals, strategies, and organizations of the FRODEBU party V arious levels of articulate d political goals from its conception amongst the branches remained c onnected to FRODEBU especially the branch developed in Bukavu by for mer Government of Burundi military officers trained at the Burundian Military Academy (ICG 2002) 6 Although the Government of Burundi armed forces were almost exclusively Tutsi, there were some Hutu soldiers, especially at the lower levels and combat battalions. Among these Hutu the most significant population possessed an additional linkage to the Gov ernment: Bururi heritage. Even though Bururi Tutsis were favored, Hutu from the same region also received benefits: for example, politician Leonard Nyangoma of FRODEBU and former Government of Burundi officer Jean Bosco Ndayikengurukiye, both of whom would go on to lead the CNDD FDD at later points. Drawing from government troops also gave the creators of the FDD other advantages: better access to weapons, knowledge of government combat styles and rallying points, and battlefield training. Wh ile not all cel ls of the FDD initially enjoyed the same level of access and communication with th e political arm CNDD, by 1995 the major groups managed to unite under the FDD banner By this point in time the group had also picked up significant players from the Palipehu tu FNL, including Hussein Radjabu, Adolphe Nshimirimana, and Evariste Ndayishimiye, who would go on to play important roles in the rebel organization (ICG 2002). The CNDD FDD, according to intervie ws with former combatants, developed a decentralized formal power structure,
182 with decisions routed from several parallel upper echelons of the movement to the combatants who were responsible for the physical acts of violence. According to interviews, individual battalions were given essentially free reign to opera te within their theater of combat, and when ordered to move to another part of the country to shore up support, did so with few linkages to other battalions of the rebel groups and messy communication linkages 7 According to a former commander, the logic o f these moves dealt with the environment in which the group developed: factions claiming to represent the interests of Hutu and FRODEBU emerging across various geographic locales and among very different groups of people from educated politicians in Bujum bura to simple farmers in Bubanza. Finding a cohesive organizational strategy and dispersing it was much more difficult than spreading a cohesive, anti government (and to a lesser extent, anti Tutsi) message 8 This is not to say that the group was unable to accomplish anything militarily: in fact, the group was able to effectively fight skirmishes with the military and coerce the civilian population into support. Dual chains of command not only caused chaos in terms of military order, but also allowed for quick decision making and flexibility. Furthermore, the movement had pockets of success as well as failure in different geographic areas. The FDD advanced before the CNDD in multiple different regions under an birth, va rious factions that emerged did not share horizontal or equal responsibilities (meaning one faction could hold power over another despite similarities in size and command) and did not have much contact with each other until later in the war; this inability to access other cells would actually help f urther civilian relationships, phenomena discussed in Chapter 5 To further compound
183 difference between cells, various regions had differing and unequal command structur es and ranks (Ninodera 2008: 108). The presence of sub ethnic loyalties (most of the early leadership was also from Bururi and most of the fighters were not) and differing strategies (where to base territories, war profiteering in Eastern Congo) increased decentralization and eventually fr actionalization of the movement. The combined CNDD FDD self described as a party that institutionally followed, at least formally, major ity rule and sought cooperation but did so through pressuring informal institutions like kinship networks and regional a ffiliation. Thus while claiming to embrace democracy, the movement built its formal institutions on the back of informal fragmented loyaltie s. 9 Membership in the CNDD FDD initially consisted of voluntary combatan ts and participants. As the civil war expl oded, forced recruitment became more and more common, especially among cells led by the so control the Burundian state security apparatus and after 1999 10 The movement open ed itself to all ethnicities, 11 although bec ause of connections to FRODEBU and geographical location in mostly Hutu dominant provinces, most members were Hutu. W hile Tutsi members gained some positions of power in the political arm of the movement, leadership in the FDD was completely closed to Tuts i because of both lack of members and lack of opportunity and, in fact, had not one Tutsi officer at the end of the war. Organization and Rebel Behavior by Period To help situate the CNDD Figure 4 1 pr ovide s a temporal diagram of the war, peace accords, and periods of rebel
184 behavior discussed further following In the appendices (A and B), I also provide a detail of the important dates in the development of the CNDD FDD t o provide overview and context a nd list the important actors in the conflict providing further clarity. 1993 1996 meaning that while tensions were high and the leaders of the movement had taken to exile, interethnic violence was common, and battles, skirmishes, and massacres brought war to the northern provinces and Bujumbura city. the center of town and moved towards the homes, especially at the Kamenge (Hutu) Cibitoke (Tutsi) 12 The violence in Bujumbura city fell directly like the ones described above. Armed insurgency also had a tendency to spiral out from the Nati onal University, where students protested the killing of Ndadaye were quickly met with retaliatory violence by both Tutsi students as well as Tutsi civilians living in nearby Nyakibega and Ngagara ( La Renoveau December 1, 1993). Protests that often turned (dead cities) days, where the entire city of Bujumbura ground to halt: markets were shuttered, offices and schools closed, a nd politicians refused to work. These occurred almost immediately after led by the Tutsi civilian population, to keep Tutsi off the street while the army did the d imprisonment of many Hutu youth, many of whom almost immediately turned rebel upon release.
185 It was also during this time that the Tutsi state army perpetuated civilian massa cres, and to a lesser extent violence against civilians by the CNDD FDD and FNL fighting together. CNDD FDD combatant. 13 While the overall goal of the CNDD FDD and FNL coincided, the tactics, strategies, and push to rebellion did not, conditions which made some, es pecially in the FNL, resentful. As described above, the FNL also faced the problem of defectors who then joined the CNDD FDD. This period culminated in the massacre of almost 400 Tutsi civilians by the CNDD FDD at a displaced persons camp in Bugendana in J uly 1996. This was the single largest massacre of civilians at the time, and transitional government, with former President Pierre Buyoya once again gaining presidential power in the name of protection and security, along with mass repression and violence against Hutu civilians. It was also at this time that what Tilly called emerged feed its functionaries, honor its symbols, give time to its service, or yield other resources [to another actor] despite the prohibitions of a still existing governme nt they formerly obeyed" (1978: 192). The rebel organizations CNDD FDD and Palipehutu FNL both used these resources (civilians, materials from civilians or materials formerly belonging to the state) to build legitimacy, strengthen their fighting numbers, and bolster territorial control, although the CNDD FDD enjoyed much more marked success in this endeavor than the Pa lipehutu FNL. The discrepancy in resource between the two groups dealt almost exclusively with the history and nature of the rebel organizations. Palipehutu FNL was created in exile, and the largest number of its supporters still lived in Tanzania
186 (Mallki 1995:130). Within Burundi itself, the movement only had limited support in certain geographic areas, whereas the CNDD FDD was able to pull members and community support from across the country. CNDD FDD almost immediately set up parallel (to the already ex isting Burundian state) administrations and local structures to provide support, protection, and relationships with the civilian population 14 (Longman 1998). These actions not only included the parallel administrations between the formal government and the CNDD FDD, but also the organization of forest courts and the collection of tributes and taxes. There also existed a band wagoning effect, wh erein which the rebel organization that seemed most likely to actually compete with the Buyoya government became the most popular. Thus CNDD FDD successes in both minor battles, especially in the Northern territories as well as success in bringing the tran sitional and Buyoya governments to the negotiating table provided the group with concrete victories that the population used in their assessments of which rebel group to support or whether or not the rebel organization ultimately succeeded against the stat e. Negotiations and Ending the Violence I turn now to one of the ways in which the political arm of the CNDD FDD was able to achieve inroads its quest for power. A regional peace initiative, spearheaded by President Museveni of Uganda and Nyerere of Tanz ania first came about in late 1995 (Crisis Group, Finalizing the Peace), although this would later be increased dramatically in scope and aim. The Mwanza Accords (the precursor to the Arusha Accords, also held in Western Tanzania) began on June 25, 1996 ( although they would only be symbolic with minimal participation by crucial players until two years later), following pressure from the international community to prevent another round of genocide in the Great Lakes 15 These negotiations also included new
187 Af rican political actors: the heads of state and dignitaries from Great Lakes neighbors, like Julius Nyerere, former president of Tanzania. The reasoning behind this involvement was more than the prevention of genocide, however; neighboring countries, especi ally Tanzania, the eastern Democratic Republic of Congo, and Rwanda, had long provided bases and, in some cases, material and symbolic support to the CNDD FDD, the PALIPEHUTU FNL and other rebel movements. This support and the concurrent existence of mill ions of Burundian refugees in the region caused stress on the already resource poor governments of these countries. This was especially the case in Western Tanzania, which had supported hundreds of thousands of Burundian Hutu refugees since 1972. In Easter n Congo, refugees had a tremendous destabilizing effect on local and regional politics, and the refugee population in Rwanda during the early part of the war directly contributed to the outbreak of genocide there (Stam 2009 ). 16 The first round of the Arush a Accords on Burundi led to an embargo by regional players, beginning in July 1996, in the hopes of forcing the Burundian government to negotiate with the rebels to a peaceful settlement. 1996 1998 During this period of time, life in Burundi, although ma de difficult by the imposition of regional economic sanctions, returned to some degree of normalcy for the civilians. With the government and the CNDD FDD in the midst of peace talks and the fighters of the CNDD FDD occupied fighting in the rapidly escalat ing civil war in Eastern Congo, violence inside Burundi declined. This does not suggest that massacres did not occur: major incidences of violence against Tutsi civilians by the CNDD FDD happened fairly often, as well as retaliatory violence against the Hu tu populace. But overall, the
188 chaos of the 1994 1996 ebbed, according to accounts published in the state run newspaper Le Reneveau Regional violence at the time increased markedly: Rwanda and the RPF invaded Zaire, backing the rebel AFDL in the hopes of b ringing down security dilemma that plagued their northern territory and could potentially topple the new post of porous borders with Congo, ordinary citizens long used Congo as a place to run in times of trouble, such as during the 1972 genocide. Rebel organizations took advantage of these as well, with the CNDD FDD using the eastern portion of the country, especially those areas sharing the Ruzizi river valley. But this came with a new set of problems. The CNDD FDD had used Eastern Congo as a supply and rear base since the civil war broke out, and now these bases and supply lines were severely lim ited by the new war. So although the rebel organization could gain new allies: the Mai Mai (non aligned indigenous fighters in Congo) and the FLDR (Forces Democratiques de Liberation du Rwanda, the remnants of the 1994 Rwandan government representing Rwand an Hutu interests), other potential benefits were lost. Tentative Steps to Agreement The political faction of the CN D D FDD took a step towards resolution of the civil a non state, non negotiating force), were thought to remove many of the barriers to peace that the regional and UN sponsored talks had because they were conducted by an outside force with n o suspected loyalties to any parties involved that would cloud their judgment
189 (Sriram and Wermester 2003: 339). President Buyoya preferred these talks for a number of reasons, the most important being that the talks were not subject to the sanctions impose Egidio carried weight amongst the heavily Catholic Burundian population, the group was unable to leverage any powers of sanction or agreement enforcement behind these opinions. The talks las ted until they broke down with no agreement between parties in May of 1997. But during this time, security in Burundi improved markedly, because a large number of fighters f rom the CNDD FDD had gone with C ommander Jean Bosco Ndayiken gu rukiye to Eastern Co ngo to fight alongside the AFDL under Laurent Kabila and elements of the RPF against the vestiges of the FAZ ( Force Armees Zairois the former Zairian Army). This division of resources and fighters amongst the CNDD FDD meant that battles and violence agai nst civilians diminished in Burundi itself along with the political maneuverings of CNDD FDD in Rome to gain political power and end violent conflict. Newspapers and public reporting in Burundi at this time gave the talks more power than they actually had, because everyday violence lessened. The advent of peace talks were both an internal and external process. Some political actors in Burundi, seeing the writing on the walls and not wishing to prolong violence any longer than necessary saw them as a necessa ry and elegant way to end the war and gain genocide. This was especially true in the highest levels of decision makers of the CNDD FDD, 17 who saw both their own safety as well as the safety of the movement put in jeopardy by continued conflict. Any organizational power or coherence gained during the early part of the civil war was severely strained by both the decisions to engage in
190 differ ent aspects of Burundian politics: the domestic in peace talks and the international in Congo. While the CNDD FDD consumed itself with these tasks, the Burundian government took advantage of the situation and tried to break linkages between the population and the rebel organization, spreading propaganda and providing so called The camp conditions lacked for basic food, shelter and security. Camps were also divided by et hnicity, either because of civilian or government choice, and camps with largely Tutsi population enjoyed far better security because of the existence of the civilian auxiliary units young men trained by the Burundian army, provided uniforms and weapons, and stipends to encourage their new duties. 18 The Break Up, Part One The dual nature of the CNDD FDD had both positive and negative consequences for the future of the movement. After four years of squabbling over the role of political and military leaders hip between the CNDD leadership and battle hardened commanders, things finally came to a head in May of 1998. On one side, political leaders under Leonard Nyangoma wanted power, prestige, and political influence, and were willing to negotiate to achieve th is even going so far as to allow the government of Egidio talks. On the other hand, military commanders wanted to thoroughly rout the Tutsi army and take power through a total military victory, storming the capital and the halls of power. The FDD militants also felt that the CNDD were willing to give too many things away in the pursuit of political power, and they felt there had been far too many deaths in the name of just ice for Ndadaye to let that happen. The FDD also gained first
191 hand experience in how exactly a rebel victory could facilitate power: fighters observed first hand at the battlefields in Congo just what kind of impact a rebel group could have, even in a coun try much larger than Burundi. In fact, Nyangoma had been accused of a month later by those participating in the Arusha process as terrorists The CNDD FDD in Congo concurrently switched al legiances, now supporting the AFDL government, who took power from Mobutu in 1997 and turned against its Rwandan backers, which allowed for the military wing FDD to gain more power and resources (ICG 2002:8). 1998 2001 During this time, the life of CNDD FDD was dominated by discussions over the future of the group and competing influences that pushed the group into different directions. With resources split amongst Eastern Congo and Burundi, fighting two wars, both the civil one in B urundi (1993 2005) an d the Second Congo War (1998 2006) seemed as though it would become daunting. Furthermore, relations with other formerly cooperative rebel groups like the FNL became strained at this time and led to intra rebel fighting in addition to rebel government batt les. Most of the battle combatants and even political members of the Nyangoma CNDD FDD left at the 1998 split to follow Jean Bosco Ndayikengurukiye, the military commander who helped to found the FDD and overall, the rebel organization intensified violence against the Buyoya government. The change in strength and power of the CNDD FDD caused international actors to lobby harder for peace negotiations, which continued at both Arusha and in South Africa. Ci vilians also played an important
192 role at this time, providing more support to CNDD FDD (some of which was forced) and losing confidence in the Buyoya government and army. The Break Up Part Two A group that has already split once can easily repeat its hi story of splitting, and a mere three years later the CNDD FDD did exactly that. In 2001, Hussein Radjabu, the number two in the organization and considered the highest ranked political officer 19 and Pierre Nkurunziza (a political officer with some battlefie ld experience), with the support of various generals in the FDD, rose up against Jean Bosco Ndayikengurukiye in a coup surprisingly similar to the historical coups of the FAB ( Force Armees Burundais) For one, Jean Bosco, the nephew of Leonard Nyangoma, wa s from Bururi province (a southern province that produced the first three Presidents of Burundi) and most of the wer structures. This was especially noted in regards to violence against civilians, who felt that southern soldiers counterparts. 20 Combatants accused Jean Bosco of orderi ng the murders of non Bururian soldiers and officers in an attempt to limit competition and maintain hegemony 21 The previous years (1998 early 2001) since the sp lit from Nyangoma experienced a marked change in intra group loyalty and cohesiveness This wa s due to two factors: one, the immediate distinction between the armed group FDD, now involved in two wars, and the political group CNDD whose only responsibility was to develop resources in Burundi, and two, the anti Bururi sentiment rising against Ndayik engurukiye and his regional kin. A stretching of material resources (guns, money,
193 and food supplies) caused serious rifts between politicians and fighters because the soldiers in Congo felt their resources were being wasted in Burundi and vice versa 22 New actors also rose to power. It was at this time that the power of Hussein Radjabu came into full effect he effectively court ed aid for the movement from African countries with Muslim leaders and populations, funnel ed material through Eastern Congo to suppo rt the war effort both at home in Burundi as well as Congo, and increase d his own power within CNDD FDD. With this new power, Radjabu was able to effectively place Peter Nkurunziza, a previous ly lower ranking officer, into the highest level of command as P resident and representative of the rebel organization and further collaborate with him for the future of the movement. Until this point, Peter Nkurunziza had been a popular, well liked, but not well tau ght physical education. Because of all these factors, the movement factioned off again, into the CNDD FDD led by Jean Bosco Ndayikengurukiye and the CNDD FDD led by Pierre Nkurunziza 23 The formerly united group was now split into two, with both factions claiming to represent CNDD FDD interests, although the government of Burundi rightly saw the Nkurunziza faction as the bigger and most important threat. After the split, combatants numbered approximately 15,000 in the CNDD FDD (Nkurunziza) and approxima tely 1500 in the CNDD FDD(Ndayikengurukiye), ensuring that the latter rebel organization a very weak position in the peace negotiations and relatively little prospect for post war governance (ICG 2002). 2001 2003 T he crux of the problem with the rebel movement emerges: it was never clear which side of the group (CNDD or FDD) reigned over the other until this later period At
194 the beginning of the war, serious questions arose about the role each arm would play. While Nyangoma and his supporters clearly th ought the political wing supreme, many commanders and leaders in the army thought otherwise. Many interviewees at the lower levels of the armed forces mentioned dual commands both in terms of a lack of hierarchy within the armed forces and between the pol itical and armed wings Foot soldiers generally did not think much about the split because the concerns of the war were of paramount importance 24 and in fact they were not initially trained or given information in political message or methods until there w as a lull in the fighting, around 1995 1997 25 Initially, the tasks of garnering civilian support and education went to the political wing, but over time, the burden shifted to the military wing to earn the trust and loyalty of the population in order to pr ovide support for the entire movement. This was because of increasingly blurred lines between political and military actors of the CNDD FDD as the war continued on, with clear distinctio ns only really known to elites. As fate would have it, Burundian polit ics settled the struggle for CNDD FDD identity, for concurrent with the split between Nkurunziza and Ndayikengurukiye was the overall several times by the CNDD FDD enticed ci vilian support. Even though the Buyoya government waned, the CNDD FDD was unable to win the capital: this led to increased participation the Arusha Accords and the Pretoria (South Africa) talks, leading to a ceasefire agreement between battle participants in December 2002, although this would not be fully implemented (meaning that fighting continued) until late 2003.
195 2003 2005 T he rebel organization slowly turned into a post war political actor at this time, and the focus shifted from fighting military ba ttles to fighting political ones. This occurred within an environment with differing levels of organization and strategy among actors within the rebel movement. The differences in organizational strategy and structure between CNDD and FDD became critically apparent, especially as the political wing purported to give up arms while battles still raged in the hinterlands The initial ceasefire signed by Nkurunziza and the political actors in 2002 led to much fanfare in the international arena, but did not actu ally change the conditions of battle still active between the combatants and the army. The CNDD FDD and FAB continued to fight skirmishes well into October 2003 (ACLED 2011). The dual chains of command so prevalent in the conduct of the civil war up to thi s point finally crystallized. Political actors would alternatively threaten the transitional government with a return to combat or jockey for post war government positions. Military commanders, on the other hand, still wanted a decisive victory 26 Battalion s remained badly organized, with ongoing battalion in the South was known for being especially brutal in enforcing its will) and without a central strategy. Political actors in the CNDD side of the organization did have a better sense of coherency and strategy going forward: they began issuing party proclamations and soliciting funds from the population. However, internal cohesion between the two arms did not fully coalesce until th e transitional government accepted CNDD FDD political leaders. This occurred most spectacularly with the appointing of Nkurunziza in late 1993. His appointment and continued support by both the armed and
196 political wings united the new political organizatio n (they shed the rebel label with the ceasefire) to face the first post conflict election of 2005. How did Organization change the outcome of the war? Because there was no defining central authority and the two branches of the CNDD FDD o ften took actions counter to the goals of the other, the actual conduct of the war be come frightfully messy. This disorder prevented either branch f rom ever cleanly achieving its goals the CNDD of negotiated political success a la the FRODEBU victory of 1993 or the FDD of a military victory and power similar to the coups of the past. Thus, the outcome of the war was greatly influenced by the fragmented nature of the organization of the group. The group was unable to exercise infrastructural power over the course of its lif espan, and rather expressed cellular power in its organization. Had clearer coordination existed amongst all branches, a more cohesive strategy could have been undertaken for how to deal with the Burundian government militarily and politically. By this I m ean that if the CNDD FDD had followed a more elite driven and tightly delineated form of command, like the RPF or NRA, decisions would have been implemented and information spread in a more uniform manner, potentially presenting the Government of Burundi w ith a stronger, more disciplined foe. Instead, internal politics, lack of communication and effective goal setting in addition to a pervasive enemy both home and abroad made for an indecisive, indeterminate war. To compound the issue, the war itself went through periods of starts and stops which contributed to further encouraging the CNDD FDD to remain disorganized. By this mean that the very nature of the Burundian civil war as a temporally long but low intensity conflict never forced the group into defin ing itself by a
197 grand battle or sustained territorial push against the FAB and Buyoya governments. Additionally, the engagement of all parties in peace talks at several points during the conflict meant that the abrupt ending of the war remained a serious c oncern to leadership of the CNDD FDD. For example, the early part of the war seemed to draw the organization together, but the middle years (1996 1998) and a competition over resources between fighting the war in Burundi and aiding allies abroad in Eastern were split. The last years of the war saw a shoring up of organizational power for the armed and political wings, but also a rise in violence against civilians, foreshado wing post conflict behaviors. Learning Strategy T he question arises of from whom the CNDD FDD learn ed and how this knowledge affected the outcome of conf lict. The role of the Palipehutu FNL must be attributed most immediately. From the onset of conflict former members of the Palipehutu FNL joined the Hutu militias that became the FDD en masse, and some specialists suggest that it was in fact Palipehutu FNL cadres who infiltrated FRODEBU in the early 1990s, and encouraged it to militarize against t he gov ernment (Uvin 2009: 12) and committing violence against Tutsi civilians. This is confirmed by personal interviews and some older militants of CNDD FDD, who admitted to first being members of Palipehutu, both in Western Tanzania in the 1980s and in Burundi i n the 1990s 27 before joining CNDD FDD at the time of the assassination of Ndadaye. As explicated above, initial Palipehutu turned CNDD FDD leaders became critical to the continued success of the rebel movement, even as the rebel organization transitioned in to a government: Hussein Radjabu was the party chairman of CNDD FDD after
198 Nkurunziza became president, Adophe Nsharimimana became the chief of the secret service, and Evariste Ndayishimiye became the private military secretary to President Nkurunziza. Furt hermore, at least during the early days of the war, the two groups fought together against the FAB, sharing tactics and strategic training in addition to weapons and human material. There were also events during the later years of the war where the Palipehutu FNL and the CNDD FDD acted as allies (in 5 incidences of violence against civilians and 3 separate battle incidences against the Government of Burundi army) according to the Armed Conflict Locator Events Dataset ( ACLED) dataset, althoug h by thi s time the CNDD FDD split and was sometimes fighting versions of itself as well as other groups. Similarly, the CNDD FDD joined forces with the FROLINA combatants at the beginning of the war, and learned strategies for fighting in unfamiliar territ ories an d with a foe trained extensively in amphibious and water combat as FROLINA was based in the lake region in the south and the CNDD FDD was more familiar with the hills of the Kabira Forest in the north. This rebel learning nexus continued to work in circul ar movements between the rebel groups in Burundi. But the rebel learning nexus was not stymied by juridicial state borders, and because the CNDD FDD had bases in two other countries, they were able to gain foreign support, training, and aid from other rebe l groups in the area. At various points of time during the Congo War of 1996 2003, the CNDD FDD worked with the Interahamwe FAR, the left over group of genocidaires from 1994 Rwanda, the Mai Mai, local militias fighting overnment, and eventually the RPF against the AFDL. These groups and their strategies, tactics and political maneuverings rubbed off on the CNDD FDD while they interacted, although directly attributable characteristics
199 are difficult to pin down This is be cause so many members of either rebel organization considered membership fluid, and thus participated in both. The Palipehutu strategy of maintaining bases in outside countries did leave a mark, however, and the CNDD FDD quickly adopted this strategy for t heir own. After exploring the historical behaviors and outcomes of the civil war, I now turn to the post rebel behavior of the CNDD FDD, in specific focusing on how the previous mismanagement in organizational power continued to prove problematic for the CNDD FDD political party and government. The Election 2005: How did Organization Shape the Outcome? The first elections in a decade and a half encouraged a hopeful population in Burundi in the summer of 2005. Although there was not to be a direct Presid ential election, ci tizens would nonetheless voice their support for the party of their choosing. The rebels, fresh from the bush, still carried the banners and symbols of the rebellion. The official party flags, symbols and even ballots of CNDD FDD all con tained the eagle carrying a machete and a cassava leaf. This sent a powerful message to both ile part of the rebellion) had weapons and plans should the CNDD FDD lose the election This group continued to use rumors that served the movement during the rebellion as a way of enticing the population and potentially threatening its enemies. The propa ganda of the movement worked its in favor in the elections, though, and the CNDD FDD won control of both the National Assembly and the Presidency, giving the m a wide base of political power The decentralized structure of the former
200 rebel group now turned party worked for and against the group. Although the military and political wings had now been joined by the party declaration, a power struggle between military posturing and political finesse continued. The CNDD FDD was aided by having multiple party hea dquarters to direct party supporters at the commune, hill, and province levels. Although these levels were supposed to report a central authority, varying levels of obedience and differences in propaganda tactics prevailed, according to election observers. 28 The group was also able to benefit from a non centralized strategy of voter recruitment, which meant that in areas where vote buying was necessary to win, this a strategy undertaken with no repercussions from a central authority either internally from party elites or externally from the National Commission on Elections (CENI). T seventeen provinces, with significant representation in both urban and rural communes. Because of its sheer size a nd nature, the message of CNDD FDD could be spread: The cellular nature of the group also worked against them in the post conflict election period. For one, the group was not able to centralize th e national message and goals of its government, so connecting rural agricultural voters who made up the majority of the population with well educate d, city dwelling voters became difficult To overcome t his, the CNDD FDD enlist ed outsiders, like Jean Marie Ngendahayo who were not origina l to the movement but add ed prestige to it. Ngendahayo, a Tutsi with strong Bujumbura connections, was active in politics since before the civil war, having joined FRODEBU in the early 1990s and appointed Minister of Foreig n Affairs by Ndadaye. Ngendahayo tried to remain part of the transitional government during the
201 civil war, until threats to his life forced him to flee the country (described in detail in Krueger and Krueger 2007). Thus, Ngendahayo made a perfect new face to the CNDD FDD: a former politician, Tutsi, who reflected the ideals and sprit of 1993 the group wanted to recapture. Even with powerful figures like Ngendahayo joining the movement, the problem remained that there was no central figure upon which the gro up could rely on to provide a physical symbol of the power of the movement. Although the rumor that Peter Nkurunziza, the presumptive President, rode to Bujumbura on a bicycle with his soldiers following closely behind, ushering in the peace agreement and cease fire, even this rumor was not enough to sustain a national cult of personality around him at the time 29 It was customary in Burundian culture even in the pre colonial era for cults of personality to be built upon rumors: the story of Samadari, a Hutu folk hero agitator against the reign of the Mwami provided an example for the balancing of Hutu peasant inte rests against wealthy cattle owning royalty. Similarly, rumors spread about Prince ever saw him in person. Smaller cults of personalities arose amongst former commanders, and it became unclear as to whether CNDD FDD supporters were truly su pporting the party or the individual This was especially true in the former stronghold provinces of Bubanza and Cibitoke, where party members sometimes referred to 30 During the war, Manasse Nzobonimpa commanded the region that included the northern provinces for the CNDD FDD, and everyone living under his command was aware of his power and authority, although interviewees expressed a lack of connection bet ween the national CNDD FDD and 31 Similar cults of personality developed around Peter Nkurunziza
202 and General Adolphe Nsharimimana, and it proved difficult to integrate fighters from these groups into a larger national group without the f ormer identity remaining. These challenges between regional and national identities also existed among the political wing and political supporters of the CNDD FDD, who felt split loyalties between the two. 32 These splits were not the result of the civil war but rather historical circumstance of the nature of the Burundian state that the CNDD FDD (and even the Buyoya government before it) inhabited. In this I mean that historical geographic tensions and regionalism, especially between the ancient regime of t he Mwami, Prince Louis Rwangasore and the hero Melchior Ndadaye (Muramvya) and the regime of the Tutsi leadership within the CNDD FDD took this tone as well, with significa nt resentment of those in power from Bururi (i.e. Nyangoma and his nephew Ndayikengurukiye). Thus to briefly return to a point made in Chapter 3 the pre conditions of the Burundian state fundamentally impacted relationships and power structures within the CNDD FDD itself, who found themselves with little alternative to change such structures. Changes in Leadership since 2005 Once in formal power of the CNDD FDD, Nkurunziza pursued strategies of interethnic and interregional cooperation and strengthened t he capacities of the movement and later the state to provide public services free healthcare for children and expectant mothers and free primary education, although these promises were largely symbolic durin g the civil war Nkurunziza, despite his fractio nalization from Nyangoma and Ndayikengurukiye, above all sought institutional stability to promote a cohesive movement, and would eve n sacrifice formal capacity (specifically, military and war
203 ma king capabilities) to gain political power ( Lemarchand 200 9:166). This means that instead of shoring up the CNDD FDD to perhaps defeat the FAB militarily, he would instead focus on building political power and civilian support in order to gain control of the Burundian state later. Nkurunziza proved to be the best of both worlds for the CNDD FDD: he had both the common touch of a man who fought in the trenches as well as political skill in negotiation and power seeking. His pattern of power consolidation continues to the present from the time he took power of the g roup in 2001. This is not to say that the results are positive, however, with one unsatisfied citizen 33 This statement was echoed in interviews I held with Bujumbura city residents, who were non combatants, who expressed that they had perceived the movement to be more bureaucratically and fairly organized than the previous regimes before it assumed power 34 Because of this willingness to trade stability and development for power, the group has experienced some hiccups in leadership since becoming a political party and gaining majority power. I elaborate on these further below Affaire du Radjabu, 2007 Hus sein Radjabu was an important figure in the CNDD FDD since the founding, acting in a hybrid capacity in both the military and political wings, although he was never a combatant 35 himself. Radjabu emerged in Burundian politics in the late 1980s as an ardent political strategist for Palipehutu FNL and worked behind the scenes to engineer FRODEBU and Palipehutu FNL cooperation. When the war broke out, he seamlessly transitioned to skilled politician for the CNDD FDD, especially the arena of foreign
204 policy, whe re his numerous connections to sympathetic hosts (Tanzania) as well as sympathetic suppliers of war material (mostly Arab nations like Libya) made him a priceless asset to the rebel organization. He was long considered the number two political operative to all three eventual leaders of CNDD FDD: Leonard Nyangoma (1994 1998) Jean Bosco Ndayikengurukiye (1998 2001) and Pierre Nkurunziza (2001 2007) Rumors abound in Burundi that Radjabu eventually planned himself to become FDD (he did in 2005 with his election as chairman) and the country following the war 36 there were still some leaders and cadre that were incredibly uncomfortable to his closeness to Muslim regimes in sub Saharan Afr ica especially as the group relied more on Christian evangelism in the later period of the war (around 1998) to spread the political message and provide new soldiers Although these countries and leaders (Libya and Sudan) fundamentally secured the power o f the CNDD FDD armed branch, the thought of a Muslim dominated Burundi (the country is only about ten percent Muslim) frightened the deeply religious Christians, including Pierre Nkurunziza, in the movement. By February 2007, these fears and moves that R adjabu had made towards power consolidation made necessary the calling of a party conference. CNDD FDD to rid themselves of him and increase their own control, again showing tension between the political and armed factions that never left (Watt 2008:198). Getting rid of Radjabu also allowed political elites within the CNDD FDD to ascend to new positions, reinforcing a recurring phenomenon within the organization; the replacement of elites through un democratic means. Shortly thereafter, Radjabu was
205 arrested in April 2007 for inciting an armed group and convicted a year later and sentenced to thirteen years in prison. He currently resides in Mpimba p rison, the most requests placed through the Government were denied three times over the period of time I conducted research. Problems with Legitimacy Corruption, and the Leavers It became very clear over the historical trajectory of the CNDD FDD that members would come and go, and although tolerated among the rank and file, high profile factionalization, dissent, and defection gave cause for concern. As a rebel o rganization, the rumored method for dealing with defection was execution by firing squad, 37 especially for defectors from the FDD. As for the CNDD political wing, few verifiable reports exist of these revenge killings during the civil war, but after the war murders and disappearances of former allies became common (Ghoshal 2010). Two such high profile dissenters emerged in the early post conflict period. After the ending of the civil war, the population surged with hope in the future, especially under the s tewardship of the CNDD FDD Surveys conducted at the time indicated a desire to return to peace and (Uvin 2009). Ethnic conflict had ended and the CNDD FDD welcomed new members, including Tutsi and former PALIPEHUTU supporters, into its ranks. Non FDD fig hters like Jean Marie Ngendahayo, discussed above, and Alice Nzomukunda became leading figures in the post rebel government. In September 2006, Nzomukunda left the party and resigned her post as the second Vice President of the Republic, charging the CNDD FDD with rampant corruption and human rights abuses. She then took a position as First Vice President of the National Assembly, attempting to
206 reform the CNDD FDD from the inside. The party removed Nzomukunda in early 2008, in a pattern similar to that of the removal of Nzobonimpa. Her ousting from CNDD FDD also allowed the government to oust her from the National Assembly, citing her non affiliation. Political observers in country deemed the move to be entirely political in nature to remove a thorn in the FDD) side. 38 This move was entirely in character with the previous behavior of the rebel organization: factionalization and removing one from a position of power on charge s of corruption or anti group sentiment. This pattern of fractionalization and desertion on charges of corruption within the party itself continued long after these first purges. After the ascendancy of the CNDD FDD to political power, the group continued and expanded upon practices of neo patrimonialist and personal rule perfected by cellular organization in the war. Cellular power during the civil war allowed for some leaders of the organization to gain followers by promising security and spoils (because the leaders could promise rewards without actually having to ask the central authority of the CNDD FDD if this was permissible, and then being able to blame the lack of follow through on this same central authority), as long as they remained under that pa rticular commander. Personal Interviews with men formerly of Adolphe 39 both confirm this. 40 But these purges like that of Nzokumunda and others accused of in party corruption and scandal turned those allegations on thei r heads: the opposition began to accuse the group of corruption in the late 2000s. It became clear from international inquiries into revenue streams and budgets that corruption occurred. Nkurunziza used these purges to advance a political
207 observers contend that the strategy and purges of party elites and political rivals are politicking (ICG 2012). These purges definitively point to cracks in the CNDD FDD that are then exacerb ated by those seeking more power to reformulate the party in a more preferable image. In this, I mean that these purges allow for certain members of CNDD FDD to gain more political power at the expense of those removed from the organization, that then stee r the party in a direction away from that promoted by those removed. These charges are not wholly dissimilar from those charged against Leonard Nyangoma and Jean Bosco Ndayikengurukiye during the civil war. Another troubling development was the gradual i ncrease in political violence in the country perpetrated by the CNDD FDD against dissenters and opposition members, which began to occur with alarming regularity in 2007. The Cult of Peter It has long been accepted in the study of African politics that history of post Independence (Patrice Lumumba, Kwame Nkrumah, and Julius Nyerere to name a few) and in Burundi in particular, including the lion izations of Prince Louis Rwagasore and Melchior Ndadaye. It is therefore unsurprising that a cult of personality has developed around Peter Nkurunziza since he came to power in August of 2005. Peter Nkurunziza, a former gym teacher from Ngozi province, enc apsulates many of the virtues that other Burundians value religiousity, devotion to family and country, and acting as a fully fledged member of a community. He is purported to detest the political life in Bujumbura and prefers to spend the weekends engage d in travaux communitaire (Traditionally, all pre colonial Burundians in the kingdom had to do work for the good of
208 the community. Now, it is a seldom enforced law that all citizens must participate in community works projects on Saturday mornings.) with c itizens, wearing his signature Adidas track suit, planting trees, building schools, or touring hospitals. state capacity enlargement projects in the post conflict period, like free primary education and free maternal and child healthcare have gained him many supporters amongst the rural populations, who laud him for these achievements as well as bringing peace to the war torn country. Nkurunziza joined the CNDD FDD in 1995, supposedly mobilized into action by the murders of Hutu students at th e Universite du Burundi, where he was a lecturer. By 1998, he was elected deputy secretary general, a political position, and during the split in 2001, was elected chairman of the group over both the military and political factions when Jean Bosco Ndayiken gurukiye was ousted. He was re elected in 2004 and thought to b e on his way to the presidency Citizens 41 and despite party missteps, usually assign no blame to him for corruption or political intrigue. The cult of Peter worked to consolidate the CNDD FDD in the post conflict period because now citizens have a visual reference and rallying figure. After he was elected, his value as a figurehead and face to CNDD FDD policy and progress grew. This effect often masks the internal disorganization of the post conflict group, and contributes to a short political memory amongst the population. For example, political observers and practitioners in Burundi accuse the CNDD FDD of having an internal political structure called the Counseil du Sage government. While Nkurunziza is part of this group, he is most certainly not the head, and takes orders rather than gives them. 42 Thus citizens see Nkuru nziza as the
209 definitive leader of the organization while in reality he may hold relatively little power. This strategy is in keeping with the cellular nature of the CNDD FDD that emerged when it was a rebel organization: the group can be represented by on e face that holds little to no sway on everyday authority, like the political CNDD during the war at the time of peace talks. Election 2010 and Consolidation of the Regime The election of 2010 provided another opportunity for the post rebel CNDD FDD to entice Burundian citizens to support their ideologies and plans. Pre electoral violence began in late 2009 (HRW 2010:3) with the CNDD FDD inciting members of its youth wing, t he imbonerakure of terror against FNL members in order to coerce silence or a favorable vote. This was in fitting with a previous increase in retaliatory violence between both the CNDD FDD an d the FNL. The election was scheduled for the summer of 2010, with five rounds of voting: Communal, Presidential, Parliamentary, Senatorial, and Collinaire (the lowest level of administration) to take place over a 10 week period from May 23 September 28 N ews outlets and international observers noted the uptick in violence in the months immediately preceding the elections (Ghoshal 2010) but in general hoped for a smoothly run election cycle. The first election, the Communal, was pushed back a day to be cond ucted on May 24 instead of the originally scheduled May 23. The Communal elections took place with little fanfare, but almost immediately both the CNDD FDD and the FNL proclaimed electoral victory. The certification of the results by the national electoral commission on May 25 ensured a modest CNDD FDD victory (about 68% of the posts went to the party, but over 90% in the rural areas) (EISA 2010) with the FNL
210 receiving the second largest percentage, about 14.5%. Allegations of fraud, especially in provinces known to be unfriendly to CNDD FDD, like the FNL stronghold of Bujumbura Rurale, quickly emerged after this initial release of provisional results by the mission issued a statement of support. The Burundian electoral code allowed for redress of potentially fraudulent election results by challenging the CENI results in the Constitutional Court, which could then decide to invalidate and order fresh elections in parts or the entirety of the country (EISA 2010). The Court refused to hear any actions and the CENI refused to redress the issue of new elections themselves (Nduwimana 2010), prompting a boycott of the upcoming Presidential elections by the five original protesting p arties (including the FNL) with eig ht additional parties joining. Nominations for the Presidency closed on June 7, 2010, and only Peter Nkurunziza remained indicating that additional parties were protesting as well, including UPRONA. The protesting organi zation emerged with a name and mission at this time, the ADC Ikibiri ( Alliance Democratique pour Changement Ikibiri), which now had twenty four members. Burundian electoral law was not clear as to how the election, if there was to be one, would be conduct ed with just one candidate standing, but the CENI decided on June 10 to have a vote an up or down vote on Nkurunziza (EISA 2010). This set off a period of tension and politically violence in the form of attacks on party headquarters, grenade attacks on pub lic locations, and assassinations of both ruling party and opposi tion activists (Boshoff 2010). Nkurunziza won the Presidential election on June July 23, 2010, with so me parties once part of the ADC Ikibiri, UPRONA and FRODEBU
211 Nyakuri, returning to take part. CNDD FDD won 81 seats and 78. 5% of the vote, UPRONA won 17seats and 16.5% percent of the vote, and FROBEDU Nyakuri won 5 seats and 5% of the vote. The elections f or Senate were held 5 days later on July 28, 2010, by indirect election by the National Assembly, with the boycott continuing and the CNDD F DD winning 32 of the 34 seats. Finally, the local elections were held September 7, 2010, with the CNDD FDD again win ing most of the seats, the boycott continuing, and noticeably, no marked decrease in political violence. Assassinations, arsons, and attacks on party headquarters continued to be common, although grenade violence decre ased (Boshoff 2010, HRW 2011). This l ed to an increasingly charged political climate marred by confrontational violence between the CNDD FDD and the FNL. Affaire du Manasse, 2011 Yet another faction began to emerge in early spring of 2011, following the tumult of the 2010 elections and subsequent political intrigue. What is most fascinating about this split is that it came from a highly regarded and well connected CNDD FDD insider, the former commander of the Western region during the war, Manasse Nzobonimpa, a crucial member of the party in the northwest. Nzobonimpa had been leader of the Conseil du Sage and a trusted compatriot of r, in his home region of Bubanza he had a large following amongst civilians 43 and had embarked on a job creation program after the war. But following the election of 2010, Nzobonimpa began to make waves about a particular loan that had transpired between t he NRA government in Uganda and the CNDD FDD during the rebel period that was to be repaid through aid (in the form of school materials and money for demobilization programs) from the Ugandan government to the Burundian one in 2008. The money trail was con voluted, and it was
212 clear that corruption had occurred. Furthermore, this touched a nerve with citizens, many of whom did not receive the promised amounts for demobilizing out of the CNDD FDD or FNL (or any amount at all) and were having problems gaining e mployment and providing for themselves. 44 In March 2011, Nzobonimpa was removed from his role as secretary general of the Conseil du Sage and forced out of the party. He then sought exile in Arusha, Tanzania, where he had been living as part of the Burundi an representation to the East Afric an Legislative Assembly (EALA). Because of this unceremonious removal from the party and the palace intrigue that surrounded it, a number of former CNDD FDD participants have left the party and joined with Manasse. The gr oup is not yet militarized, although observers warn that this is a possibility in the very near future, as Burundian politics more generally grow more contentious. This new faction shows the inability of the CNDD FDD to direct the message and create intern al cohesion that prevents factioning and creates a stronger front against the opposition. In this I mean that the splits that the CNDD FDD have undergone throughout its history, first as a rebel movement and now as a political organization, keep the moveme nt from developing stronger internal cohesion (because various entities within the movement are always fighting for control and maneuvering in case of potential or real splits) it could then use to present a stronger movement. To refer to a point made prev iously and elaborated later, these factions kept the rebel organization from providing a strong enough opponent to defeat the FAB during the civil war: subsequently, these factions kept the CNDD FDD from being able to effectively eliminate either PALIPEHUT U FNL or ADC Ikibiri challenges in the post war era. The decentralized centers of the movement also encourage party followers to leave with leaders, instead of remaining in
213 the larger movement. This is part and parcel of the Great Lakes phenomenon of local loyalties, which will be d iscussed further in Chapter 5. This incident also directly relates back to the CNDD FDD strategy of fragmentation through accusation, especially of the ebel organization was unable to manage competing personal interests during the rebellion, and thus continued to be ruled by personal regimes as a political party and as a government. Also, because of the separation of rule instilled during the rebel period wherein which multiple points of sovereignty emerged within the organization itself, fragmentation occurred along these lines as well, with some choosing to follow Nzobonimpa and others when they were removed from the CNDD FDD. Defining the Variable : Org anizational Power It is useful at this p oint to consider the type of organizational power of the group before and after the war across categories to paint a more complete picture of the ship and control. In Chapter 2 I discussed the differe nce between infrastructural and cellular power. In the previous sections, I provided examples of the ways in which the CNDD FDD exhibited cellular power as a rebel organization, a transitional political actor, and finally as a government. I now delve into specific facets of organizational power that the group displayed. Leadership, Factions, and Splits From the brief examples given above, it seems clear that the CNDD FDD suffered crises in leadership throughout its history, and that power was never surely in the hands of one person or even one branch of the group when in the bush. Furthermore, rival personal branches rose up and competed internally for greater
214 positions of power, through violence (like that of Adolphe Nsharimimana) or personal and regional loyalty (like that of Manasse Nzobonimpa). This problem seems more apparent now that the group has become a ruling party, and determining who the captain of the ship has become even more challenging. This is confirmed by outside observations and interview s within the organization that indicate dual faces or centers of This fragmented nature causes confusion in the population as to who to blame, applaud, or support. 45 Altho ugh the CNDD and the FDD branches both had separate leaders, it was always unclear as to where the cadre thought the leaders fit into the hierarchy of the group: political supporters naturally assumed that the CNDD political wing had supremacy and vice ver sa for the FDD military wing. Furthermore, post war roles were never clearly explicated, thus the party experiences something of a split personality, with some willing to accept authoritarian type command and disciplinary structures similar to that of the FDD (Nindorera 2008: 119). This is echoed again in the largesse of power distributed to prime members of the FDD after the war: specifically, the awarding ensures that fragme ntation and possible fractionalization of the movement would be curtailed by either the threat or use of violence, of which the CNDD FDD now has a monopoly of the legitimate exercise of such force. Communications Communications between different battalio ns, the CNDD and the FDD, and from general to foot soldier were often quite messy during the war. The first problem was that
215 of resources of communication cellular phones were not available before the turn of the century, and satellite phones were expensi ve and difficult to move quickly. Although the movement did rely on the use of some satellite phones, it was mostly among the high command in Kibira Forest and not distributed country wide. 46 Battalions resorted to short notes and tracts pass ed between eac h other, often using civilians to carry them 47 M ore often than not, communications were less than effective and led to many problems inside the m ovement. For example, take the period of time when the CNDD FDD went to assist various rebel groups and governm ents in Eastern Congo during the Great Congo War fro m 1996 2003: When the group decided to split resources b etween Burundi and Congo in 1996 1997 one battalion was to go to Congo. Before doing so, they hid their extensive collection of weapons in a cache in the Ruzizi river valley, on the Burundian side, in Cibitoke province, hoping to re supply and gain more while in Congo. While this battalion was in Congo, another battalion came upon the weapons cache and, assuming they were government or Palipehutu FNL weapons, took them as loot. The first group returned, assuming that their weapons had been stolen. Several weeks later when communication was finally established between the two battalions, the story of the weapons cache discovery was told by the first gr oup to the secon d group, who expressed surprise that their comrades had stolen their weapons 48 This example illustrates the utter lack of organization between the battalions, and inability to share even basic information amongst the group. This became a pr essing issue during battles, when information, especially on allies, positioning, and potential threats, was crucial to the conduct (and winning) of the war. Mixed messages also proved problematic within the chain of command of the organization: commanders would give
216 messages to sub commanders, and a game of telephone would ensue. The geographic distribution of the group also contributed to this issue: because the centers of command were in three places during the war (Kibira Forest (Bubanza/Cibitoke), near Bukavu, DRC, and in Western Tanzania), messages were incredibly difficult to coordinate, and sometimes, the group closest to the battalion would not be the one issuing orders to the battalion. This led to various units of CNDD FDD acting in various tacti cs of battle counter to each other (discussed below) as communication became faulty or ineffective. Funding The CNDD FDD undertook v arious strategies of funding during their development and while fighting the civil war. In the few territorial areas under their control, various battalions of the FDD utilized methods of looting and asking civilians for contributions of food, money, firewood, and human labor. There was supposed to be a central taxing authority that directed the gathering of the cotisation, or subscription fees, the civilians. 49 However, this was more grand theory than practice. Although certain areas where CNDD FDD presence was stronger and the battles w ere more immediate (Bubanza, Cibito ke, and in the south, Rutana and Makamba ) who could impose and collect the subscription, returning revenue to a central authority was spotty at best and often disappeared up the chain of command from the soldiers to the g enerals. The FDD armed wing was also able to secure funding and safe haven by courting foreign support President Mobutu initially provided aid and bases to the rebels in Congo in 1996 U pon his demise the rebel organization lost its bases there, and stru ggled in its war against the FAB. The AFDL initially provided no comfort or aid to the group, only
217 later did the AFDL allow the CNDD FDD to use Congo as a rear base when the CNDD FDD agreed to aid the AFDL against the Rwandan and Ugandan government forces (the second Congo War, 1998 2003). Funding of the CNDD political wing occurred along many of the same lines, with financial stability. Especially critical to this was H ussein Radjabu, whose tireless fundraising amongst other African groups and governments permitted the group as a whole to build a war chest. The lack of centralization internal to the funding operations prevented this from ever growing much past subsistenc e however. Further compounding the problem of general funding, there was the problem of individual enrichment at the expense of the group this was especially noted during the time of Jean r circle profited immensely from their Congolese adventure as part of the FDD armed wing ( Nindorera 2008: 113). This later led to his ousting and an attempt at reforming financial procedures intent on finally uniting the CNDD and FDD bodies in theory and pr actice under Peter Nkurunziza in the later war period (2001 2003). Although the collection and distribution of the funding of the group is a matter of organization, sources and development of these relationships is part and parcel of the source of social power the movement experienced vis vis civilians, and thus I retu rn to this topic in Chapter 5 I say this because these relationships require more than just the point of the gun to develop: stealing and forced tax collection are tactics of a rebel orga nization directly related to their institutional capacity to make themselves known as a physical threat. But, more importantly in this case, the CNDD FDD had to develop voluntary support for
218 both immediate (supporting the war effort) and long term (winning government power) goals. Tactics of Battle Although the CNDD FDD fashioned itself a rebel army in the style of other armed groups and even the FAB ( Force Armees Burundaise the government army under Buyoya ) battle tactics were greatly affected by the l evel of disorganization present in the group. C onfusion ruled when it came to who gave orders and where the CNDD FDD was to go to most effectively fight the enemy ruled the day. In the beginning of the war, battles were haphazard in the best of cases, and it was unclear whether there were to be actual military battles between the groups or if civilians were to be fought against as well. There was no central training force or plan, and the FDD essentially was a rough organization of officers tra ined in the F AB and the Burundian Military Institute before the war and those Although the headquarters existed in Kabira Forest, Bubanza, the process was never streamlined enough to train the numerous fighters who joined, and many left and returned at will. 50 Other battle training camps existed in Congo and Tanzania, especially as the war progressed and routes to Kibira were blocked by the FAB, but these were weakly staffed and supplied. This lack of centralized training bec ame chaotic during battles, as soldiers were unclear as to what actions were to be pursued and which tactics would be employed. This disorganization became even more complicated due to lack of centralization in communications among battalions and cells, wi th no c entral authority to direct. T his stra tegy gave the CNDD FDD a strategic advantage over the FAB in terms of flexibility
219 never allowed for any centralized large sc ale strike against high value targets, like the centers of government and commerce in Bujumbura City. This prevented the group from winning outright, as there was never a clear strategy or path to take to military victory. Political Tactics Like problems of centralization and organization when it ca me to military tactics, similar ones arose when it came to the political tactics of the CNDD FDD. While Nyangoma directed political tactics for the group until his ouster in 1998, there was a clear distinction between the goals of the FDD armed branch and the goals of the CNDD political branch and this became apparent when deciding how the group would deal with other groups and potential allies. This also became a challenge when negotiating with both former sup porter s and sometimes financial backers FRODEBU and the government of Burundi with whom certain elements of the CNDD FDD desired to broker a peace Some observers contend that the CNDD never intended to do anything but win power through negotiation ( Nindo rera 2008: 111 115), and thus deliberately stalled political negotiations at the expense of decisive military victories Which is to say that the political arm of the CNDD FDD, the CNDD never thought military victory an achievable goal, and hence never plan ned for anything but strategic and small victories through political negotiation and transitional government. This ultimately ended up being one of the causes of the removal of Nyangoma in 1998. Again, this leads to the conclusion of dual chains of command and multiple centers of power that did nothing by stifle the CNDD inroads against the Buyoya government. The problem of differing central authority and goals between the CNDD and the FDD once again raised its he ad.
220 Present Day Organization A lthough in complete political control of Burundi, the present day CNDD FDD still suffers from many of the same afflictions that plagued its organizational structure from its inception. There is still no clear division between the armed and political wings of the CNDD FDD, and the lines between war and politics are still very blur ry. Thus, the youth wing of the party, the Imbonerakure, is accused of being a militia for the group, to intimidate and commit violence against wayward civilians (HRW Report 2010). Even though the current Burundian Army is commanded by former leaders of th e CNDD FDD, dual centers of command and authority exist between the youth group, the armed forces, and the security forces (like the Secret Police) in Burundi. This helps the CNDD FDD government by allowing for plausible deniability of violence committed b y the forces. M any of the personnel in the modern party were FDD officers first, and this history is hard to forget for both civilians and the officers themselves. For examp le, when Manasse Nzobonimpa was to be removed from the party, the outcry on both sides was palpable on the one hand, he was a lauded general responsible for a great many battle successes during the war. On the other, he was still seen as a military man, a nd unsophisticated in his approach to politics and democracy. Media reported that this inability to transition from military to political contributed to his poor political performance, including accusations of corruptions 51 There are some in the Burundian media and observer comm unity who accuse the CNDD FDD of still being run as an armed group, with generals in charge instead of political leaders who respect law, order, and non violence 52 (HRW Report 2010) This belief is further influenced by the
221 existence of the so a group of five former military leaders in high positions of power who control the armed and security forces, especially the secret service. 53 Although internal party dynamics are governed by the Counseil du S age (Elder Co uncil) system, in recent years the group has been accused of having dual council systems (public and private), that compete for power within the organization 54 In further confirmation of evidence that emerged about the rebel period, the CNDD FDD appears to be continuing the behavior of multiple points of leadership and sovereignty, with little direction for the party milieu to follow. This internal struggle over organization, decentralized centers of power, and the general lack of transparency are held ove r from the days of the war, and the group has not been able to tra nsition into a government with differing organizational strategies post conflict. The FNL and the CNDD FDD It seems reasonable to suggest that the CNDD FDD and FNL share similarities in cau se and motivation. They came into great power out of the civil war, were considered representative of Hutu interests, and used violence to provoke political change. Even during the civil war, there were many instances of CNDD FDD and FNL collaboration agai nst the FAB, at least in the early days Some evidence suggests that even later when the CNDD FDD split for the second time, these groups return to coll aboration with specific factions like that of Jean Bosco Ndayikengurukiye (ACLED 2011). The Palipehutu FNL (now known as the FNL after the 2008 peace agreement that eliminated ethnic naming in political organizations ) was created in a refugee camp in Western Tanzania in 1980 by Remy Gahutu, to vocalize Hutu dissatisfaction with political and social discrimi nation in Burundi, the 1972 genocide, and the lack of ethnic
222 parity. From its founding the Palipehutu FNL expressed this message with force and violence, and during the 1980s conducted many cross border raids from their bases amongst 1972 refugees in camps in Tanzania. They are thought to be responsible for the attacks on civilians in the late 1980s and early 1990s at Marangara and Ntega. While the Tanzanian government was not a direct supporter of the movement, there is plenty of evidence to suggest that they had knowledge of the movement and their use of Tanzanian soil as a base (Longman 1998: 49) Researchers identified the existence of political parties organizing within the refugee camps and their potential abilities to impact politics, especially in regards to the development of an ethnically While an outside actor for most of the 198 0s and 1990s, the death of Ndadaye provided opportunity for the Palipehutu FNL to return to Burundi. Elements of FRODEBU and the fledgling CNDD FDD courted those within the movement trained and active in recruitment of civilians and fighters (like Adolphe Nsharimana) to build an active anti government force. The Palipehutu FNL, however, resisted the CNDD FDD overtures to subsume one organization over the other 55 because although both groups fought against the government, the Palipehutu FNL rejected the multi ethnic character of CNDD FDD. The basic ideology of the Palipehutu FNL focused on the supremacy of the Hutu people and their need to regain power lost at the hands of the Tutsi governments of the past that worked to keep them in subservient positions thro ugh discrimination and lack of opportunity. This ideology found resonance amongst would be members of the CNDD FDD in the early 1990s, especially as political repression and violence ran high and actions of the Tutsi led government to democratize were seen as lacking in follow
223 through. It was no surprise when the two groups began to work together against the common enemy FAB in the early days of the war. Although theoretically the CNDD FDD was to be multi ethnic, the FDD, the branch that the Palipehutu FNL interacted with most closely, was entirely Hutu. This led to radicalization and violence against Tutsi civilians that was often blamed on Palipehutu FNL and their influence within FDD. Palipehutu FNL tended to draw strong support from rural Burundians, e specially Rurale and the southern border provinces with Tanzania. This was because of their development as an external rebel group that had little in common with Burundian political actors living in Bujumbura. For t hat matter, the group initially even had trouble recruiting among civilians because of this outsider status. 56 In the later period of the war as stalemate became the clearer option for all involved parties, t hey later voiced their support for equality amo ngst the ethnic groups during the Arusha Accords process, a tactic directly counter to their previous efforts, which did not go unnoticed by the Buyoya government. Although the Palipehutu FNL and the CNDD FDD shared similar ideas for post conflict governan ce and power, only one could ultimately win, and competition necessarily begat conflict between the two entities. Some jealousy also emerged as a result of competition for military resources and outside support: the CNDD knowledge of actors and other rebel organizations in the region. The Palipehutu FNL receive d some support from other rebel organizations, especially the FDLR ( Forces dmocratiques de libration du Rwanda ) operating in Eastern Congo, but was unable to petition for weapons or bases from allies because of both competition and their comparative lac k of capacity and troop strength.
224 Further, there were periods of time during the war where the two groups stole weapons and material from each other 57 conditions of war that did not lead to a strong friendship. The relationship between CNDD FDD and Palip ehutu FNL can be described as schizophrenic at best, as over the civil war the groups went from cooperating to competing to actively engaging in violence against each other. At times, they actively competed for civilians, killing and torturing those accuse d of joining the other camps, against the common enemy of the FAB. After major hostilities ended between CNDD FDD and the Buyoya government in 2002 2003, the Palipehutu FNL took the opportunity to continue acts of violence against both government targets as well as civilians, including the 2004 massacre of Banyamulenge (Tutsi) Congolese refugees at the Burundian border town of Gatumba, mere kilometers from Bujumbura. Th e continued aggression towards civilians led to the government as well as political actors involved in the negotiation process, like South Africa. An initial peace agr eement was signed by the group and the Nkurunziza government in late 2006, although violence remained ongoing, especially in Bujumbura Rurale province. Direct attacks against the capital continued well into 2008, when grenades and shelling ramped up again st the capital in April and May. The Palipehutu FNL targeted government officials and buildings in an expression of dissatisfaction with the CNDD FDD government, although when questioned, expressed that the violence was mere retaliation against the CNDD FD supporters. By the end of 2008, however, political strong arming by government and
225 international actors led to a final peace agreement where the group dropped Palipehutu from its name in order to participate as a p olitical party (now known solely as the FNL) in the 2010 elections. I now turn to the problems of internal conflict and dissent within the movement. Like the CNDD FDD, the Palipehutu FNL also factioned and split over its history. The first split occurred in 1990, when members of Palipehutu FNL split to form FROLINA (name here), another rebel organization to fight the Buyoya government in the South, led by Joseph Karumba (ICG 2002 : 6 ). This split was driven by fights over leadership and direction the moveme nt would take. Palipehutu FNL continued to launch small attacks against military targets during this time, even while planning peace negotiations with the Buyoya government. Like the CNDD FDD, central control over the organization never emerged in a way th at would silence internal critics, and the movement split between Etienne Karatasi and Cossan Kabura in late 1992. Kabura was seen as more militant and willing to engage in violence than Dr. Karatasi, and thus more willing to achieve the goals of the organ ization. Karatasi and his associates remained at the head of a more political Palipehutu, rather than the Palipehutu FNL of Kabura. The Palipehutu FNL received considerable support and an increase in reputation through participation in the Rwandan genocide where they received military training and support, even returning to bases within Burundi and Tanzania with some ex FAR ( Force Armees Rwandaise ) in tow. This also led to continuing relationships between anti RPF Hutu rebel groups operating in Eastern Co ngo during the Congo Wars, who provided weapons, training, and bases. The lack of planning for the future and what to do after the war led to another split, in late 2001 early 2002, this time between Agathon Rwasa
226 and Cossan Kabura. Rwasa was a military le ader of the organization actively in Bujumbura Rurale who accused Kabura of mismanagement and accepting funds from the Buyoya government to discontinue the rebellion (ibid: 7). Rwasa proved to be a chaotic force to reckon for both the rival CNDD FDD and th e Buyoya and transitional governments, as attacks on civilians and risky bombings of government targets began. Despite this disorder, Rwasa remains in power of the FNL even today, although his role after the 2010 election shifted. After the loss of the com munal elections in May 2010, the FNL elected to boycott the Presidential and all subsequent elections as part of the opposition coalition, the ADC Ikibiri. Although Rwasa was not the spokesperson of the group, he continued to play the part of the major opp osition leader in Burundian politics. Several times after the election rumors persisted of his flight abroad to seek personal safety and protection in Eastern Congo or Tanzania. 58 It was suggested by interviewees, especially those connected to the governme nt, that Rwasa continued to direct the FNL to commit acts of violence against CNDD FDD operatives, especially in Bujumbura Rurale 59 and observers blamed the massacre at bar Chez Les Amis in Gatumba 2011, where approximately 40 civilians were killed, square ly on the FNL (HRW 2011b). This leads to a discussion of the current status of the Palipehutu FNL. Ongoing violence between the CNDD FDD government and the opposition force continues, with nearly daily assassinations in 2010 and 2011. In 2012, the FNL re mains active but the goals and even directly attributable actions remain unclear. What is clear is that Rwasa is still the head of the organization, one which continues its two decade long clash against the CNDD FDD. The group still enjoys marked support in some rural areas,
227 especially Bujumbura Rurale, although its ability to build on this support is incredibly limited. International observers continue to lobby for political dialogue between the opposition and the CNDD FDD, although the CNDD FDD continues to maintain near total political and social power. Conclusion In this chapter examining the organization of the CNDD FDD and its influence on the war and the post conflict outcome, many things become apparent. The first is that within the history of a fairly centralized state, the group emerged as a decentralized and disparate fighting unit. In direct contrast to the levels of organization present in state government, multiple points of authority with little central command emerged to challenge the stat e. These cellular pockets of power ensured that the rebel group could not pursue an overall coordinated strategy to win the civil war militarily outright. Within the history of the country itself, though, emerges a pattern of parallel pockets of leadershi p, a pattern that the CNDD FDD followe d in its trajectory into a government. This disorganization and lack of centralized strategy leads to a movement (and government) that acts in chaotic fashion, and factions and splits became common. The fact that the CNDD FDD government remains in power today is somewhat surprising, given this lack of ability to maintain internal cohesion and discipline. The CNDD FDD also seems to change very little over its history: even as a government with access to prior state str uctures of control and authority, the internal disorganization continues. Communications, tactics, and strategies remain an exercise in confusion and deniability, and rumors abound of dual centers of powers within in the organization. Similarly, the s largest rival, the Palipehutu FNL experienced similar disorder in its history, with
228 the group splitting several times due to differences in strategies and goals among rival leaders. One may assume that Burundian historical development (specifically, the competition for political and social power among the clans and regional rivalries discussed in Chapter 3 ) contributed significantly to the ways in which rebel organizations thought about the type of organizational power the groups would develop. This is t o say that the rebel groups emulated previous political developments in the state by allowing personal and geographic rivalries to determine the trajectory of these rebel movements. While the CNDD FDD and the FNL were never able to wrest control of the sta te during the civil war, the CNDD FDD despite its internal disorganization, exercised considerable more absolute strength, in terms of civilian support (discussed extensively in the next chapter) and combatants than the FNL, and this served them well in wi nning post conflict political power. However, both groups still to this day continue to suffer from lack of cohesion. Tables and Figures Figure 4 1 A Temporal Diagram of the Major Events Burundian Civil War, 1993 2005
229 Notes 1 I use various names for the Palipehutu FNL terminology interchangeably, although after 2009 they are known exclusively as the FNL. 2 Personal Interviews November 12, 2010, November 15,2010 3 Personal Interviews December 3, 2010, December 7,2010 4 Personal Interviews January 21,2011, February 4, 2011 5 Personal Interview November 14,2010 6 Personal Interview, December 3, 2010 7 Pers onal Interviews, April 21, 2011, March 4, 2011, and January 30, 2011 8 Personal Interview March 14, 2011 9 Lemarchand argues that this modus operandi continues today with the CNDD FDD under Nkurunziza formalizing and strengthening the dominant institution s of the state using informal ties. 10 Personal Interview, May 25, 2012
230 11 Researchers confirm this as well see Lemarchand 2009: 151 for greater exploration and suggest that the multi history. 12 P ersonal Interview, November 5, 2011 13 Personal Interview, December 3, 2011 14 Personal Interviews, February 17, 2011, February 19, 2011, and March 2, 2011 15 These accords are discussed in Nindorera 2008:106 109. 16 Lemarchand 2009 discusses the effect these groups have had on the region in great detail: See Chapters 1 and 2 especially. 17 Personal Interview, February 17, 2011 18 Personal Interview, January 29, 2011 19 Personal Interview, May 27, 2011 20 Personal Intervi ews, March 21, 2011 and March 22, 2011 21 Personal Interview, January 17, 2011 22 Personal Interview, February 7, 2011 23 When the group split, both factions retained the name and were only differentiated from by leadership. Civilians were often unclear as to which branch of CNDD FDD was
231 branc h favored southern positions. 24 Personal Interview, November 10, 2010 25 Personal Interview, January 23, 2011 26 Personal Interview April 3, 2011 27 Personal Interview March 10, 2011 and February 23, 2011 28 Personal Interview, April 3, 2011 29 Although this changed fairly rapidly, especially with new developments in the 2005 2007 period. 30 Personal Interview, April 2, 2011 31 Personal Interviews, March 13, 2011 and April 4, 2011 32 Personal Interviews, April 12, 2011 33 Quoted in Lemarchand 2009: 158. 34 Persona l Interview, 23 year old female Buyenzi resident, church secretary, 23 May 2008. She said that it was not until 2007 that this perception she held of the CNDD FDD changed. 35 Personal Interview, February 21, 2011
232 36 Personal Interview, January 18, 2011 37 P ersonal Interview, January 29, 2011 38 Personal Interview, March 14, 2011 39 It is difficult to ascertain whether this was true or not in regards to Manasse Nzobonimpa, for at the time he was in the process of being removed from the party. My translator sugg ested, and I tended to agree, that some former combatants of Nzobonimpa revised these practices at the behest of the regional CNDD FDD to paint an unflattering historical picture. 40 Personal Interviews, March 23, 2011 and October 11, 2011 41 Personal Inter views March and April 2011 42 Personal Interview, June 23, 2010 43 Personal Interview, March 10, 2011 44 Personal Interviews March 17, 2011 and April 14, 2011 45 Personal Interviews, March 14, 2011 and April 4, 2011 46 Personal Interview, December 3, 2010 47 Personal Interview, February 15, 2011 48 Personal Interview, October 13, 2010
233 49 Personal Interviews April, March and February 2011 50 Personal Interview, May 22, 2011 51 Bear in mind that these accusations come from the state run media, and are not generally expressed by the population without significant prodding. 52 Personal Interview June 24, 2010 53 Personal Interview June 24, 2010 54 Personal Interview, June 24, 2010 55 Personal Interview November 21, 2011 56 Personal Interview February 15, 2011 57 Personal Interview November 15, 2011 58 Personal Interviews, February, March and April 2011 59 Personal Interviews, May 1, 2011 and May 5, 2011
234 CHAPTER 5 SOCIAL RELATIONS: HOW TO MAKE OR BREAK A REBELLION Introduction During civil wars, rebel fighters and leaders often consider civilians expendable (Humphreys and Weinstein 2008 support system. However, in recent yea rs the civil wars literature expanded upon the role of the civilian as a valuable resource during war (Humphreys and Weinstein 2006, Kasfir 2005, Mampilly 2011, Weinstein 2007). Rebels realize that civilians can provide much need ed m aterial aid d uring the war ( food, shelter, and bodies for labor or sexual purposes ) Furthermore, rebel groups also recognize the value of enhancing old relationships with civilians and building new ones, creating a base group of supporters if and when they should come to power fol lowing the end of conflict, as the NRA did in Uganda in th e early 1980s (Kasfir 2005: 286 ) R ebel groups can also use targeted or selective violence against civilians in the course of war, for example, rape forced recruitment, or violence aga inst those from a different region, religion, or ethnic group. While relationship s between combatants and adults and combatants and children differ, this chapter will focus on how the CNDD FDD developed and maintained civilian relationships befo re, during, and after the war. The type of social power a rebel group exhibits captures more than just these critical relationships, however, it also denotes the ability of the rebel group to subvert and/or transform pre existing state structures. This goes hand in hand with civilian relations because states, civilians, and new challengers to the state all exist within the same realm, with the same pool of resources, seeking indivisible politi cal power and control over it. This chapter p roceeds in three
235 sections. Th e first section describe s the relation ships that the variable social power catches between the rebel group, civilians, and the state, providing a conceptual foundation. The second section details the history of civilian state relationships in Burundi most specifically, but also in other parts of the Great Lakes to show similarity in pre colonial histories. The third section explains the temporal dimensions of the rebel ves into the changing role of the CNDD FDD over time, and how the PALIPEHUTU FNL has failed to capture the same social power afforded to the CNDD FDD that allowed them to gain post rebel political power. Social Power : Civilians and State Structures Althou gh rebel organizations may not always recognize their value, scholars find relationships between the existing state and its governors, the rebel group, and civilians to be crucial to understanding rebel behavior and civil war more generally. While a wide l iterature exists time (Stam 2009, Longman 1998) scholars find the linkages between these three ac tors (rebels, the state, and the civilian population) and how and why they operate in wars and post conflict. My conceptualization of rebel capacity, however, combines these relationships to present a more cohesive concept, because in the conduct of war and politics, they are intertwined. To unpa ck this, first I discuss why scholars focus on these relationships. The state and civilians have previously existing relations, although these can and may change during the course of the war. When the relationships between rebels and the state and rebels
236 and civilians are examined, the transitive relationship between the stat e and civilians is also shown, described in Figure 5 1 at the end of the chapter. As laid out in the diagram, the relationship that exists before the war betwe en the state and civilia ns also contains the relationships between rebels and the state and rebels and the civilians, closing the triangle an d creating the variable in this research I term social power al to t he nature of social power they employ and is captured by the term. Although the actors are separated in the diagram. It is worthwhile to note that rebel organizations are initially drawn from the civilian population and well integrated within it (Weinstein 2007, Humphreys and Weinstein 2008, Metelis 2009, Mampilly 2011). So while the diagram gives the impression of separation, in truth the linkages between rebel organizations and the civilian population are far more entangled. To discover why and how social power can contain both crucial civilian relationships as well as those with the state, a review of the literature on rebel governance is necess ary. As discussed in Chapter 1 rebels can follow a Tilly like trajectory to create new states and new state str uctures out of wars. In doing so, rebels can conduct themselves as proto states collecting rents and doling out goods and services to their citizens. I do note here, that like Mampilly, the view of rebel groups subsuming the state processes and acting as of the nexus of state, civilian, and rebel relations, reconfiguration of state structures and practices is as worthy of study as the original struct ures and practices developed by rebel organizations. In many cases, this rebel state building is an explicit program put in place in the territory under their physical control for example, the Tamil Tigers in Sri
237 Lanka enacted decades long governance acti vities (Mampilly 2011, ch 3). A rebel group can engage in these activities to replace the state, or at the very least, subvert state activities with their own. But they do not have to create entirely new states separate from their origin; as Pierre Englebe rt (2009) and Jones (2010) point out, separatism is a rare phenomenon in sub Saharan Africa, and most rebel groups do not set out with the intention of splitting from the state they fight against even the SPLA/SPLM did not initially desire separation, but rather a united Sudan. So if rebels intend to change the state, it follows then that they may embark upon their own plans of governance in preparation, even if these plans do not ultimately lead to creating entirely new states. Because civilians are tied to the pre existing state, it follows then that rebel organizations would be able to gain both credible civilian support and control of state structures (local councils, for example) by courting civilian interests during the war. State structures include t hings like the aforementioned local councils, but also justice services, taxation, education, and community services (health and welfare). Traditionally, citizens understand that the government provides these public goods to civilians in exchange for conse nt to enforce laws and monopolize violence. In sub Saharan Africa, governments are often unwilling (because of a prioritization of resources along neo patrimonial relationships) or unable (because of lack of economic resources) to provide these public good s. This in and of itself is a reason many rebel organizations give for going to war against the post colonial state the CNDD FDD, RPF, NRA, and AFDL are no exception. Thus, during a civil war a rebel organization can build civilian relationships in two wa ys: either by supplanting previously existing state structures and goods with their own supporters and improving upon them or by
238 creating them entirely in places where public goods and structures were historically limited. I discuss this below and provide further examples as to how this actually occurs in the course of a civil war. The History of Burundi and Civilians: Conflict, Continuity, Change As examined in Chapter 3 the pre colonial history of Burundi played an important role in the shaping of the modern state and even rebel organizations that developed there (discussed in Chapters 3 and 4 ) More importantly to highlight for the purposes of this chapter, the pre colonial history shaped relationships between citizens and power holders. The kingdom m aintained centralized order and control emanating from the seat of the Mwami (and all political and social power) in Muramvya. Citizen subjects were generally treated well by the Mwami and power structures, and integration of outsiders, intermarriage betwe of Hutu ( wherein which Hutu could acquire property or cattle and move up in social stature ) were far more common here than in the neighboring Rwandan kingdom, similarly organized from the center out. The relation ships between citizens, the princes (Ganwa, as described in Chapter 3 ), and the Mwami and high court shaped the way Burundians historically thought about power and the state. The central kingdom was able to project power over its territory through the use of smaller units of power, although this proved detrimental to the establishment of an authoritarian dominant leadership because regional competition and Ganwa loyalties created multiple points of sovereignty and power to which citizens looked to for gover nance. Nonetheless, these relationships conditioned Burundians to expect some level of centralization in their lives from which authority sprang even if there were challengers to the central authorities It
239 also conditioned them to resp ond with regularity to named authority figures. But the Mwami and kingdom itself we re not based on authoritarian principles as in Rwanda but rather on cooperation T his added another dimension to the way civilians historically viewed relationships in that they expected coop eration, care, and in general, mutually emerged in these societ ies as a political institution that could arbitrate and further more concretely than the initiation cult the coexistence and blosso 2003: 144). T he locus of power did not rest wit h one clan, but rather with four and as such limited the absoluteness of authority there and encouraged both consociational types of governance among clans as well as intra clan loyalty. This intra clan loyalty helped to later encourage regional loyalties amongst civilians and combatants during the war. Because regional loyalties never diminished in their importance during the whole of Burundian history (see for exa mple, the Bururi dominance of post Independence Burundi describe d in Chapter 3 ), regional identity always provided a marker for citizens to rely upon. I t ouched upon this in Chapter 4 remarking on the loyalty accorded to certain leaders as a matter of reg ional affiliation. I return to this below when discussing citizen rebel relationships during the civil war. This existence of loyalties is not to say that pre colonial history was especially rosy and that clans never fought or schemed against one another. In fact, historically the four clans of the Mwami the Bezi, the Batare (the two being from Muramvya and considered people of the plains), the Bataga and the Bambutsa (the two being from Bururi and considered people of the highlands) clashed politically an d socially with each other, but major wars of conquest were not a facet of Burundian history. Peasants followed the direction of their
240 vassal lords in these clashes, and again, these actions contributed to both creating and cementing clan identities and loyalties as well as tempering power in Burundi This was incredibly beneficial to the overall health of state power and civilian life in the country: unlike Rwanda, a culture of cohesive oppression never develo ped, and thus ethnic tensions and other potentially divisive measures on a grand scale did not occur To be more explicit, the existence of multiple reference points for identity and loyalty in Burundi ensured that cleavages, when they did exist, cut acros s multiple dimensions, whereas in Rwanda, the Hutu Tutsi cleavage based on loyalty to the Tutsi Mwami became the dominant one. These tensions between Bururi and Muramvya continued even into the post colonial era, as Tutsi leaders and politicians from Burur i (Presidents Micombero, Bagaza, and Buyoya in addition to Nyangoma and Ndayikengurukiye of CNDD FDD) felt their power was retribution for past discrimination by the plains clans. Citizens still continued to follow clan and regional identities when Indepen dence came, identifying locally first. This pattern of identification reigned even after the birth of the nation of Burundi, as Bururi loyalties determined government careers, educational opportunities, and political advancement. Historically, citizens also placed the seat of local power in the hands of the interpreter, abashingantahe held their status by their respect in the local community and they often served to arti culate local concerns. Although commonly (though not exclusively) Hutu, they were fully recognized within the Burundi political system in a way unknown even adamantly opposed ,D. 2001: 371). These councils provided structure to daily civi lian life under the Mwami, and although were disbanded
241 under colonialism, were still used informally as a place of sage advice and dispute resolution. Because they were composed of townspeople and peasants and not outsiders, their authority went unchalleng ed among the population The CNDD FDD later re created and co opted these informal councils and in some cases, local elders, to create new councils in territories under their control, a way of connecting the past and present to give civilians some form of autonomy during the chaos of the civil war. They are continuing this strategy as the party in power: there are current plans to bring back the abashingantahe as a truth and reconciliation council. The CNDD FDD are widely credited with this return to a more organic and traditional manner of resolution as a way of healing the scars of the war in addition to keeping Burundian history alive. Local power also played a critical role in both colonial and independent Burundi. The colonial government inherited a state that already put in place a system of leadership that allowed for the king to exercise dominion through levels of representatives: the Belgians drew upon this system to create political chiefdoms, prefectures, and administrative zones that reflected the divisions, but not the ethnic makeup, of the system of old. Because of the linkages to the past kingdom, and the ease in which the newly Independent Burundi could be managed this way, Independent governments continued with this system as a way to reinf orce the Tutsi military state under Presidents Micombero, Bagaza, and Buyoya. Local administrative units, like the nyumba cumi (head of a ten house administrative unit), chef du zone (head of the zone, anywhere from 20 households to a small villages), and mayoral and burgomaster positions remained the backbone of political and social life for most Burundians, especially those living in rural areas far removed from the capital and national politics
242 there. These positions were in fact open to all ethnicities although Tutsi dominated the majority of the political landscape after Independence, a condition reinforced by the purge of potential Hutu challengers during the These local administrations played important roles in the live s of civilians whether in war or peacetime. In times of trouble, for example the early ethnic conflicts of 1988 and 1991 in Ntega and Marangara, local administrators were targets of violence because they were seen as instruments of the state and social arb itrators of ethnic relations. But local administrators also played the role of directing civilians to secure and safe environments and acting as the voice of authority. Thus, civilians saw local administration as both a enemy and friend. This would later serve the CNDD FDD well, especially as they began to replace administrators with their own members during the war, co opting this important state resource for their own benefit. Because of the nature of historical development in Burundi and its proximity to other like kingdoms in the Great Lakes that shared similar cultures and influence, a review of the larger regional history al ready discussed in Chapter 3 shows similarities and differences among pre colonial, colonial and post Independence political and social life. I explore this phenomenon again later with the large number of existing rebel groups in t he region (Chapter 6 ) to understand how history shaped the present in this da is integrated 42). These histories become especially critical given the comparisons many scholars make between Burundi and its norther ly
243 (Lemarchand 2009:ch. 1). In the pre colonial era, several kingdoms existed side by side in the Great Lakes region T hose of the Buganda, Rwanda, and Burundi made up the most contiguous th century. S ome peripheral kingdoms were subsumed under larger, more aggressive neighbors, like those of Western Rwanda who joined Rwanda in the eighteenth century, and others were generally peaceful and left to their own devices, like the Ha of Eastern Tanzania. While the kingdoms differ in their levels of structure to the modern state, and a historical legacy of state and soci ety relations that structured how rebellion occurs. Buganda Although not as organized and centralized as either Rwanda or Burundi, the pre colonial kingdom of Buganda provided structure and organization for the clans of what composes modern day Uganda. Referring back to Chapter 3 pre colonial Uganda comprised of many kingdoms with varying structures, most of which existed side by side with little violence or battles for subjugation between them. The largest kingdom, the Bunyoro, never controlled a large enough claim either through territory or population to create a cohesive pre colonial structure, and fell in the eighteenth century. Smaller regional kingdoms provided some structure and governance within proscribed territories, but no wars of conquest or expansion allowed for these kingdoms (the Toro, Nkore, Buzimba, Buhweju, Giseka and Bugesera) until the eighteenth and nineteenth centuries. After Bunyoro fell, the Buganda kingdom was able to become dominant
244 through territorial conquest and protectorate ship, shaping the perception of the British colonial power that Buganda historically dominated others and thus providing an internal power network to rely on. In colonial Uganda, the British first explored the territory through Protestant missionaries, se eking to establish religious networks among the Bugandan and flailing Bunyoro kingdoms. By the 1880s, the government of Britain stepped in and claimed Uganda territorially, with boundaries firmed up by the early 1900s. True consolidation by internal Ugand an kingdoms did not take shape until 1896, with the smaller kingdoms, including Nkore, Toro and Bunyoro, subsumed under the Buganda in moves by both the Bugandan kingdom reinforced by colonial actors trying to create an indirect system of rule. Although t he Bagandans dominated the political landscape during colonialism, the prioritization of the Bagandan ethnic group above all others did not emerge: this is due to British creation of an educated social class that swept across ethnic lines and acted as the British representation in country and a potentially self governing body. This history shaped relationships between civilians and the Ugandan state by providing a template of regionally based loyalties and divisions that ensured no one ethnic group was a ble to effectively dominate the post Independent state, either due to size or political strength. This also meant that the Ugandan state always reflected f ractured goals and identities. Even though British colonial rule developed a new social class based o n education as opposed to ethnicity or region, strong loyalties to region movement, although similar ethnic groups and regional affiliates would support each other acr oss political lines and create alliances. These fractures also led to dysfunction
245 in post Independence politics, specifically in the development of dictatorial regimes based on personalities rather than ethnicity, like that of military commander Idi Amin. The population did play a significant role during the 1981 1986 Bush War, Ugandan civilians suffered torture and terror at the hands of Amin. Even when Obote re claimed power in 1980, the political and social rights of Ugandans were severely limited, and most thought the elections rigged. Ugandan civilians were then willing to provide support to rebels not based on the previously existing state society relations, but rath er based on political alienation and disconnect with government regimes that seemed to care little for them. Rwanda Scholars have devoted much energy to the study of Burundi vis vis Rwanda and vice versa, because the two countries are roughly equal in size, ethnic makeup, two, simply entitled Rwanda and Burundi (Lemarchand 1970), studies in depth pre colonial history, differences and similarities in colonial policies as well as early post Independence governance, finding that while it is useful to compare the two, it may b e unwarranted to focus so heavily on their sameness. In the pre colonial era, the kingdom of Rwanda experienced far more cohesion and oppression than th at of Burundi. In pre colonial Rwanda, as in Burundi, it is thought that the Hutu and Tutsi migrated into the area in the 6 th and 14 th centuries, respectively, to live alongside the native Bantu Twa. But there is a caveat: historiography of the region is generally considered to be myth
246 around the se ventee 13). Even with this caution of over reliance on pre colonial history, a comparison of the two kingdoms yields fruitful insight. A major difference in the two kingdoms was the pre colonial seat of loyalty: whereas as local leaders had far more authority and power in Burundi, in Rwanda, loyalty was to the center. 1 This created more cohesion and less regional friction, but also left civilians less free. Catherine Newbury highlights the client patron system in its effects of re ducing local power and reliance on kinship in favor of the Tutsi elite at the center of the kingdom in Rwanda (1990). This continued even into colonialism and Independent Rwanda, where the state exercised enormous control over the citizenry. Like Burundi, the Rwandan state comprised a centralized hierarchal of administrative units, from the local to the national. This structure helped the Rwandan state to gain explicit knowledge about local communities, including ethnic makeup and social relations. Some hav e posited that this structure lead Genocide in 1994, meaning that this ingrained loyalty to a centralized authority figure caused citizens to follow blindly orders to kill their neighbors (as discusse d in Straus 2006:149 and Collins 1998) During the genocide, the knowledge the hierarchal state had spent decades collecting came to fruition, as community leaders and administrators were able to compose death lists of all Tutsi living within the commune. Administrators had both the tools (the citizens) and the information necessary to rid the country of the five percent. Thus, in Rwanda historical patterns of control dominated as opposed to Burund historical patterns of cooperation.
24 7 Eastern Congo Congolese pre colonial history and state f ormation, as it were, did not occur in the same organic fashion as in Rwanda, Burundi, and Uganda. In the intralacustrine region, Eastern Congo stands out as the primary exception to the rule of small kingdoms assimilat ed to produce the modern state. In fact, the only kingdom, as it were, that existed as an entity comparable to the other states was that of Ijiwi, a large island in Lake Tanganyika about halfway between Southern Burundi and present day South Kivu province, DRC. This kingdom did not play nearly the central role to the development of Eastern Congo in the same way that they did in other parts of the region. Although the Kongo kingdom dominated pre co lonial relationships between clans and tribes in the country and the European traders, they fell from power in the 18 th century, leaving a power vacuum upon which the colonial state imposed a top down loosely controlled network of trading posts under the g uise of the Belgian kingdom. At the Berlin conference of 1884 1885, the Congo was the only territory to be awarded to a non state power in Western Europe, given the name of the Congo Free State, to be administered and profited from solely by King Leopold I I. When international outrage over working conditions in the burgeoning rubber industry and modern slavery arose, the Belgian government annexed the possession in 1908, adding to its growing Central African empire. From almost the moment Western explorers stepped foot on Congolese soil, the size of any potential territory proved daunting. Colonial administration under the Congo Free State amounted to little more than provincial trading posts located along strategic routes with few connections between outpo sts, and difficulties in transportation and communication from one side of the territory to the other. This inability to project
248 power over much of the territorial conquest continued under the Belgian colonial administration, initially based in Boma, but moved to Leopoldville (now Kinshasa), thousands of miles from other outposts in Stanleyville (now Kisangani). Thus, in addition to facing challenges of hundreds of ethnic groups, diverse environments and conditions, colonial rulers and those who would foll ow encountered the challenge of a seemingly vast territory. This disjointedness as a result of size remains a facet of governance in the Congo, even today, with Herbst noting that the DRC is especially prone to rebellion because the population is decentral ized and far from the center (2000). Citizens in one part of the country did not resemble those of their erstwhile countrymen, and this led to disconnect not only between the state, whether colonial or independent, and civilians, but also between civilians themselves. This severely limited the impact of popular protests during the time of Independence to form a more inclusive state that accurately reflected views across the entire population. This problem allowed for political power to remain geographically concentrated, first in the hands of Leopoldville colonialists, then in the hands of Kishasa (Kinois) leadership. Even though political leaders rose to prominence from outside of Kinshasa, like Ernest Wamba dia Wamba and Laurent Desire Kabila, both from th e East, they were still seen as outsiders on the main political stage, and the differences between political figures from the East and West remained a splinter point in Congolese politics. The size of the country also allowed for more contentious ethnic an d regional clashes, both immediately preceding Independence as well as throughout the history of Independent Congo and Zaire. Social relations between the Congolese Independent State and the citizenry also suffered from the problem of separatist outbreaks frequently occurring as a result of both the enormity of
249 the country and the diversity of people within and the meddling of outside forces. Katanga, also called Shaba, broke away almost immediately on July 11, 1960, eleven days after the declaration of Ind ependent Congo, led by noted anti Mobutist and still active politician, Moise Tshombe. Kasai followed suit. On both sides, foreign assistance poured in, to either protect valuable mining interests in Katanga and Kasai, the two richest provinces or to suppo and the United States all played pivotal roles in the events of the next few years, although the interests and desires of the Congolese people themselves, and their ability to form social organization o r for rebels to rely on this kind of internal support were limited by outside force. This eventually culminated in the death of the popular nationalist Patrice Lumumba, supposedly at the hands of the United States (De Witte 2008), and the forcible return of Katanga and Kasai to the Congolese Union. Outside Influences were not only instrumental in the Independence rebellions, however, but continued to remain a problem in Congolese society and politics. In the Southern provinces, mining interests and a support ive government in Angola provided haven for new Katangan rebels to conduct raids, redirect resources, and provide materiel. In the east, Rwanda, Burundi and Ugandan co ethnics and even some political regimes provided material support and access to markets for Congolese citizens over the years. Porous cross border regions in Kivu Sud and Kivu Nord ensured that the migration of Hutu and Tutsi civilians of Congolese, Burundian or Rwandan citizenship occurred at a steady pace for the 20 th and 21 st centuries, increa sing during times of conflict. This cross border migration also proved problematic during the Great Congo War that bega n in 1996, for although Kabila received crucial support from both Rwanda and Uganda,
250 many accused him of having Kinyarwanda blood, spreading the widely accepted rumor that his son, Joseph Kabila (to become the president of Congo in 2001 after the death of the elder Kabila) was the son of a Rwandan Tutsi mother. The close linkages also sparked ethnic clashes in the border provinces during the war. After the war, the problem of ethn ically driven violence and conflict in North and South Kivu continued unabated, with Mai Mai militia (local ethnic community members) fighting against Banyamulenge rebels who often acted with Rwandan RPF support, like the CNDP( Congrs national pour la dfe nse du peuple ) or the newly emergent M23. Overall, the history of Congo shaped social relations in such a way that citizens had very little connections to the central authority in Kinshasa, most often finding bonds with either outside actors or governments or within their own ethnic or geographic communities. Even though the initial response of citizens to Lumumba was one of hope and change, this overall pattern of disconnect de emphasized loyalty to the Mobutu regime and made uniting the country under a rebel faction or potential political challenger incredibly difficult. The Great Lakes States are Feudal, Too? After this brief de scription of the pre colonial hi story of the region, the question arises of the fit of the Western understanding of the proce ss of historical state development here. It seems fitting to describe some parts of the Great Lakes region as one of the few places in sub Sah aran Africa where the kingdoms closely ma tch the modern day states like the feudal states that underpin Western Eu rope. This is exemplified by the Rwandan kingdom, historically aggressive towards neighbors, power
251 concentrated at the center of the kingdom, and the building of the kingdom through The Premise of Inequality (19 61) highlighted the elaborately feudal system that supposedly composed pre colonial Rwanda and became the dominant narrative about historical development between the lordly Tutsi and the peasant Hutu in Rwanda, later extended to Burundi by virtue of nearn ess and similarity. In Burundi although similar relationships existed, the structure of society was such that local loyalties dominated those over loyalties to the monarchy. Even in pre hs of territory associated with particular kingdoms, and even then these kingdoms understood relationships between king and subject very differently than in Western Europe or even in Rwanda and Burundi. As for the Democratic Republic of the Congo, the Kong o kingdom operated as a slave trader for European interests by conquering smaller tribes through war and violence and then selling them into slavery. This vaguely resembles feudalism in that the kingdom grew its power through conquest and expansion. Howeve r, the kingdom never existed across the whole of modern day DRC and unlike other places in the region where kingdoms expanded and grew their political and social power into the modern age, the Kongo kingdom largely disappeared by the 1700s. It would be a m isunderstanding of history to assume that feudalism shaped all citizen state relationships across the Great Lakes and provided a template for social interactions that influenced modern day rebellion. This certainly finds little ground in the Congo, the sit e of literally hundreds of rebel movements since Independence in 1960, or in Uganda, where the existence of a plethora of kingdoms and ethnic groups precluded any overarching dominant historical development. However, the existence of long
252 standing relation ships of authority and social protections that developed in pre colonial Rwanda and Burundi did structure civilian state relationships there, in ways that st ill resonate in the modern era. Thus a more appropriate statement about the links between feudalism social relations and the Great Lakes may be that in Rwanda and Burundi conditions bloomed that allowed for social structures conducive to using civilians, either by the state (as in the Rwandan Genocide) or by the rebel organization against the state (as in Burundi). Overall all, however, it is difficult to accept an assumption of a pattern of state and society formation vis vis feudal structures. Social Power and Temporal Linkages in Civil Wars The variable of social power as I define it is an extre mely important variable d uring all phases of a civil war from inception to most importantly to this study, post conflict. Furthermore, rebel organizations may likely face an uphill battle in changing the initial conditions of the source of social power. T his means that once a relationship has been established (or conversely, has failed to be established), it can be difficult to rewrite the behaviors of the rebel towards the civilian or vice versa. This is not to say that this cannot be done, as in the case of the NRA in Uganda in 1981 or the SPLA in Sudan in the 1980s, but when doing so, the rebel organization must recalculate potential costs and benefits. I mention these because building civilian support and re shaping state institutions is a potentially more forward thinking strategy than simply winning the war and hoping to rule through military strength. While the level of organization contributes to the cohesion of the rebel g roup and the unity of the movement overall, without building a foundation and having enough supporters to see
253 the movement to its goal of changing state power, these goals may go unrealized. A diagram provided in Figure 5 2 illustrates how this occurs duri ng a civil war. Civilian Support before the W ar In the lead up to civil war civilian support and the subversion of state structures helps to build the networks and capabilities of would be rebels. In Burundi, civilian support for FRODEBU, the pre cursor to CNDD FDD and the party of Melchior Ndadaye and Leonard Nyangoma, was very high before the war. This was thought to be the party that would usher in a new period of peace and prosperity and the beginning of majority rule democracy, and the honeymoon peri od following the June 1993 elections had not civil society free of ethnic violence, where citizenship would no longer be held hostage 6, p. xiii) The party itself started clandestinely among students and Hutu elites in the late 1980s, with Ndadaye and others organizing before the 1993 elections. Wh ile the election was not won on ethnic lines, (FRODEBU won with sixty five percent of the vote, and mostly Tutsi UPRONA landed thirty five percent, not reflective of actual demographics), it was non etheless assumed by those in pol itical and military circle s to be indicative of the advent of Hutu power in Burundi, despite the fact that President e lect Ndadaye immediately assembled a multi ethnic, multi party government, including several prominent Tutsi promoted to positions of power, like Sylv ie Kinigi as p rime minister and Jean Marie Ngendahayo as minister of foreign affairs.
254 (Ninodera 2012:11). Thus, projections to the outside community showed a multi ethnic movement committed to democracy, but internally, FRODEBU leaders and rousers depended greatly upon the ethnic card as a way to gain voters. This simmering ethnic tension that arose between both those who relied on e thnic calls to vote during the election (FRODEBU) and those who were threatened by a shift to ethnic power (UPRONA) made for increasingly explosive conditions in the fall of 1993, especially between the armed forces and the government. In a pattern to be r epeated often over the next two decades, civilians were caught in the middle when violence did break out with the assassination of Ndadaye in October, followed quickly by the gathering of youth militia and armed groups in the capital city. While most acc ounts understand the rising of the Hutu led insurgency as a response to the assassination and crackdowns by the Tutsi army and government sponsored youth groups, the question arises of whether support for armed insurrection arose before the election. T here is some evidence that civilians were willing to support an armed insurgency, both within FRODEBU and within PALIPEHUTU (FNL) although initially the Palipehutu more willingly engaged in violent acts against the government and civilians. In some interviews former rebels mention the influence of Palipehutu in training and driving Frodebuists (official supporters and party members) to violence. 2 The Palipehutu FNL had been making raids on towns in border communes near Tanzania since the early 1990s, destroy ing property, pillaging, and sometimes killing Tutsi civilians. A further investigation into the history of the PALIPEHUTU FNL below shows their development, influence and interaction with the CNDD FDD.
255 A Parallel Movement: The FNL As detailed in previo us chapters, the PALIPEHUTU FNL first named the P arti pour la libration du peuple hutu (Palipehutu) upon crea tion in 1980, later am ended to Palipehutu FNL ( Forces nationales de libration ) in 1985, and finally changed again to FNL in accordance with new legislation requiring no ethnic ties in names of political parties in late 2008 was created much earlier than the CNDD FDD, finding its roots among the refugee camps in Western Tanzania stemming from the 1972 genocide. Remy Gahutu created the group in 1980 in the refugee camp known as Mishamo, and in the following years held elections, congresses, and even created party pillars: self awareness, unity, the party, leadership, and the armed branch ( umuheto). In 1984, the movement created the youth league ( Jeunes Patriotique Hutu) movement ( Movement Femmes du Patriotiques Hutu ), echoes of which would be later seen in the creation of these same divisions in the CNDD FDD. In fact, many future CNDD FDD members got their start in the Palip ehutu 3 while it was actively recruiting in refugee camps in Tanzania. The refugee camps made for easy recruits for both the Palipehutu and later on, the CNDD FDD, because of the development of what Liisa historic al narrative developed about the genocide of 1972 that was taught and reinforced by survivors and refug ees. This narrative was incredibly influential in not only the foundational story of the Palipehutu, but also of the CNDD FDD, who referred to themselves in 1994. Gahutu died somewhat mysteriously in 1990, under uninvestigated circumstances while in police custody in Tanzania. After Gahutu died, Etienne Karatasi
256 took over, and cross border attacks aimed at the Tutsi military continued unabated throughout 1991. In 1992, leadership again changed to Cossan Kabura, a militant trained in Rwanda and much quicker to use armed insurrection and combat as a tactic than previous leaders. He remained in power for most of the civil war, until an overthrow by Agathon Rwasa, the current leader, in 2001. The Palipehutu FNL undoubtedly influenced the CNDD FDD in numerous ways: in their philosophy, structure and recruitmen t strategies. The CNDD FDD would later use refugee camps in both Tanzania and the Eastern DRC as potential pools of recruits during the civil war. As explained earlier, the prevailing political histories in the camps, combined with contemporary conditions imposed by Tutsi led governments created profound feelings of persecution and ethnic tension. Before the civil war, it was fairly easy to separate Palipehutu supporters from those who supported FRODEBU, although there was a fair amount of similarities and overlapping support. As Palipehutu was not yet a political party, political support for the Again, Palipehutu also had a much longer history than FRODEBU, which only trac ed its government. While both parties competed for the same pool of civilians and potential militants in the Hutu population, the tension between the two was not yet ripe, and some reports even state that members of the Palipehutu encouraged the FRODEBU to that before the war Palipehutu was nothing more than a fringe group whose main con tribution to the general Hutu population was to create more distrust and conflict in
257 the areas where they attacked between civilians and the Tutsi army, as evidenced by its attack on Tutsi civilians that prompted th e incidents at Ntega/Marangara in August of 1988 (ICG 2000: 4 ), as discussed in Chapter 3 Furthermore, the Palipehutu was unable to gain regional support in many provinces other than Bujumbura Rurale, further localizing its insurrection. Eventually FRODEBU put in place plans to start its own ar med group, called the Gnration Dmocrat ique Burundaise (GEDEBU). GEDEBU initially began as the youth league of the political party, operating in the Bujumbura suburb of Bwiza, a section of town mostly known for its large Congolese, West African, and Musl im populations. The group began to operate almost immediately upon the death of Ndadaye, suggesting that earlier plans may have been laid for the group. In fact, some informal observers note that FRODEBU was always split between those willing to work for a political solution to group mostly operated in the urban environment, at first discouraging citizens against villes mortes in French), planned protest by opposition parties that frequently devolved into violence, and later acting against the Tutsi youth militias Sans Echec (Without Failure, in French) and Sans Defait (Without Defeat, in French), in violent clashes in the str eets. For many years of the war, it was difficult for actors, civilians, and outside observers to pinpoint precisely the differences between PALIPEHUTU, CNDD FDD, and so ea rly years of the war (1993 1997 for the CNDD FDD and PALIPEHUTU together, 1993 2000 for the CNDD FDD and FRODEBU militants together), these groups often
258 acted in concert with one another, sharing the battlefields in both rural and urban areas against Buyoy Civilian Support during the war (CNDD FDD and FNL) As the war broke out, civilian support, subverting the resources of the state, and the ability to depend on a local population became critical to the CNDD FDD strategy. Previous linkages with F RODEBU lent the movement credibility, although as discussed above, confusion about leaders, titles, and goals of the movement may have lent it more support than it actually had among the population (in the way that civilians supported local heroes or leade rs associated with the movement, not necessarily realizing the greater connection between local militias and the nascent rebel organization). This is not to say that civilians were ambivalent or unsupportive of the CNDD FDD and their goals, but that confus ion sometimes ruled the day more so than anything else. By 1994, however, it was very clear to civilians, especially in areas most touched by war the Hutu suburbs of Bujumbura, Makamba, Cibitoke, and Bubanza provinces who the CNDD FDD was and their inten tions as to the outcome of the war. At first, the message to civilians was explicitly couched in ethnic terms. Parents who joined CNDD FDD were often encouraged to tell their children, especially those of militant age, the story of how sident, we must seek justice for ourselves. The Tutsi are afraid of our 4 This changed, however, over the course of the civil war as a propaganda campaign of education conducted by mobilitizeurs and civilian supporters of the CNDD FDD, in combinatio n with disastrous policies and practices enacted by the Buyoya government against rural (and Hutu) populations, both worked to change the way Hutu
259 civilians saw the role of the CNDD FDD and their potential to change discriminatory political and social cond itions. Civilian Support and Spreading the Message The role of the Mobilitizeur But how did the CNDD FDD recruit civilians to their cause in the first place? This message, of ethnic tension and retribution, helped to entice future CNDD FDD members and sup porters. This message was spread by local supporters and mobilitizeurs (mobilizers) spread ideology and recruit new supporters, both militant and political. 5 These mobilitizeurs were generally, but not always, politically active members of the community who were recruited early on and expected to maintain their political roles in the group following the end of conflict. 6 The methods employed by these members included secret house meetings and the sp reading of tracts of support and encouragement among the population. 7 These actors were extremely influential in garnering material support for the FDD armed faction of the group as well: collecting food, firewood, and potential laborers for the group when asked. 8 According to interviews, these mobilitizeurs did not directly participate in conflict, and their roles were confined to political actions. 9 Thus, they were not soldiers or combat participants, but rather acted as the political extension and link b etween actions the CNDD FDD pursued on the battlefield against the FAB and the civilian population, who could provide support for military operations as well as a symbol non armed protest against the Buyoya regime and potential voters or supporters for the CNDD FDD. This strategy grew more refined as the political philosophy and goals of the rebel organization developed: while initially, mobilitizeurs acted almost exclusively in support of increasing the brute strength
260 and capacity of the CNDD FDD fighters (through supplying troops with food, shelter, and materiel), fairly early on (according to one interview, as early as 1996/1997 in Kabere, a commune in Bubanza, 10 these actors began to organize lessons for civilians on democracy and democratic governance at the root of the CNDD FDD. The CNDD FDD pursued a strategy of using civilians as mobilitizeurs for three reasons: the legacy of relationships between authority and civilian in Burundi, the realization that victory would not come as a result of superior mil itary might or tactics, and protecting future interests. The legacy of Burundian state society intera ctions detailed in Chapter 3 show that even under the pre colonial kingdom, local authority figures patterned ways of communication from elite political an d social figures to civilians in such a manner that continued to structure behavior even in the modern era. Thus, during the civil war when the rebel organization CNDD FDD and its elite commands and soldiers were seen as the authority figure for rural resi dents, mobilitizeurs stepped in to provide a connection between citizen and ruler, and a way for civilians to both understand decisions made by CNDD FDD as well as provide feedback. Additionally, very recent memories of these kinds of political actions ste mming from the failed 1993 elections remained fresh in Burundian minds, with mobilitizeurs playing key roles in the FRODEBU victory (Nindorera 2010). The CNDD FDD also realized early on that the chance of a military victory defeating the FAB was incredibly unlikely. The rebel organization simply could not win on material strength or capability (as opposed to the RPF who were supported by foreign assistance, trained in military tactics and strategy and prepared for long term warfare), and thus would need to rely on civilian support to show both the Government of Burundi under Buyoya as
261 well as international actors observing the conflict that the rebel organization had the loyalty and desire of the population behind it. Finally, the CNDD FDD undertook specific political action from its inception as a matter of strategic planning for the future, planning action for not only winning the war, but also setting in place a political regime. Mobilitizeurs were aware of the post conflict expectation, and although it is difficult to elucidate what the probable post conflict expectations were at the time for these actors, many still hold roles in the CNDD FDD, 11 and some even mention these roles as a would choose to become mobilitizeurs. For one, participation in the war effort in this manner was far less costly that combat participation. Those civilians who found the message of CNDD FDD and its anti Buyoya government stance enticing could contribute to the war effort without placing themselves in mortal danger, although the Buyoya government did imprison, torture, and even kill those thought to be colluding with the enemy. Many mobilitizeurs also joined the cause because family or friends did 12 (Samii 2012), showing a bandwagon effect. Finally, as other researchers (Uvin 2009, Nindorera 2010) have elaborated on more fully, many civilians supported the message, goals and aim of the CNDD FDD wholeheartedly, and wanted to show this in any way they could. Targeting the population Hutus and Rural Poor Although the CNDD FDD presented itself as a multi ethnic group determined to bring Burundi back to democracy, the makeup of the rebel group, both on the political side as well as the armed one, was filled with far more Hutu than Tutsi. Part of this was the logic of the outbreak of the war, where ethnicity seemed to be the overarching political concern for all parties involved. Another part of this is simple mathematics. As
262 to begin with, and they left before the guns came. When the CNDD FDD came, we joined 13 As the war progressed, serious recruitment drives also took place in the refugee camps that sprang up as Hutu civilians fled the country into eastern Congo, Rwanda, and Tanzania. The CNDD FDD followed the same template passed down from the PALIPEHUTU FNL and FRODEBU of devising separate organizations for members, youth, and women. They did this for a number of purposes: the first b eing an organizational strategy in keeping with the way the other groups were organized, as described previously. As it seemed to be a pivotal strategy for FRODEBU for organizing citizen participation during the 1993 elections, so then, would it be a stra tegy used by the CNDD FDD in preparation for future electoral politics. Again, initial recruitment and message spreading happened among Hutu civilians, because, as elaborated above, it was a more convenient strategy to define the civil war ethnically first as a result of actions by the Tutsi government and army. But as the war continued and initial ethnic clashes between CNDD FDD and the FAB diminished after 1996, so did CNDD FDD strategy in targeting potential allies and supporters among the population, bo th as a way to increase their base and as a way to differentiate themselves from the more ethnically defined PALIPEHUTU FNL. The Abagumyabanga During the war, the CNDD FDD devised plans to maintain confidentiality and secrecy against the interrogators of the Tutsi army, and in doing so, created a name for all general members of the group. The abagumyabanga
263 Kirundi) formed as a member organization to identify and interact with fellow members both at home, as well as if forced into regroupment camps by the Tutsi army or into refugee camps abroad. Secret signals were to be used if encountering others in a drinking esta FDD bird, 14 as a way of contacting other members and signaling information to be passed, signals that remain in use. This organization formed the core of political and material support durin g the civil war, to which their name alludes secrecy was of paramount concern during the war as to avoid detection by the government army. Mobilitizeurs were the heart of this group, but it also encompassed civilians who wished to be associated with the C NDD FDD without necessarily holding a role as a propagandist. It is unclear either from interviews with participants, internal CNDD FDD records, or previous academic accounts when this terminology and group came into existence: most scholarly and policy re ports only reference this term in the post conflict period to denote the general league of CNDD FDD supporters, although interviewees referred to its use before the war. 15 The problem also exists that to claim abagumyabanga membership at this point post wa r is rather easy: because records did not exist (although mobilitizeurs and combatants in rural areas like Musigati commune claim to know all local supporters, 16 it would be easy enough for a Burundian to claim membership for material benefit post hoc, espe cially if the citizen had moved from a rural area to an urban city, where verification becomes less likely. Whatever the precise origin of the terminology, its use now separates casual supporters or non aligned Burundians from those associated with the mov ement. It was expressed to me that those with membership cards became members during the war, and thus were in better
264 positions to rely on the CNDD FDD political machine for employment, assistance, and protection, be it actual or psychological. I provide a n example of the modern cards below. According to interviews, no cards were either issued or still in existence (again, the fog of war makes these questions unclear) from the civil war period. I point to the name, symbols, and behaviors of this set of sup porters of the CNDD FDD to draw attention to the fact that whatever the term used, a group of civilians with a specific relationship and understanding of said relationship to the combatants of the CNDD FDD existed during the civil war that carry over to th e modern period. This long standing association shows that the CNDD FDD as a rebel organization thought about ties to Burundian society in ways that would continue to serve them even after the war ended, showing both their commitment to a grass roots polit ical movement (and not just armed insurrection) and to enlarging the size, scope, and goals of the group. The CNDD Abakenyererarugamba while they were a rebel movement. This group, like t he youth league, was to act as a leadership and steering committee for women members of the subsequent planning for the future. 17 n aturally limited during the civil war, the movement nonetheless planned a healthy conflict planning as early as 2003, when the ceasefires were set in place between the CNDD FDD a nd the Buyoya Government (Gasana and Boshoff 2003). Like the
265 Abagumyabanga, it is unclear exactly when the Abakenyererarugamba was formed or its precise origins: Party elites pointed to an early formation during the civil war, 18 although civilians did not m ention this in interviews even when directly questioned. There also seemed to be no mention of this particular terminology in newspaper accounts from Le Reneveau of the time. What is clear, however, is the post war role of the group, especially in its rega rds to forming part of the CNDD FDD hierarc hy. As discussed in Chapter 4, Alice Nzomukunda, the ousted member of CNDD FDD, led the president. This shows two things: first, the realization of the CNDD FDD at some point during rebellion that the inclusion of citizens, especially ones as vital to Burundian social life as women, would help to shape and share the political message following the ending of the civil war. But this of course, is somewhat cynical given that the Burundian constitution, written during the political transition of 2003 2005, includes a provision that both the National Assembly and the Senate consist of 30% female members, in proportion to party representation (i.e. the women mu st be from different political parties in proportion to the seats won in the National Assembly). Thus, some in CNDD FDD may have seen the writing on the wall early on and pushed more a more inclusive group with specified roles and sub organizations for wom en for this very reason. Secondly, however, it shows a willingness of CNDD FDD political elite to recognize the importance of society in the electoral cycle in 2010 including rallies, community development meetings, and social events. These events allowed for women to know local CNDD FDD representation, to receive and share literature and propaganda paraphernalia (which would then be worn
266 as a physical symbol of sup port), and interact with other members. It also must be noted that the CNDD FDD understood and accepted the role of women in leadership and combat even before the civil war. Sylvie Kinigi was the prime minister under the FRODEBU government (from which many of the CNDD FDD political elite originated from) and many women actively participated in the civil war as both battle field combatants (although the exact numbers are hardly known, with estimates ranging from 494 to 1200 (Lahai 2010:12) and auxiliary supp ort roles. While it may be an overstatement to infer too much about either the CNDD some recognition between civilian voter and rebel participant elite that existed during and after the war. The Youth League, Imbonerakure The use of youth wings of the various political factions in Burundi to act as party mobilizers, recruiters, and sometimes even violent actors was not original to the CNDD FDD. In fact, the youth wing of the pre civil war UPRONA party, the Jeunes Revolutionaires du Rwagasore (JRR the Youth Revolutionaries of Rwagasore ) played an important role in the genocide of 1972. Acting in concordance with the Tutsi army of the Bagaza led government, the JRR became a secondary killing force in the rural areas, especially when dealing with Hutu students (Lemarchand 2009: 413). Like other youth wings, the CNDD principle of providing a place in the movement for those young (18 35), unmarried members (typically male, but there are no provisions as to sex) to participate. It is unclear from interviews or other accounts when and where the Imbonerakure was
267 created, although there is some recent suggestion that they were founded during the war and actively participated in combat. It is thought that these youth may have been soldiers who then transitioned into these roles as an extension of their support in the post conflict context. Althoug h interviewees who were combatants were reluctant to admit membership in the Imbonerakure during the period I did research, even when questioned, 19 it is not unlikely that the transition from soldier to Imbonerakure occurred. This reluctance to admit memb ership had to do with both the popular rumor among the Burundian population (related to me in at least 30 conversations from May to August 2010) as well as reports by international observers and monitors (Ghoshal 2010, ICG 2009, HRW 2010) of intimidation, threats, and physical violence by the Imbonerakure to sway election results. Although their origins may be suspect, the role played by this youth group in the continued consolidation and shoring up of power post violence is immense, discussed further below State Structures and the CNDD FDD Although civilian relationships are an important component of the capacity factor, the story does not end there. The other half of the variable is the ability of the rebel group to co opt, subvert, or create new state s tructures, either taking over or fulfilling roles that the state would normally perform. This emphasis on new roles for the rebel organizations has a dual purpose of both preparing the group for post conflict governance as well as showing the strength of t he rebel group to the population, in its abilities to perform as well as or better than the current state and provide a credible alternative to the existing government. Hence capacity captures both axes of the triangle as described above the state as well as civilians.
268 These relationships also capture a new dimension explored in rebel studies that state capacity. While some scholars explicitly reject the ideas that taking over parts of the state enterprise are a state b uilding function (Mampilly 2011), others find that either take over or creation of these functions and institutions can create shadow states (Reno 1999, Reno 2000) that exist within a larger functioning state or alternatively, can provide state resources w here none exist (Keister 2011). This section of the chapter focuses on how the CNDD FDD used pre existing state structures (and the resources available through these) and the revival or creation 20 order, and support while they were a rebel organization. They did so as part and parcel of a strategy to cement relationships with civilians, subvert badly needed material resources, and plan strategically for the future of the organization after the war. These strategies undertaken by the CNDD FDD severely limited the options of the Government of Burundi to pursue effective counter insurgency tactics to deter civilians from joining the movement and providing support, overall contributing to a civilian s desire to see the government overturned as well as creating a legacy of support for the post war CNDD FDD political organization. Parallel Administrations Often times in modern civil wars, a rebel organization will set up a parallel (to the already exi sting state government) structure of local administration (Wickham Crowley 1987), as was the case in Uganda under the NRA, which set up local administrative councils run by the NRA and its supporters during the Ugandan bush war (Kasfir 2005). A number of t hings can occur within these rebel proto state systems: coercion of the
269 local leaders previous to the war, persuasion of existing leaders, replacement of unfriendly leaders with rebel supporters, or creation of entirely new administrations separate from th e state enterprise. The CNDD FDD had great success in creating parallel structures of administration for civilians under their control (Longman 1998, Nindorera 2008, 2010, ICG 2002). Especially in rural areas, where administrative control prior to the civi l war already suffered from the problem of being isolated and far from the capital and seat of government, the CNDD FDD set up local councils with elder administrative figures with tax collecting capabilities, judicial capacities, and links between combata nts and civilians (as part of the mobilization and education efforts) (Longman 1998:17). These councils provided security in the form of a physical symbol of CNDD FDD control over the territory, as well as a sounding board to bring civilian concerns to the ongoing military operation. Additionally, the councils could act as intermediaries in the flow of information between those who participated as look outs or leadershi p. These parallel administrations provided a way for the CNDD FDD to organize its civilian operations, not only providing internal logic for the movement, but more importantly, providing civilian support to the movement and vice versa, again, cementing th e relationship in hopes of sustaining it for the future. These thrived in areas under CNDD FDD territorial control, which served to protect combatants during transit, control flows of information, and corral resources, especially in the areas between Kibir a Forest (in the Northeast of the country) and Ruvubu National Park (outside of Bujumbura), including Ngozi, Karuzi, Kayanza, Bubanza, Muramvya, and Gitega provinces (Longman 1998:31). Two pictures emerge from interviews conducted in my
270 fieldwork: the fir st, from elites and combatants, of these administrations as self run entities designed to sup port the larger combat effort. The second picture that emerges, from civilians, especially in areas under direct CNDD FDD territorial control, is one of unclear un derstanding of what exactly these administrations did beyond providing logistical support (Longman 1998:48) although their existence was undisputed. Thus, there seems to be a concentrated effort among elites to reframe these councils as key to the war eff ort, whether or not civilians actually agreed with this. In interviews, civilians of local elders created under the Mwami and gradually done away with after Independen ce) who heard disputes between neighbors, 21 although the nature of their authority and what the specif ic issues to be resolved were. The documentation of these councils in other accounts of the war confirms their existence and provides new evidence for thei r extent. Local Councilmen and Co Option of Members of Government The CNDD FDD also embarked on strategic co optation of local members of the Buyoya government, to both shore up support as well as use state resources that the local members ha d access to, against the state. They did this in plain sight of the government by choosing Hutu members who found the CNDD FDD cause palatable, usually former FRODEBU members. These members were already seen as susceptible to the cause, and might have been more easily enticed to joining the CNDD FDD as a way of planning for a political future under the presumed post war CNDD FDD regime. The government of Burundi reacted to this by removing most Hutu politicians from positions of power, even at the lower lo cal levels of government. By 1997, only 31 of the
271 121 communal administrators were Hutu, and 22 of them UPRONA (Longman 1998: 1996 Coup, strengthen UPRONA political a nd social control, and shore up military dominance on the battlefield. One mobilitizeur in Bubanza province (Bubanza commune) expressed that co optation of local nyumba cumi and zone administrators was seen as a vital role the mobilitizeurs would play, bec ause co optation of local officials operated as both a way to get state resources and a way to spread information about the CNDD FDD through a respected community figure. 22 This strategy does not seem out of context with the larger descriptions of the creat ion of parallel administrations in other accounts, although it was difficult to question those who led communes during the war without some interviewer effects: i.e. I was seen in those cases as someone influenced by (and possibly spying for) the CNDD FDD, and the answers to these types of questions were met with a fairly standard response that the leader had always supported the CNDD FDD, even during the war. 23 My translator and I discussed extensively whether or not co opted leaders actually supported the CNDD FDD during this time, or if they were simply hedging their bets against victory for either the CNDD FDD or the Government of Burundi. The fog of war and the current regime made these answers difficult to ascertain with any certainty, but it seems like ly that both cases existed: local leaders who were co opted who genuinely supported the CNDD FDD and their mission of government turnover, as well as those who wanted to protect their own political power and personal safety should either group win the war.
272 Creation of a Court System Forest Justice Many interviewees mentioned the existence of a localized court system to resolve disputes, sniff out potential spies and informers, and judge local crimes 24 in CNDD FDD controlled territories, espe cially near its operating base in Kabira Forest (Longman 1998, Nindorera 2008 and 2010). These courts were staffed by CNDD FDD members with community ties, mostly belonging to the political faction CNDD rather than the military one FDD. 25 There were, according to surv eys with former combatants, separate justice systems for offenses committed by FDD combatants as opposed to civilian actors. Although some accused the courts of being slanted towards CNDD FDD members, the courts were open to all civilians who lived in the area, not just active members 26 and disputes between neighbors could easily be brought before the courts, especially ones involving conflic ts over land or farming areas. There was some suggestion in interviews that these courts did operate as a way for CNDD FDD sympathizers to gain material advantage through adjudication over their peers who did not support the rebel organization, by bribing judges and administrators 27 although this only came up in one singular interview in a particular area (near the rebel base location of Kibira forest) and was never highlighted again, even when I prompted civilians with this line of inquiry. As explained above, these forest courts also were presented in interviews as part of the re emergence of the abashingatahe/village el der system of rule. 28 There was also rumor of a high court connected to the elite levels of the CNDD FDD operating at the main rebel base in Kibira Forest in Bubanza province, but it seemed to be use more for military violations among combatants than for ci vilians. The purpose of creating a court system for civilians was two fold for the CNDD FDD: one, by
273 allowing civilians to met out justice themselves, it removed a potential source of work (and headache) from the combatants of the CNDD FDD, freeing them to focus solely on battle related matters. Secondly, it allowed for the CNDD FDD to build stronger relationships with civilians as they could later express the sentiment that the rebel organization tried to maintain societal order and relationships with and between civilians eve n in the midst of a civil war. The inclusion of civilians within administration and judicial power of the CNDD FDD provided a credible history of the group as the true and legitimate representatives of the Burundian people and their in terests, in a way that other rebel movements and even the existing Government of Burundi under Buyoya could not. Planning for a Political Future Observers of the political history of the CNDD FDD point to the distinct future orientation of the group in the willingness of the political faction CNDD to negotiate and jockey for post conflict political positions (Ninodera 2008:106 108), both in terms of dialogue with the actors involved in the civil war (the Government of Burundi, international actors who w ould presumably enforce provisions of peace and transitional agreements, and even fellow rebel grou ps) as well as with civilians. These actions were more than dialogue and political negotiation, however, towards the end of the conflict (around late 2002), the CNDD FDD began a program of amassing a war chest to prepare for elections and the transition to a fully functioning, fully funded political party. They did so through the process of raising cotisation (contributions) organizing those previously involv ed with CNDD FDD as either administrators or mobilitizeurs to act as collectors. Small amounts were collected on a regular basis and receipts were issued to
274 civilians. The physical representation of a contribution, while illegal to possess, provided physic al evidence of the relationship between the rebel organization and the population. These pieces of paper were carefully hidden and kept for many years after the war (all of the examples below were obtained from 2010 2011), and treasured as a sign of CNDD F DD support and membership. A few examples are illustrated in Figure 5 4. Violence against Civilians during the Burundian Civil War Although the chapter thus far focused on the positive aspects of the relationships affiliated civilians enjoyed with the CN DD FDD, it cannot be said that the CNDD FDD did not engage in violence against civilians especially those of Tutsi descent or those accused of supporting the Palipehutu FNL But even violence against non supporters was matched in some geographic areas wit h violence against would be Hutu supporters as a tactic of intimidation and battle strategy. Violence against civilians was unsystematic in type, temporality, or purpose. As one Burundian proverb expresses, This was because as the general combat violence in the war waxed and waned over its twelve years, so did violence a gainst and between civilians. In the ACLED (Armed Conflict Locator Events Dataset) dataset, the most compreh ensive dataset (to date) on specific, geo referenced incidences of violence during civil wars in Africa, The CNDD FDD committed 136 separate, delineated acts of violence against civilians, defined as any armed/violen t group attacks unarmed civilians. Rebels, governments, militias, rioters can all commit violence against civilians. This is the only event that involves civi 11) This was over the course of the Burundian civil war that the data set codes for, from 1997 2005. In the
275 same period of time, the Palipehutu FNL committed 138 separate, delineated, and reported acts of violence against civilians. The reports of these events all came from international media sources, and thus while there m ay be some data issues, the evidence provided shows that in fact both rebel organizations committed violence against civilians. Additionally, another 543 separate events of violence against civilians ither the Palipehutu FNL or the CNDD FDD. In terms of absolute numbers of casualties, the Uppsala Conflict Dictionary Dataset of One Sided Violence v1.4 2012, during the war CNDD FDD was responsible for 155 civilian fatalities, the Palipehutu FNL, 427 civi lian fatalities, and Sided Violence v1.4 2012). While credible quantitative data on the exact number of fatalities committed against civilians is sparse in regards to the civil war, these data nonetheless point to the use of violence against civilians by both rebel and government as a tool of conflict. By far, the most data that exists about the CNDD FDD and Palipehutu FNL committing violence against civilians exists in the form of reports by Human rights o rganizations and international actors compiled during the war. Many of these reports specify what type of violence was committed against civilians (ethnic, sexual, or political). This helps to show the targeted nature of the use of violence against civilia ns as another tool to either intimidate or coerce the civilian population into supporting the rebel organization. I now move to specific discussions of the type of violence employed by the CNDD FDD in relation to civilians.
276 Hutu and Tutsi Violence Early o n in the civil war as issues remained explicitly fixed in ethnic terms, so too did violence against civilians with the FDD armed branch of the CNDD FDD attacking Tutsi civilians and Tutsi held strongholds. Wholesale ethnic cleansing of neighborhoods occurr ed in Bujumbura city, and in the countryside roving bands of combatants fought the army and the civilians that supposedly supported them. It is hard to distinguish the source, nature and scope of ethnically targeted violence during the early part of the ci vil war (1993 early 1997) because many reporting agencies, like the United Nations or Human Rights Watch were unable to travel throughout Burundi or gather data. Many civilians when interviewed paint a picture of this time period as one of extreme chaos an d violence specifically colored by ethnic tension, especially in rural areas further away from the central military capacity in Bujumbura. 29 Scholars describe the events as genocidal acts by the Tutsi government army against civilians with Tutsi and Hutu civilians fighting each other in the milieu (Lemarchand 1996). In 1998, scholars estimated that over 150,000 civilians had died during the w ar (Ndikumana 1998:30). While the CNDD FDD actively participated in violence against civilians and the the rebel movement in droves, because of the abject failure of t he state (described in Ndikumana 1998, Uvin 2009, and Lemarchand 1996) to act as anything other than a eerily similar fashion to the events of 1972. Journalistic accounts a t the time (Jelinek Army. Again, while it was clear to even casual observers that the violence was ethnic in
277 nature and the sides clearly demarcated between Hutu and Tutsi, t he exact role of the CNDD FDD in violence against civilians at this time remains murky. This could be due to the fact (el aborated earlier in Chapter 4 ) that the CNDD FDD had not yet coalesced, and many would be CNDD FDD belonged to smaller, unorganized or unnamed militia. Rape and the CNDD FDD Rape became a facet of the civil war as a tactic favored by rebel groups and the FAB ( Force Armees Burundais the official name for the government army), especially as the first round of negotiations broke down in 1997. In asking questions about rape during interviews, many civilians were initially reluctant to give me or my translator information, because of community shaming and fear of reprisal attacks from persons in power inherent to the nature of the discussio n of sexual violence in wartime (Wood 2003). We approached the questions with delicate phrasing specific to the terminology of the civil war in Burundi to determine whether rape had occurred in the geographic area, if the civilians knew someone who had bee n a victim, if they themselves had been subject to sexual violence, and finally, if the perpetrators were known. Interviews of civilians by human rights organizations confirm the occurrence of rape in conjunction with civilian fatalities during the war (Lo ngman 1998:88). In interviews, a pattern emerged of rape as a tool of violence against civilians to coerce support in rebel held areas in one of two ways: by combatants not native to the area (i.e. by Southern rebels operating in Northern provinces) or by combatants in areas thought to be too sympathetic to the Government of Burundi. Women interviewed in Kibira Forest, Musigati and Bubanza communes in Bubanza were quick to note both of these occurrences 30 with the understanding through the interview that th e perpetrators were
278 CNDD FDD 31 An interviewee in Mpanda, a commune on the road between urban Bubanza ville, the seat of Bubanza province, and Bujumbura, noted that Government soldiers raped women first, 32 which is fitting with the long held notion by both civilians and international observers that the CNDD FDD did not commit atrocities against civilians (Longman 1998: 81), a myth long disproven. Violence agai nst the Palipehutu FNL and its S upporters 1997 was also the year in which cooperation between the CN DD FDD and the Palipehutu FNL broke down (although other accounts note that the two rebel organizations had begun to fight each other almost immediately after the war broke out), and the former collaborators began to fight each other as well as the FAB. Ac cording to ACLED dataset for the period 1997 2005, 21 separate events of battles took place between the CNDD FDD and the Palipehutu FNL group s where control of the contested location does not change. If the government controls an area, fights with rebels and wins, this is the correct code. If rebels control a location and maintain control after fighting with government forces, this is the corre ct code. If two militia groups are fighting, this is the correct code. Battles are the most common activity and take place across a range of actors, including rebels, militias, and government forces, communal groups These events are no t coded to include CNDD FDD violence against civilians who might be affiliated with Palipehutu FNL. The Palipehutu FNL was a smaller group with a far fewer territorial strongholds, and often in the unlucky position of finding themselves in close proximity to the Government of Burundi army, which severely impacted their abilities to fight against both the
279 Government of Burundi army and the CNDD FDD. In rural areas in Cibitoke in the north or Tanzanian border provinces in the South, clashes between CNDD FDD a nd Palipehutu FNL occurred, and these areas are the places where the CNDD FDD committed violence against civilians suspected of harboring Palipehutu FNL allegiance (Longman 1998:88). This pattern of violence against civilians intensified as the civil war c ontinued, and CNDD FDD began to plan for the future and the elimination of potential political rivals. This was in fitting with the changing character of the civil war and violence, from a decidedly ethnic struggle to one of political power and authority. This pattern of violence towards civilians with Palipehutu FNL ties spilled over into the post conflict period, with a serious of extrajudicial killings, intimidations, and political violence against FNL members in both the pre electoral period (Ghoshal 2 010) and 2011 2012 (ICG 2012, HRW 2011). The Palipehutu FNL Civilians, and the War Effort The Palipehutu tended to attract those Hutu civilians who felt more ethnically marginalized than others, because the message and strategy to attract followers was framed in such explici tly ethnic and anti Tutsi form. This strategy resonated especially with those in marginalized areas of the country: the hinterlands, especially those bordering Tanzania with easy access to the refugee camps that had been there since t he 1970s, and the parts of Bujumbura city that were most affected by the Tutsi strangehold on commerce, Kinama, Kenoysha, and Carama, which bordered Palipehutu FNL strongholds in Bujumbura Rurale province. The Palipehutu FNL embraced a strategy of recruit ing from the CNDD FDD and other rebel movements, poaching fighters who were disgruntled with treatment and
280 conditions who could bring other recruits and war materials with them. This strategy was especially embraced as the civil war continued and the rebel movements grew in size and strength. This was not always a voluntary defection, with the Palipehutu FNL being the rebel group with the most noted practice of abducting and forcing children into soldierdom, although other rebel organizations, CNDD FDD incl uded, also relied on forced abduction as a means of growing combat strength, 33 although many combatants expressed they went voluntarily to join the war effort 34 The exceptions being those accounts found in either those that left the CNDD FDD 35 or in report s produced by international observers (UNHCR 2004). Voluntary recruitment age during combat was supposed to be set at 16 for both rebel and government forces, but children as young as ten participated in combat. My own housing security guard in Burundi was a child soldier recruited at age 12 to join the CNDD FDD during the war when it spread to his commune in Mwaro province. 36 But by and large, the convention wisdom was such that the FNL and the Government of Burundi forces employed more child soldiers gain ed through forced abduction in combat. The Palipehutu FNL also set up parallel administrations similar to those set up by the CNDD FDD in their own rebel controlled territory (Bujumbura Rurale province) (ICG 2002). Like the CNDD FDD, the FNL also prepar ed for the future elections while operating as a rebel organization, also collecting cotisation, even when it was still an active fighting force (2005 2009).
281 Civilian Support after the war (CNDD FDD) The Building of a Political Party: Disparate Goals and N ew Cleavages With the war coming to a close with the signing of the Arusha Accords and various ceasefires in 2003, the CNDD FDD was now tasked with the difficult process of changing a rebel movement into a political party expected to conduct itself as such: i.e. avoiding violence and gathering supporters while navigating a complicated transitional government system. At least, initially, however, public support and the backing of international authorities painted a rosy picture for the incoming CNDD FDD ( Uvin 2009). At elite levels, a new program to re invigorate former Tutsi politicians who participated in FRODEBU or those seen as willing to join with CNDD FDD began. This involved courting politicians through promises of new posts and increased ethnic and political dialogue in the hopes of bringing democratic reform in the spirit of Ndadaye and the 1993 elections to Burundi. The CNDD FDD courted Jean Marie Ngendahayo, the Tutsi former member of FRODEBU who was appointed Minister of the Foreign Affairs and the Interior under Ndadaye from 1993 1995 and his wife, Antoinette Batumubwira, to become part of the new political movement and highlight the multi ethnic character of the movement. 37 But the new recruitment was not limited to Tutsi, it also included form erly non political Hutu, like Hafsa Mossi, who left Burundi to pursue a career in international journalism at the beginning of the civil war, returning when the CNDD FDD took power to become the presidential spokesperson. The continuing use of the militia : Imbonerakure as Intimidation and Enforcement As the nature of the CNDD FDD has changed in the transition from war to peace, so too, has the nature of the Imbonerakure, the CNDD FDD youth league. As I
282 explicated above, there exists an assumption positing that many members of the Imbonerakure were combatants during the civil war, and thus the Imbonerakure is merely a continuation of political violence in the post conflict period. The league became especially active in politics around 2007 2008, when the fir st accusations of political assassinations were leveled against the CNDD FDD. In the years since, many former party is enforcement, including within party discipline as we ll as attacks on Palipehutu FNL members (Ghoshal 2010, HRW 2012). Their roles only grew as the CNDD FDD attempted to consolidate power in the post conflict era. But this consolidation was only possible because of the foundation built during the war, attrac ting members to the cause, setting in place networks, and gathering intelligence for the post war political climate. Furthermore, some interviewees not in CNDD FDD living in urban environments (where violence tends to be concentrated) suggested that former combatants were being paid to participate in Imbonerakure and carry out violent acts. 38 Why CNDD FDD and not FNL? When examining the two rebel groups, CNDD FDD and Palipehutu FNL, the question that most frequently arises is that given similar histories an d trajectories, as well as similar objectives during an ethnic civil war, why did the CNDD FDD succeed in transforming itself into a dominant political movement, while the Palipehutu FNL has had to settle for localized levels of support and has proven una b le to win national elections? The reasons for varying levels of support are context dependent upon the particular province and commune, but several global reasons can be found for the CNDD conflict political struggle. The CNDD
283 FDD earned more general widespread support than the Palipehutu FNL, which chooses to pursue deeper ties with smaller communities rather than global linkages in several provinces across Burundi. While it was true the CNDD FDD had its own territor ial strongholds, these were less a factor of inability to expand (like that which plagued the Palipehutu FNL) than a matter of strategic location of bases to engage the FAB in the civil war. The CNDD FDD was also able to win more support than the Palipehut u FNL in terms of the messages delivered to civilians. The CNDD FDD, although mostly Hutu, always embraced a message of multiethnic composition, welcoming Tutsi into the political factions, appointing Tutsi to important roles in the post conflict governmen t as a further sign of the commitment to ending ethnic conflict. The Palipehutu FNL, on the other hand, routinely acted with violence against Tutsi civilians, including mass killings at Bugendana displaced persons camp and other gatherings of Tutsi non com batants. And the group did not change its message until forced to by the post conflict government in 2006, when it began to actively recruit Tutsi to the cause (with little success). Historically, the group never opened itself to Tutsi participation Even presently, the Palipehutu FNL represents itself as the party of Hutu and non urban (read as explicitly Tutsi, Business or Political Elite) interests. The CNDD FDD was more willing to negotia te with community actors and promise post conflict political gain s to win support and encourage civilians to follow the rebel movement. This ingratiated the movement to the communities they operated in, and as word spread of the actions of the movement, engendered good will on a larger scale. The Palipehutu FNL, on the other hand, kept power localized to specific regions of Burundi. Thus, the CNDD FDD was seen as an internal movement, created for all of Burundi, while the
284 Palipehutu FNL were seen as outsiders with desires to only represent their own interests. Finally, t he CNDD FDD had great success in dictating the narrative to civilians of being provoked into war by an unyielding Tutsi oligarchy intent on murdering all potential Hutu challengers, whereas the Palipehutu FNL was seen as an agitator force, from its attacks on civilians in 1988 to its attacks on the city of Bujumbura when peace negotiations broke down in 2008. Decentralization and Building Civilian Support in Civil Wars Decentralization of aut hority, as examined in Chapter 3 plays an important role in t he battle conduct of a civil war. But how then, does decentralization affect the relationships between civilians and rebels during the war? It is a question worth considering when looking at the political history and trajectory of the CNDD FDD, to understa nd how this organization could have not succeeded militarily and succeeded politically. Or to be blunt, why the FDD (armed branch) failed and the CNDD (political branch) succeeded, both during the war and after. Whereas in armed conflict, separately organi zed cells are less advantageous than having an overarching leadership, in political organizations, whether leadership is global or local is unimportant (up to a certain point) because support is built in either case. Thus, the fact that decentralization wa s a central feature of both the CNDD political branch and the FDD armed branch of the movement was a favorable development in building long term political support among civilians. Furthermore, while both arms of the movement can be characterized by this le vel of organization, the decentralization was not nearly as severe among the CNDD political faction, as the struggles over leadership and tactics simply never emerged among the mobilitizeurs and political organizers. And the political faction
285 did at one po int gain a strong leader (and an increase in centralization of authority) in the form of Hussein Radjabu, the second in command of the CNDD FDD under both Jean Bosco Ndayikengurukiye (1998 2001) and Peter Nkurunziza (2001 2005) with de facto authority ove r the political activities of the CNDD. Radjabu knit a cohesive political strategy that he was effectively able to use especially in regards to post conflict planning for both political power of the CNDD FDD in Burundian politics, as well as his own person al political power. I use these examples to illustrate that while decentralization in the armed branch of the CNDD FDD worked as a potential deterrent to victory through combat, when it came to managing civilian relations, subverting the state, and buildin g credibility among a population needed to sustain the rebel organization long term, the effects were opposite. Even though for most of the disorganized and decentralized, this positively affected the ability of the group in regards to its source of social power. And even though resulting loyalties post conflict are uneven (rural areas tend to favor CNDD FDD in much higher proportions than urban areas, and certain areas (Bubanza and Cibitoke most notably) are thought to be strongholds), the memory of CNDD FDD and civilian support during the civil war matters more to the abilities of the CNDD FDD to retain political power than the manner by which it was accomplished. Conclusion This chapter explored how civilian support and the ability of the CNDD FDD to conflict
286 political power holders. While not the materially most capable actor during the civil war, the CNDD FDD nonetheless emerged victorious, gaining control of the state during the transition and maintaining political power of Burundi since 20 05. In this chapt er, I examined how the roots and nature of the social linkages between the CNDD FDD, civilians trapped in between the government and the rebels during the war, and pre existing Burundia n state structures interacted. The CNDD FDD used civilians during the w ar, creating social structures for them, both new and borrowed from the existing state. These actions provided a legitimate way for the CNDD FDD to connect and build credibility among civilians in a way that extended beyond mere physical protection and saf ety, a good the rebel organization was sometimes incapable of providing. In doing so, the rebel group kept a sharp eye on the future of the movement as it planned for a post conflict existence gaining political power in Burundi, with the help of the citize ns who supported it in the bush. Tables and Figures Figure 5 1. Diagram of Relations in Civil Wars
287 Figure 5 2. Temporalities and Interactions with Civilians Figure 5 3 Abagumyabanga membership cards, CNDD FDD
288 Figure 5 4. Examples of Cotisation Receipts, 2001 2004
289 Notes 1 This is important because in the post colonial period, rebel organizations in Burundi found more ground among civilians when challenging the center, whereas in Rwanda they did not. Clearly, this history had left its mark. 2 Personal Interview, March 21, 2011 and January 22, 2011 3 Person al Interview, March 21, 2011 4 Personal Interview, October 10, 2010 5 Personal Interview, April 2, 2011 6 Personal Interview, April 2, 2011 7 Personal Interview, March 21, 2011 8 Personal Interview March 24, 2011 9 Personal Interview February 27, 2011, Jan uary 13, 2011 and March 23, 2011 10 Personal Interview, February 12, 2011 11 Personal Interview, March 7, 2011 12 Personal Interview, November 13, 2011, December 3, 2011, and January 24, 2011 13 Personal Interview, February 12, 2011 14 Personal Interview, Janua ry 29, 2011
290 15 Personal Interviews, November 14, 2010, January 27, 2011, March 12, 2011, April 6, 2011 16 Personal Interview, February 17, 2011 17 Personal Interview, November 13, 2010 18 Personal Interviews November 21, 2010 and December 5, 2010 19 Personal Interviews December 11, 2010, March 13, 2011, April 3, 2011 20 21 Personal Interviews March 4,7,8,9, 23, 25, 2011 and January 28, 29, 2011 22 Personal Interview, March 14, 2011 23 Personal Interviews, March 2, 2011 and March 21, 2011 24 Personal Interviews, November 11, 14, and 17, 2010 and February 16, 21, and 25, 2011 25 Personal Interviews February 12, 2011 and February 9, 2011 26 Personal Interview, March 9, 2011 27 Personal Interview, March 23, 2011 28 Personal Interviews, December 10, 11, 2011 and May 3,7,8, 2011
291 29 Personal Interviews, January 12,18, 21, 23, 2011 30 Person al Interviews, March 11, 17 and 19, 2011 31 Again, asking questions of who perpetrated rape, especially if the CNDD FDD was responsible, was incredibly difficult because of m y suspected ties to CNDD FDD and more general suspicions that community leaders wou ld find out and punish these women. 32 Personal Interview, March 22, 2011 33 Personal Interview, March 23, 2012 34 Personal Interviews November 13, 17, and 22, 2010 and March 7 and 9, 2011 35 Personal Interview, March 9, 2012 36 Personal Interview, May 21, 201 1 37 Personal Interview with Jean Marie Ngendahayo, February 19, 2011 38 Personal Interview, November 13, 2010 and December 7, 2010
292 CHAPTER 6 REBELS, WAR, THE GREAT LAKES AND SUB SAHARAN AFRICA Introduction In the Great Lakes, rebellion began in earnest with the birth of the National Resistance Army/Movement (NRA/M) as a revolt against the regime of former collaborator Milton Obote in 1980. Although seeming to affect only the internal politics of Uganda at the time, this new movement would cause reverberations throughout the region spa nning the next three decades, fundamentally reshaping the dialogue and action of political change there. From its inception, the NRA gave weight to a new philosophy in post colonial politics: that change was possible, and indeed, necessary from the one par ty dictatorial regimes that took shape after Independence. Taking insights from the in depth examination of the CNDD FDD in Burundi, this chapter explores the other rebel turned ruler transformations in the region: the NRA in Uganda, the RPF/A (Rwandan Pat riotic Front/Army) in Rwanda, and the AFDL ( Alliance des Forces Dmocratiques pour la Libration du Congo Zare ) in the Democratic Republic of Congo. While these movements differ in their organizational strategies and sources of social power, they all came about in a region sharing remarkable political, social and economic similarities (especially between Rwanda and Burundi) and emerged out of individual civil wars. Some states in the region also shared analogous pre colonial, colonial, and post colonial hi stories, and thus I examine the legacies of state conditions and antecedent development in the creation of the rebel organizations as well. In doing so, this chapter illustrates the core logic of the argument presented about rebel to ruler transi tions and post rebel outcomes. After examining these case studies, the chapter
293 then proceeds to a larger examination of rebel to ruler transitions across sub Saharan Africa. Why the Great Lakes? As explored in Chapter 3 some countries in the Great Lakes share similar pre a nd post colonial hi stories, especially in terms of previously existing ethnic structures and state cohesiveness and the fragmented politics that resulted after the colonialists left. As independent nations of their own, these countries all exp erienced military and dictatorial regimes, economic pressures, and eventually, massive political upheaval and civil war. These wars ranged in scope and duration of violence, from the relatively mild Ugandan Civil War sometimes called the Ugandan Bush War (1981 1986) with about 100,000 estimated casualties to the all out destruction of the Civil War in Congo, which all share primary dependence on agriculture, with the exception of the Democratic Republic of Congo, whose natural resource wealth provides most of its gross domestic product, and rank poorly on the Human Development Indicators Index Rwanda 166, Burundi 185, the DRC 187, and Uganda 161(in 2011), indicating similar lack of development, advancements in education and public health. With all these similarities in country conditions and civil wars, however, different patterns of organizational power and the sources of social power emerged among the various rebel groups that would eventually come to power, as referred to in Table 1 1. The NRA The first movement to rise from rebel to ruler in the region was demonstrated in Chapter 3 Uga ndan history finds similarity to the kingdoms found in
294 pre colonial Rwanda and Burundi, although in Uganda because of the sheer number of emerge until the British colo nial power and their policies elevated the Buganda kingdom above the others and favored them in education, social, and political advancement. But, this was done in such a manner that violent anti Bugandan sentiment never became the overarching statement th at drove ethnic relations in Uganda. Anti Bugandan sentiment political movement at Independence. But to again draw a finer point on the interaction between ethnic history and reb ellion there, it simply never existed as a serious cleavage between Ugandans. Rather, Museveni saw the purpose of the NRA, even before it was known as that, as an explicitly reformist movement. This view stems from his interact ions (described in Chapter 3 ) with other post Independence revolutionaries at the University of Dar Es Salaam, Tanzania, in the 1960s and 1970s. He took these ideas back to Uganda with him, first working as a political researcher in the first Obote regime (Ngoga 1998: 92), before see king exile during the Amin regime in 1971. During this exile, Museveni had many opportunities to hone his philosophy on anti government movements, linking and learning from both the pan Africanism of Julius Nyerere and the ( Frente de Libertao de Moambique ). These interactions prompted Museveni to start the pre cursor movement to the NRA, FRONASA (Front for the National Salvation), which even sent troops to overthrow Amin, with little success, several times over the decade The opportunity to put this philosophy into practice emerged in the late 1970s when Tanzania and Ugandan political elites decided the Amin regime must fall. The Tanzanians invaded with several exile groups,
295 including FRONASA. The overthrow of Amin led to short lived political change in Uganda, with Museveni even taking a political role in one of the new regimes. The controversial election of December 1980 rapidly escalated a deteriorating political space in Uganda opened by the elimination of Museveni. Al deemed free and fair by the Tanzanians, many Ugandans, especially in the political classes found the results suspect and were not willing to sacrifice the political gains of the 1978 1979 war for another potential dictatorship. And for former agitators like Museveni, the path to political change was clearly defined by rebellion and overthrow. Museveni and his very small band of fighters (about 35) previously trained with both FRELIMO and fellow actors in the Ugandan Tanzanian War (UNLA, FRONASA), gaining valuable tactical battle experience. The war began in the Luwero Triangle region (a rural area hostile to Obote), gaining civilian participants there, and the main rmy engaged in heavy handed tactics against the rebel organization, routinely raping, looting, and arresting those with suspected links, only serving to push students and political dissidents into the movement. Acto rs in the Development of the NRA Ugandan : Yoweri Museveni leader of FRONASA, UNLA, and later of NRA (1981 present), gained political power of Uganda 1986, elected President 1991, 1996, 2001, 2006, 2011 Milton Obote Ugandan President 1966 1971, 1980 1985, leader of Ugandan (UPC), deposed by military commanders Rwandan:
296 Paul Kagame second in command of the RPF, Head of Military Security under Museveni, assumed Presidency of Rwanda 2000, re elected 2003, 2010. Fred Rwigyema original commander of the RPF, killed October 2, 1990 under mysterious circumstances in battle, former Ugandan military Chief of Staff under Museveni NRA Rebel Strategies during the Civil War (Ugandan Bush War), 1981 1986. This section details the history, development, and nature of the NRA in its rise to political power in Uganda from the beginning of the civil war there to illuminate key similarities and contrasts between the NRA and other rebel groups in the region that also rose to power. First I discuss the role of the Ugandan state during the civil war, and how the nature of this state affected rebel development, civilian support, and the ability of the Obote government to mainta in political power during war. Second, I discuss the type of organizational power the NRA possessed, and how this was developed and maintained over the civil war and into post conflict governance. Finally, I address how the source of social power for the NRA directly contributed to its abilities to win the civil war and become a credible, strong, and durable post conflic t governor. Role of the State during Conflict As discussed in Chapter 3 the Ugandan state never experienced the kind of cohesion or control that the Rwandan and Burundian states did, as the kingdoms that formed modern day Uganda were disparate and scattered, and consolidation only came about during the colonial exploration p eriod, and was met with significant social resistance by lesser ethnic groups than the Buganda, who were favored by the British with employment, education, and political authority. Post Independence Uganda continued to be ruled under the one party state of
297 control of the state arose in the form of the Ugandan military, which took power in 1971 emerged as a response to ethnic cleansing carried out by Amin, dictatorial policies that left the state economy weak, and lack of progress and advancement opportunities. The Amin regime essentially left the Ugandan state in shambles, with little control over wide territorial areas, and regular purges left the once orderly and disciplined Ugandan military and security forces unable to credibly resist internal challenges to state legitimacy (Baker 2004). Once Obote resumed power following the 1979 overthrow, the sta te apparatus continued to erode: corruption in the police, government officials and judiciary, torture of political enemies and victimization of civilians (Baker 2004, Kabwegyere 1995) became commonplace. This lack of state structure and responsibility hel ped to foster pro NRA sentiment among the areas in the Luwero Triangle where the rebel organization was first able to gain a foothold. It also made it so that the NRA confronted a disorganized government with little ability to exercise power over the milit ary, governmental structures and civilians. Organizational Power within the NRA: Infrastructural The NRA was one of the most highly organized and disciplined rebel groups in Sub Saharan Africa. Several high ranking officers, including founder Yoweri Muse veni) in the rebel movement were members of the Ugandan state intelligence service, military, and then other previous This history contributed to providing a foundation of discipline, order, and central ized authority within the movement. From its inception, the movement was able to project power over the territories they controlled, had a centralized leadership
298 command under Museveni with no factional splits or major internal discipline issues. This was in direct comparison to the Obote government, and the NRA used these differences to rob state banks, raid hospitals, and loot private businesses (Mwenda 2007: 26). Like other rebel organizations, the movement developed both the political faction, the Natio nal Resistance Movement (NRM) and the armed unit, the National Resistance Army (NRA). They did so as a way of separating combat duties, goals and planning with post conflict political duties, goals and planning, although according to Kasfir, this distincti 33).The NRA was similar to the early development of the RPF in that one person (Museveni) was the chairman of both the National Resistance Movem ent (NRM) and the commander of the National Resistance Army (NRA). The legislative branch of the rebel organization, called the National Resistance Councils (the branch of the NRA that most closely dealt with civilians and the Ugandan social structure) was also under his command (ibid: 27). Thus, the rebel organization was arranged in hierarchal fashion under one central authority also able to centralize all internal structures of the rebel organization while fighting a difficult military war against better equipped and far more numerous foes. Recruitment was carried out through using the NRM and NRC movements, and accepted members from any ethnic group, although the initial majority was Banyankole, like Museveni. The choice of location of the rebellion also played a part in centralizing the rebel organization: by choosing to operate in the center of Uganda, in the Buganda dominated Luwero Triangle, rebel soldiers were unable to rely on external support or bases and thus could build more internal discipline and cohesion. The movement also practiced meritocracy
299 in leadership and advancement (Amaza 1998:32), regardless of whether or not the combat recruits had previous military training or not, were all subject to basic training over potential favoritism early on, preventing future splits. Sources of Social Power The NRA established a pattern of contact and conflict with civilians that other rebel organizations used to replicate, over and over again all over sub Saharan Africa. Like other rebel organizations, the NRA never followed a static pattern of relationships with civilians that lasted throughout the five years of the Ugandan Bush War, but was one of the first to i nstitute local rule of civilians perpetuated by civilian governors, but set in motion and controlled by the NRA themselves. The National Resistance Councils, headed by Museveni, established levels of administrative and judicial authority for civilians livi ng in the rebel held territories at the village (RC 1), sub county (RC 2) and county (RC 3) level (Baker 2004: 2). These councils served as political forums for civilians living under the NRA, similar to the role that the local administrative councils late r established under the CNDD FDD in Burundi held (ibid:3). But in contrast to other rebel organizations, the plan of these councils was incredibly structured and well documented, with all adult members of the village able to vote on council issues, with a head of nine civilians forming the RC1 committee, who then in turn acted as liaisons between the NRA, NRM, and civilians. These councils served not only political roles for the civilians, but also funneled aid and supplies to the NRA, protection from betra yal to the UNLA Government army, and military intelligence on positions, strategies, and tactics of the UNLA (Kasfir 2002: 2). These clandestine political networks between the rebel organization and civilians helped to not only draw
300 potential civilian reso urces (in terms of either support or material) away from the Obote government, but also helped to build legitimacy and faith in the message and future governing strategies of the NRA. specifically laid out the social, political, and economic goals of reform for the post conflict Ugandan state. This development was largely the result of the leaders of the NRA (including Museveni) and their early experiences with ideologically based rebe llions like FRELIMO, who developed practices based on popular support and civilian inclusion in liberated Northern Mozambique (Kasfir 2002, Museveni 1997). These ten points included plans for the dismantling of traditional structures of oppression utilize d by the Colonial and post Colonial regimes, creating an independent economy and cooperation with other African nations in a solidarity gesture against imperialism (Amaza 1998:29). It was the express goal of the NRM wing of the NRA to educate civilians thr ough the Resistance Councils and spread future goals and plans through propaganda and inclusion (ibid:33). The Outcome of the Civil War and the Enduring Power of the NRA In 1986, Museveni seized power from the military coup that overthrew Obote just a fe w months earlier. It was clear that the Ugandan state held very little internal capacity or ability to govern. In coming to power with the support of both a well disciplined armed group unafraid of combat and very battle tested as well as a civilian suppor t and governance structure, it seemed as though the NRA (who now transitioned to using the NRM as its primary designation) was poised to lead Uganda into a brighter future.
301 The Structure of the Ugandan State under the New NRA When the NRA took power in Ug anda in 1986, the government and its structures and institutions were in December 1985, state resources were low because of looting by the regime as well as theft by the NRA. C ivilians also had very little faith in the Obote regime and by extension the Ugandan state, and while they supported the NRA, the rebel organization still had to rebuild lost legitimacy and show that they were capable of running the state. Continuing Inte rnal Discipline After the NRA became the Ugandan government in 1986, virtually nothing changed in terms of the internal logic of organization of the former rebel movement except the name, which the leadership of the group thought would signify the changin g of the movement from a rebel organization to a political one, intent on implementing their ideas and political goals for the future. The Continuation of the Resistance Councils and Local Administration One place the NRA did not lack resources was in th e support of civilians. The NRA/M established the local resistance councils as part of the overall strategy to ensure credible and legitimate citizen centered politics throughout the movement. This was more than just war time strategy, however: the NRA sa w these as a way to govern post conflict. This was reflected in both deed (the actual creation and management of the councils) as well as word (within the Ten Point Programme distributed throughout the Bush War). The NRA almost immediately made these coun cils and committees into statutes when they assumed power (Commonwealth Law Bulletin 1988: 566 and Katono and Manyak 2010: 7 8). Although initially more centralized, reforms at the behest of donors occurred during the early 1990s which culminated in the Lo cal
302 Governments Act of 1997 (Katono and Manyak 2010: 9), still in place today. Rwanda The History of the RPF I now contrast the development of the RPF and its conduct during the Rwandan Civil War (1990 1993) and the Rwandan Genocide (1994) with the other rebel organizations in this study. The RPF provides an interesting comparison to both the NRA, who essentially trained and provided the breeding grounds for the Rwandan rebels, as well as a comparison to the CNDD FDD, who operate within a similar state en vironment to the RPF. Rwandans in Uganda Rwandans had long been refugees in the southwestern which saw the violent overthrow of the Tutsi monarchy there and subsequent et hnically charged massacres ensuring the mass exodus of Tutsi from the country in the early colonial history and colonial history, while similar to the monarchy that existed i n Burundi, was much more rigid in its social categorizations (Hutu, Tutu and Twa) and the kingdom was the only one in the region to grow through military conquest of territory, thus, ethnic violence there took a much more static tone, and flight was common following ethnic massacres and turnovers in political power in the 1960s and 1970s. Consequently, in Uganda, entirely Rwandan communities sprang up around the initial refugee camps established (especially the one called Gahunge), communities that establis hed themselves as part and parcel of the Ugandan landscape. Rwandan students enrolled in Ugandan schools, took employment, even in the Ugandan government, and joined the armed forces. Even though these moves were interpreted
303 by some Ugandan leaders (notabl y, Milton Obote, President in the 1970s and over thrower of Amin in the 1980 Tanzania Hutu led governments in Rwanda bega n almost immediately when Rwandans of Tutsi origin crossed the border bands of rebel groups known as the Inkotanyi or Inyenzi (described in Chapter 3 ) raided and attacked government of Rwanda positions during the Kayibanda regime. The Rwandans living in e xile held far greater ideals than that of small armed bands pestering the regime: many influential members of the refugee community wanted to build a political organization to petition the now ruling Habyarimana regime for the right of return of refugees a nd the equal and fair treatment of all Rwandans, regardless of ethnic origin. The creation of the RPF was itself a was created outside the country where it intended to operate, its members were initially recruited among the armed forces of a foreign power, most of its combatants had never set foot in the land where they were going to fight, and they never managed to get any support from the masses of the population in w (Prunier 1998: 119). Initially, the Rwandan community in Uganda was not the seat of opposition to Kigali: rather Burundi was, with a much larger refugee population and sympathetic Tutsi ed Rwandan Tutsi against the Ugandan population, recruiting into the State Research Bureau service (ibid: 120). Thus, popular sentiment turned against Rwandans there. These circumstances led to the development of the Rwandese 1 Alliance for National Unity (RANU), a political organization founded by the refugee intelligentsia in 1979 after the
304 overthrow of the Amin regime, and quickly radicalized by the rigging of the 1980 election that brought Obote to seemingly false power. Obo te already earned the ire of Rwandan most populous area in the southwestern part of the country and the seat of the Banyakole ethnic group) where many Rwandans settled, thus i ncreasing competition over land, cattle and resources. seeking the weapons, military training and experience that would be necessary to provoke change in Rwanda. In fact, Paul Kagam e and Fred Rwigyema (the leaders of the RPF) were experienced combat veterans by 1980: both had participated in FRONASA during the Tanzanian Ugandan war, with Kagame even winning a post as interim Minister of Defense in the Obote government. Thus, the woul d be RPF foundations were laid: one, a political organization designed to stir up emotion (and funding) among the Diaspora, and two, military training and organization in the NRA. Rwandans played a key role in the victories of the NRA, and by the time the group took the capital Kampala in 1986, at least 3000 Rwandan troops made up their forces. The Development of the RPF as an offshoot of the NRA When the NRA took power in Kampala in January 1986, Rwandan Tutsi refugees celebrated their slice of the topp ling of the Obote regime. RANU, which had been operating out of Nairobi during the war, returned to Uganda to assess a new strategy for possible duplication of the NRA victory. The seventh congress of the organization held in Kampala in December 1987 expre ssed these goals more decisively by assuming a new name for the organization: The Rwandan Patriotic Front (ibid: 125). At first, signs in post Obote
305 Uganda indicated a rise for the Rwandan population: Kagame, Rwigyema and many other Rwandan NRA fighters we re given important government and defense posts, and Museveni even agreed to change citizenship laws to allow for refugees to attain citizenship after ten years. But by 1988, the Museveni and the NRA government backtracked from this position, removing Fred Rwigyema from his post, reshuffling Rwandans active in the military, and dragging out the citizenship question. These actions caused many young Rwandans, especially those in the NRA and government, to question as to whether pursuing life in Uganda was as attainable it seemed to be a couple of years before. The shift among the population was exemplified by their until his firing from the NRA, only supported lukewarmly the idea of a return to Rwanda or armed invasion (ibid: 127). Even though the removal of Commander Fred seemed to indicate a turning point in NRA RPF relations, many other would be RPF members remained in positions of power and access within the Ugandan military, which allowed for RPF leadership to gain weapons, training, and prepare for invasion. Explaining Rwanda: The rise of the RPF, state conditions, and power The Rwandan Patriotic Front (RPF) and its named armed group, the Rwandan Patriotic Army (RPA), grew out of a long history and association between the major leaders of the group and the NRA, but conditions for rebellion simmered in Rwanda long before this association. Listed below are the major actors in the history of the rebel organization. Actors in t he Development of the RPF Rwandan:
306 Paul Kagame second in command of the RPF, Head of Military Security under Museveni, assumed Presidency of Rwanda 2000, re elected 2003, 2010. Fred Rwigyema original commander of the RPF, killed October 2, 1990 under m ysterious circumstances in battle, former Ugandan military Chief of Staff under Museveni Juvenal Habyarimana former military officer assumed Presidency of Rwanda through military coup 1973, killed in plane crash over Kigali airport April 6, 1994 Ugandan : Yoweri Museveni leader of FRONASA, UNLA, and later of NRA (1981 present), gained political power of Uganda 1986, elected President 1991, 1996, 2001, 2006, 2011 Lead Up to the October Invasion By 1990, it became apparent to the RPF members living in Uganda that matters would have to be taken into their own hands, and tacit plans began to form for battle. Although plans had been made to discuss the status of refugees in detail with the Habyarimana government during the many meetings of the UN supported Rwandan Ugandan Joint Committee on the Refugee Question, Habyarimana undertook little action on this front, and many RPF members believed it accepting the returnees and gai said. 2 Although the RPF was extensively prepared with a large support network of both civilian political supporters in the refugee camps (Kinzer 2008: 46) and with the quiet support of the NRA leader ship, they sorely lacked Rwandan internal support. Some Hutu opposition to the Habyarimana regime existed and chose to join the RPF around this time, for example, future president Pasteur Bizimungu, but most were leery of the strategies and goals the RPF w ished to employ. The decision to invade Northern
307 Rwanda via Uganda was planned for October 1, 1993, exclusively by Fred Rwigyema and Paul Kagame. While extensive planning took place in terms of gathering weapons and vehicles from the Ugandan army and choos ing government targets within Rwanda, what happened during those first crucial days shocked the hardened veterans. Rwigyema and Kagame planned the October invasion for a date when both President Museveni and President Habyarimana would be abroad at a Unite Summit. They knew they would be confronting an enemy force with little battle experience, but did not account for how their lack of knowledge of the terrain, lack of understanding of the international context, and more importantly, th e population, would stymie even the best laid plans. The Civil War Battles took place over 1991 1992 in the northern provinces, the RPF proved decisive, victories did not provide a stable citizen support base: in fact, all Rwandans, Hutu and Tutsi, escaped the path of the RPF due to the influ ence of government propaganda. The civil war was typical of what war would look like in sub Saharan Africa in the co ming decade, with low level violence occurring in fits and stops. When it did appear that the RPF was gaining momentum and would potentially reach Kigali, the capital, outside forces, most notably France, stepped in to bolster the government army to push b ack the rebels. Growing concern by international and regional actors also helped to keep violence at a minimum: negotiations at Arusha, Tanz ania, backed by a UN Chapter VI peacekeeping and assistance mission (United Nations Assistance Mission in Rwanda (UN AMIR) occupied the main leaders of the
308 rebel organization and the Government of Rwanda and kept the war from developing much. I now turn to how the RPF employed and manipulated the type of organizational power, the source of social power, and the anteceden t conditions of the Rwandan state. The type of organizational power the rebel organization utilized was a direct result of organizations developed infrastructural power, meaning that they employed centralized, hierarchal leadership strategies and tactics and projected both physical and political power over the territory they controlled and within the rebel organization itself. Although seemingly chaotic at first, the return of Pa ul Kagame to the leadership (he had been training with US Army officers at Fort Leavenworth, Kansas, and returned immediately upon hearing of the failure of the invasion), ushered in an era of successful battle and political strategies to force the Habyari mana regime to the negotiating table. The Rwandan State In nowhere in Central Africa does the state play such a crucial role to the development of Post Independence politics as in Rwand a. As explained in Chapter 3 and detailed in this chapter, the intern al power and control the state exhibited over the population ensured that no rebel movement could credibly find traction among civilians. Even currently, the state the RPF inherited shows this kind of domination. Although the RPF is largely seen as a Tutsi instrument of political power, and well over eighty percent of the population is Hutu, no serious threa t of rebellion exists, and has no t since 2000 (Thomson 2009). However fortunate this was for the RPF when it inherited the state in 1994, during the civ il war this prov ed to be a burden for the RPF. At the beginning of the civil war in Rwanda, state power was so ensconced
309 that the Habyarimana government was able to convince civilians to flee rebel held territory by using the state propaganda machine, from the lowest levels of the nyumba cumi system to mayor and burgomasters all the way to the President distributing convincing messages about t he terror the RPF would bring. The power of the Rwandan state was also such that even though Tutsi had lived in fear of the Habyarimana regime since its coming to power in 1973, they followed the instructions of the state and its authority figures instead of the potentially more agreeable rebel organization. This message was not only state driven through government chan nels, but also backed up by a state sponsored and financed media, especially Kangura the Hutu power paper and Radio Television Libre des Mille Collines (RTLM), a private radio station close to F as invaders and anti Rwanda, and encroached on any potential civilian support that the RPF may have received. The state army was also strong enough to rebuff the RPF invasion (with the help of arms and training from foreign sources) and keep the territor ial gains of the rebel organization limited to the northeast portion of the country. Some called Rwanda a state sponsored media and the rise of extremist political parties who seemed poised to control in the north also signaled problems with potential state failure. But the true power and capacity of the Rwandan state under Habyarimana r emained the tight grip that the government held over the population, and the ability to entice, threaten, and coerce civilian participants into committing acts of extreme violence during the genocide.
310 Organizational Power in Rwanda Despite initial setback s on the battlefield, the groups itself experienced relatively little internal conflict or disorder over its two decades long history. As with the NRA, the RPF had a hierarchal command structure with one central figure at the top, issuing commands and batt le decisions with no alternative chain of command or source of power. Upon the death of Commander Fred Rwigyema, Major Paul Kagame assumed command of the RPA, the armed faction of the RPF. Like other rebel organizations, the movement was separated between the RPF and the RPA, although this distinction was not made clear (or even utilized, according to some (Prunier 1998) until after the RPF took power in 1994, when assuming the government mantle necessitated a split between political functions of the new st ate belonging to the RPF and the military functions to belong to the RPA. The RPF relied heavily on funding through external patrons, most notably the NRA, who now had access to state coffers and resources to support their former comrades in arms with mone y, material, and weapons as well as from the diaspora, mostly Tutsi, with a fairly high level of education and willingness to support armed struggle against the Habyarimana government, with whom right of return had long been a contentious issue. The rebel organization also took advantage of military victories during the civil war to steal and loot government resources when possible, as the NRA did during the Ugandan Bush War. The RPF was also incredibly dissimilar to other rebel organizations in that it alw ays saw itself as a potential state government (and a very strong one at that) in training, and thus from the beginning, in its behavior during the early part of the civil war and then through the negotiations at the Arusha Accord, showed infrastructural p ower, the capability to bureaucratize and
311 delineate specific functions, and the establishment of a system of governance headed by a strong executive, all of these characteristics being practices already in place with the rebel organization itself. It seeme d that although the RPF lacked in size compared to the Government of Rwanda, the abilities to control and shape the Rwandan state into their own image were incredibly high, and this potential became increasingly threatening to Habyarimana and his cronies. Social Power Because the RPF developed as an almost entirely external rebel organization in Uganda (as opposed to the AFDL, CNDD FDD and NRA, all of which developed internally to the state they wished to challenge), the opportunities to recruit from the civilian population and utilize state resources were extremely limited from the beginning. While the RPF could rely on the diaspora for funding and recruitment drives, this only exacerbated the notion that it was a foreign, invading force, rather than a gr oup of reformist minded Rwandans. Furthermore, propaganda and intimidation of the civilian population inside Rwanda kept the movement from gaining many supporters after the civil war began. That being said, however, when the Rwandan state under Habyarimana turned to murdering the civilian population as a control tactic, RPF recruitment inside Rwanda experienced a boon (Prunier 1998:132), as Tutsi Rwandans now saw the RPF as The type of warfare practiced by the RPF in the civil war against the Habyarimana government of Rwanda also eliminated the potential for 132) style, meaning that the RPF approached the Rwandan Government as an
312 the citizens stuck in the middle. The RPF was also unable to rely on the civilian population as simple demographics dictated their positionality with eighty five percent of the civilians belonging to the same ethnic group as the hated regime, it would be an uphill battle to convince civilians in an ethnically divided society to support a rebel group composed of ethnic usurpers reminiscent of the historically oppressive ki ng. Despite these challenges to finding inroads with civilians, the RPF was also fundamentally uninterested in pursuing civilian relationships, as it prioritized victory and dominance o ver any other goal or strategy. Besides failing to build civilian relati onships for support during the war, the RPF also failed to pursue the building of legitimacy with civilians in a key way during the genocide in that the rebel organization failed to protect civilians in the Government of Rwanda army, or in the massacres committed by the Government army in conjunction with Interahamwe militias (Stam 2009, Reed 1995). This would only exacerbate the disconnect between the RPF and both Hutu and Tutsi civilians, both during c onflict and after the assumption of power. During the civil war and subsequent genocide, there was also evidence to suggest that the RPF failed to protect Tutsi civilians as a matter of strategy that to focus efforts on civilians would take away from the overall goal of winning the war minded approach and Stam 2009 explores the phenomenon of military positions vis vis civilian strongholds or potential protection zones). This did not help to build relationships betw een civilians and the rebel organization, as Tutsi living in Rwanda (genocide survivors) realized the RPF had little desire to forge relationships with them. On the other hand, the RPF also employed
313 violence against civilians when it could (Reyntjens 2008, Reyntjens 2011, Prunier 1998, 2009), especially in terms of massacres of Hutus who could be potential genocidaires. This was a practice employed during the civil war and genocide, as well as later in the Great Congo War. The Genocide Although the RPF fo ught impressively during the civil war, international pressures to negotiate, the presence of the UNAMIR mission, and general government (under the RPF) made it so that event s were at a standstill in early 1994. Plans, however, were in development under the Habyarimana government with the collusion of the local militia recruited by government forces called Interahamwe ( those who work together in Kinyarwanda) to create killing forces capable of eliminating thousands of Tutsi civilians a day. These forces were provided weapons and training by the Government of Rwanda army in preparation for an ethnic war that would decisively settle the question of who would govern Rwanda, and e liminate the so power had become largely figurative by the spring of 1994, with the akazu ( little house) of Hutu extremists closely connected to Madame Habya rimana planning genocidal massacres and potential coups behind closed doors. Habyarimana was killed in a plane FRODEBU) April 6, 1994, setting off these plans and engulfing Rwanda in a nightmarish 100 day long genocide. Researchers and scholars speculate as to which group (Hutu extremists or the RPF) was responsible for the downing of the plane, but little concrete evidence has come to light 3 The genocide was only stopped b y the advancement of the
314 RPF and overtaking of Kigali on July 4, 1994, thoroughly cementing the RPF as the victors of the civil war and the new Rwandan government. The RPF in Comparative Perspective organizational and social power compared with other rebel organizations in the Great Lakes. The type of organizational power exhibited by the RPF directly contrasts the haphazard development and inte rnal structure of the CNDD FDD. Similar to the NRA in the formation of a militant organization with centralized leadership, tactics and strategies emanating from a hierarchal command, the RPF practiced infrastructural power, in that they used their centralized authority to create legitimacy throughout the rebel organization during the war, es tablishing internal coherency. They were also able to project power internally throughout the movement (meaning there was virtually no dissent, and when problems of authority arose, were quickly quelled) as well as externally over the territory they eventually came to control during the civil war. The RPF was different from other rebel organizations in that the group had very little interest or necessity in pursuing a cadre had nearly always lived in Uganda, the connections to Rwandans living under the Habyarimana regime were not forged until after the civil war began, and even then, tension between returnees (the RPF) and those who suffered in silence under the regime remained high. The RPF was eventually able to do some recruiting of armed combatants among the internal Rwandan population, but the imposition of political order for civilians, as in the case of the NRA and the CNDD FDD, the collection of material goods a nd resources for the use of the rebel organization, or the education of a population on political matters
315 and philosophy never occurred during the war the way it did with the NRA and the CNDD FDD. e, 1994 Present This next section describes the changes the RPF undertook after transitioning from a rebel organization actively fighting a civil war to a newly established government ate in July 1994, the rebel organization chose to implement the existing government agreements and structure made at the Arusha Accords in 1993, with a few notable new stipulations, like a strong executive (as was the case within the rebel organization) (R eyntjens 2011: 24). of the post conflict state. in civil service, government and judicial positions, Tutsi lik e the RPF who had grown up outside of Rwanda and most notably, were educated and used the English language instead of French, the official lang uage of the Habyarimana state. This was known as f the Rwandan state (ibid: 28). This exempli structure and philosophy on political power: that it was entirely up to the RPF high command to dictate the terms of combat, structure, and engagement. Repression and violence continued to be used against civilians as a tactic of int imidation, and rumors of massacres and killings of Hutu civilians plagued the new regime. These killings were seen as a continuation of the type of behavior the RPF practiced as a rebel organization (DeForges 1999, Reyntens 2009: 27). By 1996, the threat o f civil war was a very real possibility: attacks on Tutsi civilians by Hutu rebel groups who used the chaos of Eastern Congo to reform and rearm, to which the new RPF government had a
316 responsibility and duty to address. This new security challenge to the f ledgling state allowed for the RPF to continue behaving as a militant organization, as it had not yet made the transition to a political one. This is in direct contrast to the way that the NRA assumed power in Uganda, which almost immediately softened in s tructure and rigidity, focusing less on war and violence and more on governance. By 1996, well over half of the Rwandans state budget was spent on military expenditures, 20.6 bil lion Rwandan francs (ibid: 32). These conditions and the overt threat that lax Congolese borders and security constituted left the RPF little choice: it seemed as though war with Congo was imminent. Domestic matters, especially when it came to building support and legitimacy with Hutu civilians or Tutsi genocide survivors, took a bac kseat to these immediate priorities. That being said, however, the RPF did have a large advantage over other rebel groups that transitioned to state power in that although state resources were completely drained, the Rwandan state itself survived the violen ce. The administrative structure that existed before, with vertical hierarchy and even the ability to collect taxes and other forms of revenue despite poor conditions, was re instituted almost immediately following the genocid e. (ibid: 33, Lemarchand 2009) This enduring power of the state was also co opted by the RPF to keep it running like a military organization, using the structures to institute law and policies in centralized hierarchal fashion, and projecting complete power over the state and citizens through physical control, like the forced villagization process called imidigudu, which purported to address population pressures on the land, but actually served as a control measure for the RPF and a way to reward RPF loyalists with land, businesses, and growth opportunities.
317 Comparing the RPF to the CNDD FDD as Post Conflict Governments In Rwanda and Burundi, a natural comparison exists because of the closeness in ethnic makeup, pre colonial kingdom structure, and Colonial experience. It is with this in mind that I now turn to a brief comparison of the RPF and the CNDD FDD in their post conflict governing styles and legacy. First and foremost, both former rebel organizations had to address the issue of ethnicity in relation to governing and building le gitimacy with civilians. The CNDD FDD continued to do so by focusing on social programs (like universal free primary education and free healthcare for pregnant women and children under five) and allocating resources to citizens (services, structures, and p olitical order). The group focused less on developing or even maintaining internal discipline, and seemed to accept that former members would leave or be expelled as the needs of the political organization changed. The RPF on the other hand, played to thei r strength, developed and tested when they were a rebel organization, by focusing on projecting political power through dominance, hierarchy and organizational structure that reinforced internal logic and cohesion. Dissenters were treated harshly, either i nternal to the organization or in the civilian population, and the group used targeted violence and repression as a way to build power through fear. Although both organizations took control of highly organized states, the difference between the two pre exi sting states and their structures became clear: the Rwandan state under the RPF returned to a nearly identical system of centralized governance when the organization took power. The Burundian state, on the other hand, followed a more consociational model u nder the CNDD FDD.
318 The AFDL: Herding Cats in Congo As mentioned in previous chapters, the AFDL and the environment in which they operate, the Democratic Republic of the Congo, exhibit chaos and weakness in terms of organizational structures, projection o f power, and basic ability to function. Although the AFDL under Joseph Kabila still controls the government, many active rebel groups still exist in the Eastern part of the country, and many outside observers predict that the regime will continue to face i nternal challenges to authority and legitimacy (Stearns 2012, Seay 2012). Congo, Kleptocracy, and the Struggle against Mobutu As elaborated in Chapter 3 the Democratic Republic of the Congo, formerly Zaire, experienced many stalls in its post colonial de velopment. Thirty years of separatism, poorly run state institutions and bureaucracy, and a dictatorial, kleptocratic leader who used the state as the 1990s. State s ervices and government resources were virtually non existent, and much of the state existed as a black market. Although Mobutu had proclaimed the freedoms, and all owing political parties, this was more a symbolic gesture than one designed to truly allow political liberalization (Prunier 2009: 72). Mobutu also meddled in African affairs during the 1990s, hoping that in supporting other heads of state, they in turn w ould prop up his regime and provide material support. These relationships, especially the close one shared with Rwandan President Habyarimana, would ultimately seal his fate. Mobutu even provided arms ( Reno 1999 ) to the Rwandan government in the pre genoci de period. After the fall of
319 Habyarimana, the genocide in Rwanda caused severe problems for its western neighbor, as over two million Hutu refugees flowed over the border and into the eastern regions (already notoriously ungoverned and highly problematic f fearing RPF reprisal. This left the both the Mobutu government of Zaire and the new RPF government in Kigali in a difficult situation. Because Mobutu had supported the Habyarimana regime and was seemingly allowing the Rwandan genocidai res (those that committed violence during the genocide) to infiltrate refugee camps in the East and potential re group and re arm, the RPF saw the Mobutu government as an impediment to the security of Rwanda. When cross border raids began to occur, the RPF realized that threats of war would soon be realized. Serious ethnic violence had also erupted over much of the early part of the 1990s in North and South Kivu provinces, as a result of the spillover of ethnic conflict in Burundi and Rwanda. Congolese Tuts i, called or other ethnic groups). The influx of refugees from neighboring conflicts (Burundian Hutu and Tutsi into South Kivu and Rwandan Hutu into North Kivu) vastly chan ged the dynamics of ethnicity in these regions: crystallizing identity into a political tool and mobilization strategy, importing tensions, and providing the wheezing Mobutu regime new political targets. The new violence as well as the uncertainty also cau sed regional damage and tension: the RPF and the NRA realized the value that rebel groups currently organizing against them (the reformed FDLR/Interahamwe for the RPF, the LRA and other groups for the NRA) saw in regrouping in the undergoverned provinces of Zaire, as well as the burgeoning small arms market flourishing through cities and ports there. By the summer of 1996, it was clear to both Paul Kagame and Yoweri
320 Museveni that something had to be done. Because of their mutual histories of organic revolu tion organized in principle on reform and change (which was more important for rebels to overthrow Mobutu. This would provide a Congolese face to the action while also prote cting the fledgling NRA and RPF regimes from security threats. To make matters worse in the region, the civil war in Burundi continued to rage and Mobutu was suspected of providing aid and comfort to both Palipehutu FNL and the CNDD FDD. A series of meetin gs in both Eastern Congo and in Kigali in the late summer of 1996 cemented the AFDL from the remnants of several rebel movements and Congolese leaders (Stearns 2011, Prunier 2009: 113), signing the creation agreement at Lemara, South Kivu on October 18, 1 996. The primary leader and spokesperson was Laurent Desire Kabila, a rebel leader who trained with Che during his adventures in Congo in the 1960s and lived for many years in exile in Dar Es Salaam, Tanzania (where he met Yoweri Museveni). He previously h eaded a rebel organization known as the Parti de la Revolution Populaire (PRP). With Kabila, there were three additional figureheads in the movement: Deogratias Bugera of the Alliance du Democratie et Progres (ADP), Anselme Masasu Nindaga of the Mouvement Revolutionnaire pour la Liberation du Zaire, and Andre Kisase Ngandu of the Counseil National du Resistance pour la Democratie (CNRD). Of the four rebel leaders, only Ngandu laid any real claim to having combatants at his disposal, although the t roops numbered roughly 400 (Prunier the Mobutu regime since its inception (although often from Dar Es Salaam), and considered himself the political philoso pher of th e newly formed AFDL. With a manifesto
321 outlining the political strategy of the AFDL created and signed, the RPF and the NRA could now put a Congolese face on to their armies, who made up the majority of the battle troops. Furthermore, the war effort was almo st entirely funded by external forces, with little input or resources from Congolese citizens themselves. The war began in of about 2000 Congolese Tutsi troops who had bee n trained in Rwanda in the RPA (the Rwandan Patriotic Army, the new name for the army of the Government of Rwanda). These troops were able to recruit an additional number of Congolese Tutsi once the military operations began (Reyntjens 2009: 48), exchangin g gunfire in small battles with the Zairian army. The Mobutu government in the West responded in September with violence had erupted in the Kivu provinces, with the Zairian army fighting against self defense groups of Congolese Tutsis. By late October, violence had spread to most of the two provinces, aided surreptitiously by Rwanda, Burundi, and Uganda. By the end of 1996, the territory under rebel control stretched for about 800 kilometers long and 100 kilometers wide, buffering the Eastern borders. Although this constituted little of the whole territorial area of the Congo, in early 1997, the Angolan army joined t he rebel cause, providing direct support enough to overtake numerous provincial capitals in the march westward. By May 1997, Kabila, the AFDL, and the regional coalition of willing supporters had overtaken Kinshasa. Actors in the Development of the AFDL Congolese:
322 Laurent Desire Kabila leader and spokesperson for the AFDL (1997 2001), trained with Che Guevara, educated and exiled in Dar Es Salaam, former head of the PRP Joseph Kabila son of Laurent death, Susp ected by other Congolese to be of Rwandan origin Andre Kisase Ngandu figurehead and leader in the early years of the AFDL, leader of the CNRP, more militant than other leaders in AFDL and commander of a previous fighting force. Deogratias Bugera leader of the ADP, North Kivu Tutsi working as an architect in Goma before the war. Anselme Masasu Nindaga leader of the MRLZ, half Mushi, half Tutsi political agitator in Bukavu before the war. Mobutu Sese Seko the Zaireian president and dictator from 1966 1997, Paul Kagame vice president of Rwanda and director of state security, highest ranking military commander in the RPA Yoweri Museveni president of Uganda Pierre Buyoya newly re installed President of Burundi The Civil War and the Great Congo War: The Demise of the Kleptocratic State I now turn to a brief discussion of how the variables I focused on in the study of CNDD FDD in Burundi mattered to the outcome of the AFDL rebellion in the Democratic Republic of the Congo, providing further illuminatio n. The State The Congolese state in the early 1990s barely functioned the strongest institutions centered on President Mobutu, including the Praetorian Guard and exclusi vely. (Turner and Young 1985). This was the case from the very beginning of
323 preserve the kleptocracy. Any potential ways of shoring up ethnic or regional power in conte station to the regime were suppresse d. As described in Chapter 3 the lack of a pre colonial structure followed an inept and cruel colonial experience left Congo in of n atural resources. Any potential creation of a road network linking the thousand miles between East and West was stifled by corruption and mismanagement, inhibiting post Independence growt h. The one bright spot in post Independence Congo was that the country never experienced the outbreaks of state directed political violence in the same way that Uganda, Rwanda, and Burundi did. This, however, did not alleviate the suffering of the Congolese people, who by the early 1990s were subsisting largely through black market developments and subsistence agriculture, as most businesses and foreign aid fled the neg lect and corruption of Mobutu. Although multi partyism was best. genocidal regimes in those places caused internal ethnic issues to rise to the top. Coupled with land reform issues, citizenship and belonging as Congo approached the 21 st cent ury led to distinct uneasiness about what the future held for the country, uneasiness that the AFDL, with foreign backers, was able to capitalize on to use to their advantage. Rebel Organizational Power: the AFDL From its inception, the AFDL was as misma naged as the Congolese state itself. Formed from the remains of four rebel political movements and cemented almost
324 entirely by foreign backers willing to provide the material necessary for rebellion, but not the training or support neces sary to manage the insurgency. Furthermore, competing goals and strategies among the four leaders, including serious ethnic tensions bound by both the rise of autochthony questions in Congolese politics in the 1990s as well as more general questions of civil war, refugees, a nd spillover in the region, made it more Turner 2007 22 25). Although Kabila studied under or from famous rebels, including Che, Museveni, and Kagame, he was ultimately unab le to put theory and philosophy into practice that would unite disparate actors and provide a clear sense of leadership and vision within the AFDL. To complicate the situation further, whereas other rebel movements were composed almost exclusively of hom egrown troops, with homegrown commanders, and thus might experience some sort of internal logic and loyalty because of this, many of the initial fighters, mid level commanders, and even highest ranking officers of the AFDL were not even Congolese, or if th ey were, like Laurent Nkunda, were again the wrong kind of Congolese, i.e. Tutsi who had fought with the RPF during the Rwandan Civil War and Genocide. But even without these larger issues of internal conflict within the AFDL, the rebel organization was un able to organize on any sort of structural level: funding, organization, battle tactics and communications were all arranged and otherwise handled by Rwandan and Ugandan forces operating in the Congo. Because foreign troops made up so much of the fighting forces, discipline and enforcement became major issues, as the Congolese troops would easily defect against their foreign commanders, and foreign troops would find Congolese commanders lacking in order
325 and ability. Once Kabila came into power, much like th e rebel movement itself, he to create any sort of internal support and forced instead to provide autonomy to regional institutions and actors ( Turner 2007 : 33). This was reminiscent of the Mobutu state in wherein which the tools necessary to build internal capacity and structure were never utilized. Open conflict between separate fac tions of the rebel movement (both domestic and international) erupted mere months after the rebel group took Kinshasa, necessitating even more intervention from third party actors, like Angola, and causing a new counter rebellion by other actors, some form erly aligned with the AFDL. Relationships with Civilians: t he AFDL, Rising Ethnic Conflict, and Managing Interests The AFDL was unable to build relationships with civilians, and in fact, like the RPF, never had an initial need to in any capacity other than finding a face for the movement and a cause to nominally rally behind. Because financial support was provided by the regional allies for the movement (Uganda, Rwanda, and later Angola), the need to rely on civilian resources and in turn build relation ships not present, much like the case with the RPF in Rwanda. Soldiers were well equipped on their mission, which from the beginning of the movement was almost exclusively focused on toppling Mobutu, and less so on garnering or building civilian political support. Although the rebel organization designed a manifesto upon their creation, the spreading of the message was almost never practiced, and citizens had a hard time discerning goals other than
326 Unlike the CNDD FDD or the NRA, the move ment was also unable to provide security services and public goods to civilians in their quest, as crime and Also, the clear ethnic character of the movement that arose in the middle of conflict erupting in the 1990s, persuading those Congolese who were not Banyamulenge or Tutsi, even though Joseph Kabila was not Banyamulenge himself, to support the movement was an uphill struggle. Allegations of the targeting of Congolese Hutu civilians did not help to persuade potential supporters of the multi ethnic na ture of their cause. The AFDL in fact often used the opposite tactics, actively pursuing violence sided violence in 1996, the first year of the conflict (HSR 2012: 203). This was not surprisin g given their training and association with the RPF, a rebel group that had been accused (accusations later substantiated by Stam and Davenport 2009) of also pursuing wanton violence against civilians. This behavior is wholly different than the killing of strategy targeting non combatants and more an issue of collateral damage. The Dust Settles: The First Governance under AFDL, 1997 20 01 I now briefly turn to the system of government in Congo after the overthrow of Mobutu, and how the first AFDL government, which achieved power through military victory and the storming of Kinshasa, continued strategies, organizational types, and charac teristics developed during the rebel period into their new government. The Kabila State. Fresh from the horrors of the civil war and the legacy of
327 government took p ower of a chaotic state barely recognizable as a modern entity in May 1997, following a long march across the country. One of the priorities the new regime must put in place was basic governance, including law, order and public goods. Sadly, though, very l ittle changed (Reyntjens 2001: 317). The state remained in pieces, with some places seeing more government services than others, but generally remaining fractioned. The renewing of war a year after Kabila took power did provide for additional state buildin g activities, however, these came in the form of rebel state building by contra Kabila groups, like the Rassembleme nt Congolais pour la Dmocratie ( RCD, Rally for Congolese Democracy), lead by Ernest Wamba dia Wamba, a pr ofessor of History, and supported by Uganda who turned against former ally Rwanda in the course of the second war (Mampilly 2011). Internal Organization of the New Regime. When Kabila assumed power in Kinshasa, Congolese and International observers alike hoped that the government would be established quickly, fairly, and capably able to maintain the political or der that had long been absent. H owever, the disastrous, fractious and complicated internal politics of the AFDL continued to plague the organization as a government. Kabila, with no logical internal structure, pursued policies of personal rule, creating a cabal of leaders around him, inc luding a praetorian guard made up of kadago of his power remained small, and he instituted more and more divisive policies and arrests to prevent the AFDL from completely fracturing to the point of losing power (Reyntjens 2001: 314). also caused disruption within the fledgling regime (Nzongola Ntalaja 2002: 197), as
328 disputes over how m uch power Rwanda, Uganda, and later Angola and Zimbabwe should exercise over mining rights, governance, and economic benefits flared up. to intimidation of press an d the oppositions, the banning of political parties and factions and the concentration of power around Kabila (Reyntjens 2007: 307 308). regime rather quickly is unsurprising, given that during the rebel per cobbled from diverse interests, political philosophies, and styles. Kabila was also an incredibly fickle leader, willing to turn on friends, colleagues and foreign backers when it suited him, with l ittle regard to how this structure future affairs of state. By August of 1998, a new rebellion started, once again in the ungoverned East, which test ed the Civilians, Social Power, and the New Regime. Be cause the AFDL did not rely on civilian support to build legitimacy as a rebel group, strong relationships were never formed between the rebel government and those they governed. While the regime was not violent towards civilians in the same manner as duri ng the war, it did nothing to quell the culture of violence quickly becoming a m ainstay of life during the war. This severely goods and services to civilians. It was thus easier for new rebellions to emerge almost instantaneously, because civilians felt no loyalty towards the new regime and lacked physical control over vast amounts of territory where other rebel groups, like the RCD, could provide services and gain legitima cy. The RCD did this, creating a state within a state in the east (Mampilly 2011) that swept streets, provided security and political
329 order, facilitating loyalty. Many civilians also found the increasingly ethnic nature of the AFDL, whether real or imposed by Rwandan support to be highly problematic. By the time the AFDL took power in 1997, the group was understood to be entirely supportive of Banyamulenge causes to the detriment of others. This was in direct contrast to the disavowal of Tutsi support the A FDL government claimed once in power, leading to another dimension of conflict and new rebel organizations. Conclusions about Case Studies The additional study of rebel movements in the Great Lakes show the differences in outcomes depending of what type of organizational power and nature of social power the rebel movement pursued in its drive to political power. In the case of the NRA, infrastructural organizational power and inclusive social power ensured both the ability of the rebel movement to succeed in the civil war as well as building long term support for the post conflict government. The RPF also chose to use infrastructural organizational power in their desire to overturn the Rwandan government, but chose a strategy of exclusive social power and interaction. In using these types of power, the RPF was able to beat the Government of Rwanda in the course of the civil war and centralize political power after, but relationships with civilians remain rocky. The AFDL, which employed cellular organization al power and exclusive social power was able to win the civil war and reach political power in the DRC with the help of other rebel organizations and supportive governments, alt hough from its inauguration in 1997 experienced numerous problems, divisions, a nd inability to build strong relationsh ips with civilians. The next section tests these variables globally in a larger setting.
330 Testing Rebel to Ruler Transitions Globally Other rebel to ruler transitions are indicative of the applicability of testing my theory globally. In sub Saharan Africa alone, I compiled a dataset of sixty rebel groups operating from the years 1970 2011. In this section I discuss how I took the concepts of organizational power and the nature of social power, quantified them, and appl ied them to the data set elaborated on in Chapter 1 and provided in Table 1 1. The Dataset I select ed possible rebel groups through their inclusion in Uppsala incidences and actors of conflicts, state, non state, and one sided in the world since 1946. I then culled p ossible rebel movements as described in the Encyclopedia by fact checking major news outlets, journalistic accounts, contemporary research and previous scholarship for descriptions of the groups as rebels and to search for evidence of reform mindedness. 4 L ike Clapham I separated out and did not include rebel groups with specifically separatist aims, for while they might be loosely considered reformers, the ultimate goal of their effort is the creation of a new state and not a change in power or structure in the existing state. It is not enough to be simply identified as a rebel by the media. In this research, I take statements and accounts of rebels in media to be indicative of desires of state ch ange and reform. These desires need not be indivisible, but th eir presence is a necessary condition to be included in the dataset I divided rebel groups into their various factions where it was appropriate to do so, i.e. when one group was born from another. I then used the date of the split as the date of inception of the new group. This will be further illuminated when discussing post conflict outcomes and how factions have won or lost. This is acceptable in fitting with the conditions of the
331 analysis: not all factions behave the same, and many have different struc tures, type of organization and features Translating Variables One of the more difficult tasks here is translating the large breadth of qualitative research conducted here on rebel victory in the Great Lakes into a parsimonious and quantifiable rposes of statistical analyses. This first analysis tests the concept of state takeover as defined quantitatively. State takeover is, for the purposes of parsimony and explanation through comparable and available data, the winning of simple majority power (51%) in government following the cessation of battle related conflict. Thus, in this analysis there are three possible outcomes the rebel group failed in i ts challenge to state takeover (case 0), the rebel group took ove r the state through control delegated in a negotiated power sharing arrangement (case 1), or the rebel group took over the state through military victory (case 2) 5 T he figure described in Chapter 1 listing rebel movements active in sub Saharan Africa incl udes the outcomes of civil wars in which the rebel movements fought according to this terminology. Of the sixty observations, forty five cases are failures in that they did not gain majority political power, five cases are examples of case 1, where they wo n majority power after the implementation of a negotiated peace and/or power sharing arrangement, and ten are examples of case 2, where they won majority power through military defeat of the government forces. It seems fairly intuitive that the most desira ble outcome would be that of case 2, wherein which a rebel group challenges the state and gains victory through conflict, a victory scholars have posited is more likely to lead to post conflict and longer term, stability (Toft 2010b) In this case, rebels are more easily able to replace governments and policies with those of their own party or
332 inclination. Part of the assumption of reform mindedness implicit in the definition above is that the goal of the rebel group is to succeed at state power through in a manner like this, because of the weak nature of the state and state power in Africa, a phenomenon to which scholar s have devoted much attention The analysis that follows provides a first step in identifying characteristics of the rebel gro up that inf luence the ability of it to gain political control of a state The assumption is that these characteristics a) continue to affect rebel behavior in the post conflict time peri od and more importantly, b) ensure post conflict success for the rebel organization 6 While the subject of rebel behavior is not unexplored in the literature so far, the causal logic for post conflict transition and how these behaviors influence outcomes of civil conflict is. In keeping with an organ izational approach to post conflict and governance, I hypothesize the following : that organizational power matters to post conflict success of a rebel government, thus groups with a wider range of internal ly cohesive organization al structures and functions in the p re conflict period are able to more effectively take over the state that groups with hierarchal centralized leadership structures are mo re likely to be effective in achieving post conflict gains, including winning the war That groups with inclusive rela tionships with civilian populations matter to successful post conflict transitions, including winning the war These three hypotheses tested provide a preliminary answer to the que stion of whether groups with high levels of organizational power are more lik ely t o be effective in capturing the state and thus more effective in post conflict gover nance than cellularly
333 organized groups who are less able to build credibility or legitimacy While these hypotheses only in directly test the implication of rebel suc cess post civil war they do provide a promising first cut. Data Collection, Coding and Methods Th is section of the analysis detail s the data collection and statistical methods employed here While there has been much prior exploration of rebel, anti gov ernment or non state actors in conflict, there has heretofore been li ttle quantitative analysis focused on the impact of rebel behavior on conflict outcomes or post conflict legacies 7 Part of the reasoning behind the lack of previously collected evidence on rebels groups in sub Saharan Africa may be the sheer number of rebel groups i n existence since i ndependence and our inability to code rebel groups motives in a way that makes sense T he data I employ here does in fact, suffer from some of these problems, but provides a first attempt at testing these relationships in conjunction with th e qualitative data established. I determined the sample of non state actors for inclusion in the da ta set through inclusion in the UCDP Conflict Database Encyclopedia, which codes for non state actor, one sided and armed civil conflict in t he world from 1946 until 2010. This dataset is well known and well utilized (Gleditsch et al. 2002, Miguel et al. 2004, Haborm and Wallensteen 2005, 2007), making it appropriate for later elaboration and replication efforts building on these initial findings Once I identified possible rebel groups, I performed searches of news media outlets, including journalistic, statements of goals against participation in conflict. I also engaged in literature reviews of accounts of the rebel groups identified. I accepted the description of rebel behavior in media and publication as indicative of goal, whether pre hoc or ad hoc, during conflict. I
334 searched for corroborating evidence amongst various claims for rebel groups 8 and prioritized information that came directly from rebel sources, like statements of purpos e, party statements during elections, fieldwork accounts and historical records. Although I am somewhat skeptical of these statements, because rebels can and do self valorize after the fact, the qualitative analysis discussed in earlier portions of this ch apter (vis vis rebel organizations in the Great Lakes) as well as Chapters 3, 4, and 5 (which highlight specifics of the CNDD FDD) illustrate that rebel groups tend to follow through in some capacity on statements, motivations and goals described during the civil war. Thus I can reasonably accept such as evidence of their commitment to reform and change. Once I identifi ed these groups, I then gathered data on the independent variables. I proxy Rebel Strength with a me asure of the number of troop the rebe l group commands Numerical Years of Activity (Age of Rebel Movements) provides the count of how many years the group was active as a rebel movement (and not a government) Rebel Capacity is a measure of capacity and strength devised by Wood 2010 It compr ises the ratio of the number of rebel troops to the number of government troops, scaled for multiple anti state groups in conflict 9 For the most part, I used his ratios when available, but calculated them myself according to his logi c when they were not. I averaged across years of rebel group activity because in this particular set, the unit of observation is not group year, as in Wood, but rather group size aggregated across all years of activity This variable represents a proxy of the type of social pow er the rebel group possesses, because the measure of rebel capacity vis vis the state gives an expression of relationships with civilians and the building of support by amassing volunteers who could have otherwise joined the state army The variable for measure of
335 centralization and leadership is an ordinal variable coded for specific behaviors. A 1 value indicates a group with hierarchal structure and centralized leadership that projects power within the rebel organization. In reality, this means that a rebel group reports to a centralized command structure that is known to most, if not all, members, and that this leadership is accepted as the supreme head of the movement (centralization). The second dimension, hierarchy, implies that positions are subord inate and not parallel, and further, that positions are accorded by merit based rank, akin to a state army. These cases are best exemplified by the NRA of Uganda and the RPF of Rwanda, who took power through military victory in 1986 and 1994, respectively. These movements were well prepared in the bush, and took power against a more formidably equipped state foe, one of which was a former rebel mov ement (UNLA of Uganda) itself. The capacities of these groups were equal or greater than the capacities of thei r state foes. They also included former military soldiers which will be discussed later as a contributing factor for victory. A 2 value indicates a hierarchal group with decentralized command center, meaning that the group may have subordinate ranks and n ot parallel positions, but the leadership is divided amongst two or more heads with no supreme command. groups in Eastern Congo. These groups often splinter and exhibit less survivabil ity in post conflict 10 transitions. A 3 coding indicates a non hierarchal, centralized group. This would be a group like Palipehutu FNL in Burundi, where fighters are not ranked based on merit, there are parallel leadership roles, but yet, the movement has one central command figure or location. These are the rarest of all rebel groups in sub Saharan Africa. A 4 value indicates a non hierarchal, decentralized group. This would include
336 groups like Al Shabaab in Somalia, ancephalous in nature, difficult to det ermine membership and harder still to determine leadership. There are two variables in the dataset to account for alternative explanations and testing: Peacekeepers, which is a dummy variable coded for the presence of an international peace keeping force, and Outside Help, which is a dummy variable coded for support from another government or outside group. The dependent variable here is Outcome, a categorical variable coded based on results of the conflict, according to the case 0, 1, and 2 schema as descr ibed above. Summary statistics and example values for the major cases in the study are provided in Table 6 1 and 6 2 respectively. Using this dataset, I then test my initial hypotheses. I provide the results of a multi no minal logistic regression at the e nd of the chapter in Table 6 3 testing the possible outcomes of the ending of conflict wherein which the rebel group is victorious. The results of this analysis (Table 6 3 ) are quite interesting Rebel Capability, the ratio to measure strength in troops o f the rebel army versus strength in troops of the government army is statistically significan t in neither case of victory, settlement or military might. The measure of Centralization and Leadership is only weakly and negatively statistically significant wh en looking at cases of military victory, which is in fitting with the hypotheses. Less hierarchal organized and centralized groups are less likely to win state power through military victory. Surprisingly, rebel strength, the sheer number of anti governme nt troops, is weakly and positively statistically significant when speaking of groups that have won power through negotiated settlement, but not during cases of military victory.
337 Now, I perform the same regression analysis but add in an additional variabl e for age of the rebel movement. It could be the case that groups that succeed have had longer learning histories and struggles, and thus have learned by experience. This provides an additional robustness ch eck, illustrated in Table 6 4 at the end of the c hapter. The results of this analysis are somewhat confusing. Rebel Capability, the ratio to measure strength in troops of the rebel army versus strength in troops of the government army is statistically significant in neither case of victory, through settlement or military might. The measure of Centralization and Leadership is only weakly and negatively statistically significant when looking at cases of military victory, which is in fitting with the hypotheses and the results of the first regression an alysis Rebel Strength, the sheer number of anti government troops, is weakly statistically significant when speaking of groups that have won power through negotiated settlement, but not during cases of military victory. This is most likely due to the imme diacy of the threat large numbers of anti government forces can produce when sitting down to negotiate. The age of the rebel movement is weakly and negatively correlated with the outcome of negotiated settlement, indicating that older groups are less likel y to see victory in this manner. This could be, because like the Palipehutu FNL in Burundi, governments have become used to seeing these rebel movements as weak actors over time, and are thus less likely to give into demands or allow for any entrees into p ower. Now, I perform the same regression analysis but test for the most hiera r chally organized a nd centralized behavior factor, providing the results in Table 6 5 The results of this analysis are surprising, there are no statistically significant relation ships when
338 controlling for the most hierarchal and centralized groups. Rebel strength and capability seem to be insignificant factors in victory here. Logistic regression models also allow for the testing of alternatives to my theory by allowing for test in g for dummy variables. It could be the case that having outside help (i.e. foreign support through either material or territory) co victory, indicating a bolstering effect. I provide the r esults in Table 6 6 The results of th is analysis shed some light as to how outside help and influence can change the outcome of a r Rebel Capability and Rebel Strength are both weakly correlated with success in negotiated settlement victories when controlling fo r outside help. This is not surprising, as scholars have pointed to transnational effects of rebellion and costs of supporting a rebel movement from a neighboring country 11 But this support is not always a case of cross border co ethnicity or similar poli tical beliefs it can also be, in the case of the AFDL in Congo, a way to minimize security threats (Rwanda and Uganda) or gain financially (Angola). Rebel Capability is also positively correlated with Military Victory in this analysis, suggesting that out side help and material favor. This could be because outside help allows for rebel movements to better arm and prepare them or to gather more members from cross border refugees, like the case of the RPF in Rwanda, whose material, logistical and training support from the NRA government in Uganda was critical to winning the war. Finally, I test for the effect of the presence of international peace keepers. It could be th e case that having an international or regional peacekeeping body encourages rebel
339 victory t hrough negotiated settlements with guarantees of security protection I provide the results in Table 6 7 The results of this analysis indicate that no variables of the characteristics of the rebel group are significant when controlling for the presence of peacekeepers in the conflict. This is somewhat unsurprising, as one could logically expect that the presence of an armed body of peacekeepers would not contribute to the rebel group being able to win power militarily, but rather for conflict to stagnate or end (Walter 2002, 2004) Ho wever, it seems counterintuitive that peacekeepers would have no effect on victory through negotiated settlement, which might be more a ttainable if international forces are providing security and force necessary to ensure credible commitments, although this is untested here. Robustness Checks I ran additional tests on the robustness of my initial findings, as it seems likely that some of these variables suffer from the problem of collinearity. In the results, provided in Tables 6 8, 6 9, 6 10, 6 11, 6 12, 6 13, and 6 14, the robustness of my previous findings in the statistical evidence are confirmed. Summary Analysis and Possible Data Ch allenges The results of this analysis provide a first cut at the effects certain characteristics of rebel groups have on their prospects for successful state takeover through military victory or negotiated settlement. In all models except for the Peacekeepers and Centralization and Leadership ones, rebel strength, the physical number of rebel troops is statistically significant and positive for the Case 1(won through settlement) model. This could be for a variety of rea sons one, the more troops a group has, the more
340 effective their potential to intimidate, terrorize or persuade both government forces and civilians (Weinstein 2005, 2007). Age of the Rebel Movement, a variable added to the original hypothesized model, is weakly negatively significant in Case 1 victories, because governments might be more familiar and less willing to negotiate with groups that have been in existence longer. Centralization and Leadership is negatively weakly statistically significant in all models but those with controls, meaning that less centralized and less hierarchal groups are less likely to attain Case 2 military victories, although there is no effect on Case 1 victories. When controlling for outside assistance, Rebel Capability becomes a weakly statistically significant variable, meaning that groups with higher capabilities are more likely to see Case 1 or 2 victories when they have an outside ally or base. This raises serious questions about the abilities of rebels to rely exclusively 12 Although the relationships are only weakly correlated in this particular analysis, the door is nonetheless opened to exploring the relationship between capacity and organization fur ther. Although this set only has sixty observations, it could be expanded to include group behavior over time, changing the unit of observation from group aggregate to group year dyadic form. This might potentially capture changes in rebel behavior and e fficacy over time and could even be compared with battle deaths per year or other data to test other hypotheses about capability D ata collection on this topic at this juncture is in the early stages and requires muc h background analysis at the qualitative leve l, but this does not hinder my nascent conclusions The next step in the analysis would add in the determination of winning political power in an election one and two cycles post
341 conflict. The problem with the second phase of statistical testing in th is regard is that the sample sizes of successful rebel movements become too small to be considered significant, but are worth considering for their potential value added to the qualitative evidence provided in earlier chapters as well as future analysis. F inally, I wish to address the question of whether even temporally slight control of the state equals power. This analysis is severely limited to fully answer that question, and I accept the criticism that the outcome variable may not capture adequately the range of all outcomes. For example, there are many groups within the dataset that signed peace agreements and became political parties and legitimate actors who gained some control of the state, but not all o f it. For example, the CNDD Nday iken gu rukiye in Burundi, that hold seats in Parliament but not the entirety of state power. Chapter Conclusions In this chapter, I explored potential similarities and differences among rebel groups in the Great Lakes regions using conclusions reached in earlier chapters about the role of the prior state and its institutions, the type of organizational power the rebel organization possesses, and the nature of the relationships that it pursues with civilians and the prior stat e in the case of the CNDD FDD. The range of out comes illustrated in the Great Lakes show that rebels pursue different types of organizational power in their bids to take over power, and as such, their post conflict governments reflect these strategies. The differences among the RPF, AFDL, NRA, and CNDD FDD governments can be illustrative of how these rebel governments mature, and potentially demonstrate how they will continue to develop in the future. Rebel organizations with infrastructural power, that exhibit internal logic and discipline, like the RP F and the NRA, are able to
342 effectively dominate their states politically and militarily. While their governments may be accused of turning authoritarian, they nonetheless maintain control, hierarchy and power. Rebel organizations with cellular capacities, meaning they lack this internal cohesion while a rebel group, have a tougher time governing: they are often unable to bring disparate actors to the table and build new governments or cannot exhibit total sovereignty. Similarly, the comparative case studies show how rebel organizations pursued relationships with civilians and the pre existing state in the territories they now control. This also matters to post conflict governance, as well as the potential for new violence. Groups that aligned themselves wit h civilians during the war, both in using them as a valuable resource and providing for them, like the CNDD FDD and the NRA exhibit a different kind of stability and behavior in post conflict governance. In opposition, those groups who do not use civilian resources often find an uphill battle in connecting with those they are now responsible to. This can prove problematic in maintaining post conflict power without resorting to violence or authoritarianism to provide stability in the absence of civilian supp ort. In quantitative tests among rebel groups, these two variables and their effect on how the civil war ends and the potential shape of the post conflict government are borne out. More organized groups are able to win the war through military conquest, sh owing an inner cohesion that allows them to effectively control chaotic post conflict states. Also, groups with higher measures of rebel strength (a proxy for civilian relationships) are able to win the civil war through negotiated settlement, which helps them to gain power in the post conflict state. While the quantitative results require further research and refinement, they support the initial findings of this work and provide for a new application of rebel studies research.
343 Tables and Figures Table 6 1 Summary Statistics of African Rebels Dataset Table 6 2. Examples of Measures on Major Cases in Study
344 Tabl e 6 3. Multinomial Regression Results **Statistically Significant at the .05 level Table 6 4 Multinomial Regression Results with Addition of Independent Variable Age of Rebel Movement *Statistically Significant at the .10 level **Statistically Significant at the .05 level
345 Table 6 5 Multinomial Regression Results using the Proxy Measure for Organizational Power Table 6 6 Multinomial Regression Results holding Foreign Assistance constant *Statistically Significant at the .10 level **Statistically Significant at the .05 level
346 Table 6 7 Multinomial Regression Results, the Effect of Peacekeepers Table 6 8. Robustness Check on Centralization and Leadership Table 6 9. Robustness Check on Rebel Strength
347 Table 6 10. Robustness Check on Rebel Capability Table 6 11. Robustness Check on the Natural Log of Rebel Strength *statistically significan t at the .10 level *stastically significant at the .001 level
348 Table 6 12. Robustness Check on Centralization and Leadership and the Natural Log of Rebel Strength Table 6 13. Robustness Check on Rebel Capability and the Natural Log of Rebel Strength
349 T able 6 14. Robustness Check on Rebel Capability and Centralization and Leadership Notes 1 An interesting note about the terms Rwandese and Rwandan in most American difference between British English and American English that ultimately bears little significance, as the group answers to both names. It is telli ng that the group does not identify itself using the French acronym FPR (Front Patriotique Rwandais), although French speaking scholarship sometimes does. 2 Personal Interview, April 22, 2011 3 Although France has attempted numerous times to try Paul Kaga me in French court for the action.
350 4 I am well aware of the somewhat arbitrary method of identification of reform mindedness through news sources. However, I chose to accept rebel descriptions and statements by leaders at face value as indicative of desir e. 5 These designations are ordinal, not categorical. 6 These assumptions are merely posited at this juncture. 7 Which is not to say that the future will not bring more, especially with new analysis opened up by Weinstein 2007, Cyrus 2010, and others. 8 The same claim being made by more than two sources. 9 within a conflict system and then multiplied by government forces(G). This is represented by ((i/I)* G). He discusse s the problems with the assumptions of such a measure on the same page. We agree with his logic regarding the difficulties in relying on this measure but find it to be a useful tool for analysis. 10 At least as described in secondary literature. 11 See Saleh yan 2007 for a discussion of this. 12 Salehyan 2007 points to it in Iraq and Congo.
351 CHAPTER 7 CONCLUSION Introduction This study provides new findings that contribute to the expanding rebel studies and civil wars literature. In exploring the question of why some rebel groups win control of the state following the cessation of conflict and continue on to become durable post conflict governments, several new developments emerge. These developments specifically focus on how rebel organizations can build both internal legitimacy and strength through organizational type and strategy as well as through relationships with pre existing stat es and civilians. Rebel organizations then draw upon these organizational and social powers before, during, and after violent conflict to gain political control of the state, thereby becoming governors themselves. In this concluding chapter, I summarize th e findings of the research conducted on the CNDD FDD ( Conseil National Pour la Dfense de la Dmocratie Forces pour la Dfense de la Dmocratie ), the AFDL ( Alliance des Forces Democratique pour la liberation du Congo Zaire ), the RPF (Rwandan Patriotic Fron t) and the NRA (National Resistance Army) and provide a final overview of the contributions to the discipline and literature. I then discuss the future implications of the conclusions of this research and potential expansion. Guiding Theory : I began this w ork with a statement of proposed theory as to how rebel organizations take power from states they fight against and eve ntually hold power themselves. My theory holds that the post conflict behavior of rebel organizations is not dependent upon the way in wh ich the war was won militarily, or the size or composition of the rebel group, but rather, that characteristics of the rebel organization
352 matter to the way and manner in whic h the rebels achieve post conflict power. Specifically, I claimed that three variables mattered to both civil war and post conflict governance outcomes: First, that the type of organizational power the rebel group possesses its ability to maintain intern al order and discipline, project power over a territory under its physical control, an d centralization of authority strongly influences the likelihood of victory. Second, that the nature of social power that the rebel group pursues its relationships with civilian populations and the pre existing state strongly influence the likelihood of victory in combat and post conflict transitions. Inherent in this second hypothesis is the existence and shape of the pre conflict state (a third intervening variable th at acts upon both organizational and social powers), its institutions and structures. The state and its development are important to rebel organization in sub Saharan Africa because it structures the way rebel organizations, and those citizens of the state that comprise them, are socialized into understanding authority and legitimacy. Testing the Theory In this work, I tested the theory laid out above at several levels first, by providing qualitative accounts garnered through interviews with rebel partici these data to compare how rebel movements in the geographic region closest and most similar to Burundi fared in their armed struggles against state actors, focusing on the d ata gleaned about the transformation of the CNDD FDD and how these variables are present in other organizations. Finally, I presented a dataset of Reformist rebels in post
353 Independence Sub Saharan Africa, of which the CNDD FDD, RPF, NRA, and AFDL belong to showing how these variables apply in other contexts. In this analysis, I showed how in the case of the CNDD FDD, unable to win the civil war through combat methods due to a lack of coherency, internal discipline, and order (what I term organizational po wer), the rebel organization developed a program of inclusive social relations in addition to subverting, transforming and recreating state structures to ensure that civilians would support the group long after the civil war ended. This support mattered d uring the war, for although the rebel organization was unable to coordinate battle plans and thoroughly rout the army of the Government of Burundi, civilians still understood which organization provided them protection, gave them rule of law, and fostered positive relationships. Although Kriger argued that guerillas are able to win conflict while depending on coercion to gain civilian support (1992: 238 and Kasfir 2002: 1), winning the war was not the primary function of these interactions. These interactio ns served the CNDD FDD well in planning for a post conflict political future, and continue to remain a vital part of their strategy in maintaining power. Thus, the argument I make helps bridge the gap between the history and development of a rebel organiza tion and its post conflict trajectory, showing how behaviors established and perfected during the rebel period matter to if and when the rebel organization achieves post conflict political power. I then extended the study of these behaviors to rebel groups operating in the immediate geographic area. Across the region, a surprisingly high number of similar transitions from rebel organization to government occurred. The RPF in Rwanda, the state with the most pre conflict similarities to Burundi in terms of et hnic makeup and power structures, experienced a different pathway to power than the
354 CNDD FDD. There, the RPF managed to take political power by strength of force by focusing the goals and strategies of the rebel organization not on civilians or available s tate structures (as was the case with the CNDD FDD) but rather, on strict internal discipline, logic and cohesion that created a comprehensive combat strategy that allowed for military victory over a larger and more well equipped enemy. The State and Rebe llion In Chapter 3 of this research I provided an explanation of how the nature and shape of the state before and leading up to conflict influenced the behavior of the rebel organization. The state influences the type of organizational power a rebel grou p possesses because a) experiences with the military as an institution of absolute power in the state can determine what patterns of authority the rebel group will pursue, and b) state domination and order may be tied to a historical socialization process that rebel groups are unlikely to outgrow, as in the case of the RPF in Rwanda. Moreover, the historical trajectory of the pre conflict state may determine important aspects of the civil war that go on to influence rebel behavior. In Burundi, four aspects of the state were discussed: two pre Independence and two post Independence. These four factors ethnic structures and the degree of state centralization or coherence before 1962, and political instability and the late 1980s transitions as a result of inte rnational restructuring of aid in the post Independence period fundamentally shaped the nature and cause of rebellion there. And because state histories are similar across the Great Lakes region, the chapter then compared these factors to the shaping and outcome of rebellion in Rwanda, Uganda, and the DRC as well. Although these four countries share similar state conditions that led to civil war, the outcomes of their rebellions differed greatly. Although Rwanda and Burundi share
355 the most similar histories the more authoritarian nature of the Rwandan state and the lasting legacy of relationships cemented as ruler and ruled there led the RPF to become a highly structured, disciplined and more organized rebel group than the CNDD FDD, in which authority reste d in multiple hands and organization and loyalty were far more lax. In Uganda, multiple sovereignty in the pre colonial period followed by authoritarian rule post Independence meant that the NRA also followed hierarchal patterns of leadership and strategy during the Ugandan Bush war, ultimately winning Kampala and the governmen t through military domination. In the DRC, a history devoid of any real political or social organization meant a continuation of the old style of dys function and competing agendas. Th e state clearly left its mark on all of these rebel organizations, as well as the civilians who lived under it and then forged (or did not) relationships with as well as the nature of social power the rebel organizations pursue, it is an intervening variable in this analysis. The Internal Logic of a Rebel Organization: Type of Organizational Power In C hapter 4 of this work I explore how the organizational power chosen and implemented by the CNDD FDD affected its capability to carry out a civil war against a much larger and more capable enemy. The type of organizational power utilized by the rebel group, infrastructural, mea nt that the group was often disorganized, prone to fractionalization, and unable to exercise strength through hierarchical relations and discipline. This made for a longer civil war, tenser negotiations, and a distinct possibility of non victory. Although the Burundian state was far more centralized than other states on the continent, it still experienced a high degree of consociationality and regional
356 tension in its pre war governance, so it is not altogether unsurprising that the CNDD FDD reflects some of this legacy. Thus, in this chapter I explored both the imprint of the state left on the rebel movement in addition to the organizational power characteristics of the group. I also showed how the CNDD FDD continued the patterns of behavior implemented whil e they were a rebel organization when they took the government followi ng the end of the war in 2005. Decentralized authority, splits among members, and problematic discipline plague the newly formed political party, although they have maintained political power through the 2010 election cycle. I also examined the CNDD Palipehutu FNL to test for in country variation. The Palipehutu FNL also experienced many problems with organization and leadership, also showing inf rastructural organizational power. This is largely due to the way in which Burundian history, state and political development shaped the nature of conflict there. Rebel groups emulated previous political developments in Burundi by allowing personal and geo graphic rivalries to determine the trajectory of these rebel movements, undermining any potentially successful military campaign that would have taken B ujumbura during the war. The CNDD FDD, however, was able to win political power by employing another fac et of the rebel organization to better use, that of their loyal networks of civilians who supported first the war effort, and then political power.
357 Civilians, States and War: Social Power and Rebellion In Chapter 5 I explored an alternative ro ute that rebels can take to post conflict success, that of rebels utilizing social relationships with people and the states they live in to sustain political power and change. Inclusive relationships with civilians and the state the rebels operate in creat e lasting legacies serving the rebel organization well: Not just supporting the insurgency materially, tangentially, and philosophically during the civil war, but also working towards building legitimacy and credibility in a future rebel controlled state. These types of interactions can provide long term benefits for all actors, although they may be skewed toward the rebel organizations. Civilians benefit from security services, local governance structures, and food and materiel benefits during the war. Reb el organizations benefit from civilian loyalty and philosophical support, the passing of information and keeping of secrets from the hostile state actor during the war, as well as the ability to use civilian resources for their own purposes during the war to finance and prolong the battle campaign. If the civil war should go to a stalemate then proceed to settlement talks, as many contemporary civil wars do, these relationships can provide much needed pressure on actors to settle disputes and hold elections especially if the state government involved in the civil war feels citizens turning against them and wants to maintain some post conflict political power. After the civil war ends and government is restored, civilians can then be mobilized for voting in post conflict elections, providing the party structure and support networks, and serving as a protesting force if necessary. Both the CNDD FDD and the Palipehutu FNL utilized this resource while operating as rebel organizations during the Burundian Civil W ar from 1993 2005. The CNDD FDD was more successful in building these relationships with
358 civilians than the Palipehutu FNL because they did not practice discriminatory ethnic practices, did not use violence against civilians as a tool of regular warfare, a nd exercised in a wider geographic area. The evidence provided by studying the behaviors of these two rebel organizations shows how rebel groups have strong incentives to utilize civilians and pre existing states if possible, and how these can provide a di fferent pathway to power instead of relying stri ctly on fighting capabilities. This initial definition organizations in the Great Lakes region who may not have utilize d the same strategies during their respective civil wars. Rebels in Comparative Context In Chapter 6 I focused on how the experiences and development of the CNDD regio nal neighbors: Rwanda (the RPF), the Democratic Republic of the Congo (the AFDL), and Uganda (the NRA). By expanding the geographic area of study, I show that rebels in Sub Saharan Africa and in the Great Lakes region are neither wholly unique nor wholly s imilar in their transitions from revolution to regime. In the case of the RPF, The RPF employed infrastructural organizational power to overthrow the Habyarimana state, a brutal civil war that lasted from its inception in 1990 until the genocide that saw a million civilians massacred in 1994. They were unwilling and unable to adopt p ositive relationships with civilians of either Hutu or Tutsi ethnicities, conflict governance, this has translated to an incredibly strong, centralized state power
359 apparatus, but an authoritarian domination over civilians. In the case of the NRA in Uganda, infrastructural organizational power and inclusive social power ensured rebel success not only in the Ugandan Bush War, but also the building of long term support for the post c onflict government, although this has waned over time as the NRA/NRM government has sought to control total political power in Uganda for the last 26 years. As for the AFDL in Congo, the use of cellular organizational power and exclusive social power did n ot prevent the rebel organization from winning the civil war initially and reaching political power in the DRC: The help of other rebel organizations and supportive governments in the region, especially Rwanda and Uganda, provided enough material and resou rces to overcome these challenges. However, from its inauguration in 1997 the group experienced numerous problems, divisions, and inability to build strong relationships with civilians. This chapter also explored the implications of the data gleaned from the intensive fieldwork on the CNDD FDD in an even more comparative context: testing the variables organizational power and the social relationships rebel groups built to see if patterns or significance emerged across 60 rebel groups in sub Saharan Africa operating from 1975 2009. Although the initial tests of the two variables only looked at how the civil war ends, the potential shape of the post conflict government are borne out by the relationships tested and shown. More organized groups are able to win the war through military conquest, illustrating an inner cohesion that allows them to effectively control chaotic civil wars as well as the potential for governing post conflict states. Groups with higher measures of rebel strength (a proxy for civilian re lationships) are able to win the civil war through negotiated settlement, which helps them to gain and maintain power in
360 the post conflict state. While these quantitative tests require much further research and refinement, they support the initial findings of the earlier chapters and provide for new applications of work in this area. General Contributions to Academia the Discipline This research contributes to the discipline of political science as well as the study of Sub Saharan Africa in several innovat ive ways. First, by adding to the methodological debates presently occurring about the nature and viability of mixed method research and secondly, by incorporating anthropological, historical and other disciplines into political science promoting a more ho listic research agenda and a focus on African Studies By integrating a mixed methodological approach that uses qualitative interviewing ethnography and participant observation as well as quantitative data studies, the project highlights how they compleme nt each other to provide more complete evidence of the likelihood of my theory. This corresponds well to recent innovations in studies grounded in field work and confirmed in data testing put forth (Lindberg 2010,Varshney 2008, Kalyvas 2006) The research also contributes to the enlargement of interdisciplinary studies, incorporating methods, concepts, and approaches from other social sciences, notably history, anthropology, psychology, and sociology. By utilizing these, my project contributes to work withi n African studies and these disciplines, and thus fit across a broad range. Beyond these general acade mic contributions, my study also adds to the study of political science. By focusing on understudied aspects of political behavior, the linkages between studies in comparative politics and other sub fields in political science can be forged. Instead of treating rebels and other non state actors as non rational or
361 unimportant contributors to the governance of the state, this work shows that rebel organizations have goals, make decisions (whether rational or not is another matter), and become important players in both domestic and inc reasingly international arenas. The research also contributes to bringing in understudied regions and areas into discussions of broader phenomena, using case study examples to solidify, refine and even criticize existing concepts and theories in political science. For example, the study wholly critiques theories about war and democratization whereas some see these as separate phenomena that can be imposed at will without regard to previous conditions (see Bermeo 2003 for a lengthy discussion), this works s hows that behaviors during civil wars matter to post conflict outcomes, and that governance styles of rebels can be traced genealogically back to these. Contributions to the Specific Civil Wars/Rebel Literature In this section, I detail what specific con tributions the research makes to the subset of political science literature on civil wars and rebel stud ies. As described in Chapter 1 while Civil War studies enjoy a prominent and seemingly permanent focus in the discipline of Political Science, only in the very recent past has the subset of Rebel Studies within this broad er literature come into vogue. Beginning with Clapham (1998) and Reno (1997), studies on civil wars moved from macro dynamics (understanding wide scale and philosophical questions about the nature of rebellion) to focusing more on specific questions in rebel studies, i.e. types and distinct goals of different kinds of rebellion, emphasis on specific actions pursued, means of support and authority building, an d rebel civilian interactions. This research fits squarely within these new literatures on rebellion, civil war, and revolution and political change. It does so by
362 examining cause and effect of rebel behavior by distinct time periods before, during and after political violence, to sho w that these actions are more interconnected than the scholarship may express. The work provides an extension of rebel behavior beyond the conflict itself, showing that rebellion is a process of reform that may manifest itself through violence, but consequ entially and fundamentally changes political processes in the state in which it occurs. Generating Future Tests These first tests of theory can also offer a way forward for tests of rebel behavior, whether they ar e case studies or quantitative data collec tion. I plan to expand the research design to include other qualitative work on other suitable cases both inside and outside of Sub Saharan Africa. Some potential cases to draw more data from include the Liberation Tigers of Tamil Eelam (LTTE) or Tamil Ti gers, in Sri Lanka, who were defeated in 2009 by the Sri Lankan state. The LTTE set up a shadow state in parts of Sri Lanka under its territorial control (Mampilly 2011), but was not able to effectively win the civil war. Thus, additional tests of partial control or political power could show how important these variables are, when they become crucial, and what other factors might matter. Other cases, like the Zimbabwe African National Union (ZANU), might illustrate how ideology (in this case, a liberation/ independence struggle) might affect post conflict political outcomes. I also have plans to expand the dataset in several ways. The first is to test the implications of the variables in the post conflict setting. I can do this by running multinomial logi stic regressions with the outcome variables based on post conflict elections and the rebel groups winning or losing of the elections. Additionally, I could
363 turn thi s into a few journal articles in the immediate future. Future Implications and Research As previously elaborated, the theory presented here offers several new avenues for future work on rebels, rebellion, and post conflict transitions. Across sub Sahara n Africa, approximately twenty governments have come to power after rising from a rebel group, and new political developments in the Sahel in 2011 and 2012 may increase this number in the coming years. The implications of this research can provide a way fo rward to test other transitions in sub Saharan Africa, especially in terms of ideological or ethnic dimensions of conflict. That being said, sub Saharan Africa is not the only continent to experience political transition in the form of civil war and rebell ion. I put forth a theory that holds no specificity to African transitions, and thus could have potential implications for other rebel turned rulers, in Latin America (as in the case of the Sendero Luminoso in Peru) or in Southeast Asia. Testing the theor ies in other geographic areas also provides a test of regional effects. If in fact, rebellions are contagious (Ulfelder 2012), then it might matter as to how contagion happens differently (or not) in various regions. Manuscript Development I plan to expa nd on the research descri bed above to comprise my fir st book following the completion of my degree While the main thrust and hypotheses of the work will remain integral, new chapters will be added on emerging rebel movements that complement the original r esearch. Specifically, I aim to expand the components of the research on the RPF and the AFDL into full chapters, conducting more field research in Rwanda and the Democratic Republic of the Congo. I also plan to
364 lengthiest time in office post conflict. This research will further illuminate how long rebels continue to maintain programs and plans developed while the group was still a rebel organization into its time in governance and state power. This is in fitting with additional plans for the book: The book project also advances the initial findings of the strength of the pre rebel state, the organizational power a nd source of social power of the rebel organizations by testing the argument far removed from the immediate post over a longer history of rebellion, as in the South Sudanese SP LA. The original dataset will also be expanded beyond the borders of sub Saharan Africa, including rebel groups (the notion that characteristics only fo und in African states make African cases a wholly unique set) in the broader context of rebellion and change in political order By expanding the dataset to other geographic areas, potential geographic particularities owing to specific histories and post colonial experiences may be illustrated, further allowing for refining and testing of the theory presented in this study Conclusion This brief conclusion attempted to consolidate the bread th of material presented in my work by focusing on the k ey points, concepts and theory of how rebel organizations in the Great Lakes region of Central Africa have successfully transitioned from anti government actors to government power holders. My fieldwork and analysis presented contribute to the growing rebe l studies literature and will continue to expand in my career as I apply theory generated here to other cases and contexts.
365 A PPENDIX A TIME LINE OF THE BURUNDIAN CIVIL WAR October 21, 1993 Assassination of Ndadaye; Hutu groups begin to gather November 16,1993 UN Special Envoy dispatched to Burundi January 1994 Election of Cyprien Ntaryamira February 1994 Plans for GEDEBU militia made in exile in Belgium Spring 1994 Development of Hutu defense groups in Northern parts of Bujumbura April 6, 1994 Death of Transitional President Cyprien Ntaryamira in plane crash, Kigali, Rwanda Summer 1994 Development of FDD in Bukavu with Hutu former military officers of the Burundian Army September 1994 Election of Sylvestre Ntibantunganya November 21, 1994 Official Christening of Group with combinatory name, CNDD FDD 1994 1995 Collaboration between FNL, FROLINA, and CNDD FDD against FAB May 1996 Beginning of Peace Process in Mwanza, Tanzania June 15, 1996 Arusha Accords Summit convenes July 1996 Pierr e Buyoya retakes power from transitional government of Burundi July 31, 1996 Regional Summit and Sanctions Imposed on Government and Rebel Forces 1996 1997 Talks with Buyoya and CNDD 1997 1998 Group of CNDD FDD fighters under Jean Bosco Ndayiken gu rukiye goes to fight against Congolese militant groups together with RPF elements and AFDL May 1998 Splitting of CNDD FDD into CNDD FDD Nyangoma and CNDD FDD Ndayiken gu rukiye June 15, 1998 Official resumption of Arusha Accords Summit
366 January 1999 Sanctions officially end August 28, 2000 First Peace Agreement signed at Arusha No Rebel Participation Late 2001 Ousting of Jean Bosco Ndayikenrukiye by Peter Nkurunziza and Hussein Radjabu November 1, 2001 Launch of three year t ransitional government October 2003 CNDD FDD signs Pretoria Agreement, effectively ending violent conflict June 2005 First post conflict elections; CNDD FDD and Pierre Nkurunziza win
367 APPENDIX B LIST OF ACTORS MENTIONED CNDD FDD: Jean Bosco Ndayikengurukiye Leonard Nyangoma Pierre Nkurunziza
368 Manasse Nzobonimpa Hussein Radjabu Agathon Rwasa Government of Burundi: Pierre Buyoya
369 Cyprien Ntaryamira Key International Nelson Mandela Julius Nyerere
370 Jacob Zuma Congolese Actors: Laurent Desire Kabila Joseph Desire Kabila
371 President Mobutu Sese Seko
372 LIST OF R EFERENCES ACLED C odebook. 2011. Version 1. Dublin: Trinity College. Latin American Politics and Society 48, no. 4: 137 162. Amaza Ondoga ori 1998. Kampala: Fountain Publishers. Autesserre, Severine. 2011. The Trouble with the Congo : Local Violence and the Failure of International Peacebuilding. New York: Cambridge. Baker, Bruce 2004. Uganda 1981 International Journal of the Sociology of Law 32: 333 348. Barth, Frederick.1969. Ethnic Groups and Boundaries: The Social Organization of Cultural Difference. Boston: Little an d Brown. Beardsle y, Kyle and Brian McQuinn. 2009 Rebel Groups as Predatory Organizations: Journal of Conflict Resolution 53, 4: 624 645. Berdal, Mats and Malone, David.2000. Greed and Grievance: Economic Agendas in Civil Wars Boulder: Lynne Reiner Global Governance 9, no. 2: 159 177. Boas, Morten and Kevin Dunn. 200 7. African Guerillas: Raging Against the Machine. Boulder: Lynne Reiner Boone, Catherine. 2003. Political Topographies of the African State: Territory Authority and Institutional Choice New York: Cambridge. Boshoff, Howard. 2010. The Burundi electoral process: a time to act", Institute for Security Studies, 23 July 2010, Available online < http://www.polity.org.za/article /the burundi electoral process a time to act 2010 07 23 (accessed 15 March 2012). Brady, Henry and David Collier. 2004. Rethinking Social Inquiry: Diverse Tools, Shared Standards Lanham: Rowman and Littlefield. Cederman, Lars Erik, Brian Min and Andreas Wimmer. "Ethnic Power Relations
373 dataset". Accessed 2011/04/21. Available online at http://dvn.iq.ha rvard.edu/dvn/dv/epr/faces/study/StudyPage.xhtml?studyId=3658 2&studyListin gIndex=1_97e4560afdf3e4074f22e4414bc5 Chandra, Kanchan. 2001. Ethnic Bargains, Group Instability, and Social Choice Theory PS: Politics and Society 29, no. 3 : 337 362. Chretien, Jean Pierre.2003. The Great Lakes of Africa: 2000 years of History translated by Scott Straus. Zone: New York. Clapham, Christopher, ed. 1998. African Guerillas. Oxford: James Currey. -----------------------------. 2002. Africa and the Inter national System: The Politics of State Survival. Cambridge: Cambridge. CNDD FDD Internal Party Records. Accessed 13, 14, 15 November 2010. Collier, David, Jody LaPorte and Jason Seawright. 2012. "Putting Typologies to Work: Concept Formation, Measurement, and Analytic Rigor" Political Research Quarterly 65, no 1: 217 232. World Bank Conference on Development Economics, Washington, D.C., April 20 21, 1998. Collier, Paul and Hoeffler, Anke,1998. On Economic Cause s of Civil War ." Oxford Economic Papers 50, no. 4: 563 573. -----------------------------------------Oxford Economic Papers 56, no 4: 563 595. Collier, Paul, Mans Soderbom and Hoeffler, Anke .2001. Policy Research Working Paper Series 2681, The World Bank. Collins, Barrie. 1998. Obedience in Rwanda: A Critical Question. Sheffield: Sheffield Hallam University Pres s. Commonwealth Law Bulletin 14: 566. Daley, Patricia. 2007. Gender and Genocide in Burundi: The Search for Spaces of Peace in the Great Lakes Region. Bloomington: Indiana University Press. Daniels,Morna, compiler. 1992 Burundi :World Bibliographical Series. Oxford: Clio Press.
374 Davis, Diane and Anthony Pereira. 2003. Irregular Armed Forces and their Role in Politics and State Formation. New York: Camb ridge Day, Christopher. 2011 Comparative Politics 43: 439 461. Journal of Peace Research 41, no. 3: 303 320. Des Forges, Alison. 1999. Leave None to Tell the Story: Genocide in Rwanda. New York: Human Rights Watch. De W itt e Ludo. 2008. The Assasination of Lumumba. London: Verso Publishing. De Zeeuw, Jeroen. 2008. From Soldiers to Politicians: Transforming Rebel Movements after Civil War. Boulder: Lynne Reiner Diamond, Larry, ed. 1992. T he Democratic Revolution: Struggles for Freedom and Pluralism in the Developing World. New York: Freedom House. Doyle, Michael W. And Nicholas Sambanis. 2006. Making War and Building Peace. Princeton: Princeton University Press. Eichstaedt,Peter 2011. Consuming the Congo. Chicago: Lawrence Hill Books. EISA 2010. Burundi: 2010 election boycott. Available online. < http://www.eisa.org.za/WEP/bur2010candidate1.htm > (accessed 11 March 2012). Africa? Understanding and Pre Journal of African Economies 9, no. 3: 244 269. Eller, Jack David. 1998. From Culture to Ethnicity to Conflict: An Anthropological Perspective on International Ethnic Conflict. Ann Arbor: University of Michigan. Elman, Colin. 2005. International Organization 59, no. 2: 293 326. Englebert, Pierre. 2009. Africa: Unity, Sovereignty, and Sorrow Boulder: Lynne Reiner Fearon, James D and David International Organization 54, 4: 845 877.
37 5 --------------------------------------------. American Political Science Review 97 1:75 90. ----------------------------------------------.2008. the 2007 Annual Meeting of the American Political Science Association. Journ al of Peace Research 41 no. 3: 275 301. Fortna, Virginia Page. 2008. Does Peace Keeping Work? Princeton: Princeton University Press. Freedom House.2009 Report: Freedom in the World 2009 Washington, D.C.: Freedom House. Fukuyama, Francis. 2004. State Building, Governance and World Order in the 21 st Century. New York: Cornell University. Gahama,Jean. 19 83. Le Burundi sous Administration Belge. Paris: Karthala. Gasana, Jean Marie and Henri Boshoff. 2003. Burundi: Critical Challenges to the Peace Process. Institute for Security Studies African Security Analysis Programme Situation Report: Cape Town: ISS. Greed and Grievance: Economic Agendas in Civil Wars ed. Mats Berdal and David M. Malone. Boulder: Lynne Rienner: 157 172. Gebre Medhin, Jordan. 1989. Peasants and Nationalism in Eritrea : a Critique of Ethiopian studies. Trenton: Red Sea Press. Geertz, Clifford. 1963. he Integrative R evolution: Primordial Sentiments and Civil Old Societies and New States. ed. Clifford Geertz. Glencoe: Free Press : 105 157. --------------------.1973. The Interpretation of Cultures: Selected Essays New York: Basic. Gellner, Ernest.1983. Nations and Nationalism. Ithaca: Cornell University Press. American Political Science Review 98, no. 2: 341 354.
376 Gersony, Robert (August 1997). "Results of a field based assessment of the civil conflicts in northern Uganda" (PDF). The anguish of northern Uganda Kampala, Uganda: USAID Retrieved 2009 03 18 Gilligan, Michael et al. 2010. Reintegrating ex rebels into civilian life:Evidence from a quasi experiment in Burundi CAPERS Working Paper. Ghoshal, Neela. 2010. New York: Human Rights Watch. Available online. < http://www.hrw.org/reports/2010/05/14/we ll tie you and shoot you 0 (accessed 10 November 2010)>. Gleditsch, Nils Petter, Peter Wallenstein, Mikael Eriksson, Margareta Stollenberg and Hvard Strand Journal of Peace Research 35, no. 3: 615 37. Gourevitch, Phillip. 1998. We wish to Inform you that Tomorrow we will be Killed with Our Families. New York: Macmillan. in Conflict and Ethnicity in Central Africa Goyv aerts, Didier, ed. Tokyo : Tokyo University Press Gurr, Ted Robert.1970. Why Men Rebel Princeton: Princeton University Press. ---------------------.1980. Handbook of Political Conflic t : Theory and Research. New York: Free Press. Harbo m Lotta and Peter Wallensteen .2 005. Dimensions 1946 Journal of Peace Research 42, no. 5: 623 635. ---------------------------------. 2007. Journal of Peace Research 44, no. 5: 623 634. Herbst, Jeffrey. 2000. States and Power in Africa. Princeton: Princeton University Press. -------------------.2004. Journal of Peace Research 41, no. 3: 357 369. Hintjens Helen. 1999. Journal of Modern African Studies 37, no. 2: 241 2 86. Hochschild, Adam. 1998 : A Story of Greed, Terro and Heroism in Colonial Africa. New York: Houghton Mifflin. Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: UC Press.
377 Horowitz, Irving Louis.2002. Taking Lives: Genocide and State Power 5th edition. New Brunswick: Transaction. Human Rights Watch (HRW) 2010. Burundi: World Report 2010 New York: Human Rights Watch. Available online < http://www.hrw. org/world report 2010/burundi >. Human Rights Watch. 2011. Burundi: World Report 2011 New York: Human Rights Watch. Available online < http://www.hrw.org/world report 2011/burundi >. Human Righ Human Rights Watch. Available online http://www.hrw.org/news/2011/09/20/burundi investigate deadly bar shooting Human Rights Watch. 2012. Burundi: World Report 2012 New York: Human Rights Watch. Available online < http://www.hrw .org/world report 2012/burundi >. Human Security Report (HSR) 2012. The Human Security Report 2012. Available at < http://www.hsrgroup.org/human security reports/hum an security report.aspx > Humphreys, Macartan and Jeremy Weinstein. 2008. Who Fights? The Determinants of American Journal of Political Science, 52, no. 2: 436 55 Huntington, Samuel P.1968. Political Order in Changing Societies New Haven: Yale University Press. Hyden, Goran. 1980. Beyond Ujamaa in Tanzania : Underdevelopment and an Uncaptured Peasantry Berkeley: University of California Press. -----------------. 1986. No Shortcuts to Progress: African Development Management in Perspective. Berkeley: University of California Press. ------------------. 2006. African Politics in Comparative Perspective. New York : Cambridge University Press. International Crisis Group (ICG). 1 December 2000. Burundi: Neither War nor Peace. Africa Report No. 25. Brussels: ICG. ----------------------------------. 6 August 2002. The Burundi Rebellion and the Ceasefire Negotiations Briefing No. 9. Brussels: ICG. ----------------------------------. 28 August 2007. Burundi: Finalising Peace with the FNL Africa Report No. 131. Brussels: ICG.
378 ----------------------------------. 30 July 2009. To Integrate the FNL Successfully. Africa Briefing No.63. Brussels: ICG. ----------------------------------. 21 March 2012. Burundi: A Deepening Corruption Crisis. Africa Report No. 185. Brussels: ICG. --------------------------------.5 April 2012. Uganda: No Resolution to Growing Tension s Africa Report No. 187.Brussels: ICG. Juridical Journal of Modern African Studies 24, no. 1: 1 31. The Chicago Sun Times: 10. Johnston Patrick. 2007. Civil Wars 9, no. 4: 359 277. Jones,Bruce.2001. Peacemaking in Rwanda: The Dynamics of Failure. Boulder: Lynne Reiner Jones, Cara. 2010. cessation ist Movements in Sub the Internationa l Studies Association. Kaarsholm, Preben. 2006. Violence, Political Cultu re and Development in Africa. London: James Currey. Kabwegyere, Tarsis. 1995. The Politics of State Formation and Destruction in Uganda. Kampala: Fountain Publishers. Kagame, Alexis. 1952. Le Code des Institutions Polititiques de Rwanda Brussels: IRCB. --------------------.1959. Brussels: ARSC. Kaldor, Mary.1999. New and Old Wars: Organized Violence in a Global Era. 1st edition. Palo Alto: Stanford University. Kalyvas, Stathis 2006 The Logic of Violence in Civil War Princeton: Princeton University.
379 Kasfir, Nelson. 2002 National Resistance Army in Uganda, 1981 California, Los Angeles. -------------------.2005. Guerrillas and Civilian Participation: The National Resistance Army in Uganda, 1981 86," Journal of Modern African Studie s 43, no. 2: 271 296. African Studies Quarterly 11, no. 4: 1 20. Kaufman, Stuart. 2001. Modern Hatreds: The Symbolic Politics of Ethnic War. Ithaca : Cornell. Kay, Reginald. 1987. Burundi Since the Genocide New York : Minority Rights Group. Keister, Jennifer. 2011. States within States:The Social Contracts of Armed Groups PhD researc h University of California, San Diego. Kilcullen, David. 2009. The Accidental Guerilla : Fighting Small Wars in the Midst of a Big One. London : Oxford University Press. Kiley, Sam.1994. UN Acts to limit Burundi flare The Times London. King,Gary, Robert Keohane and Sidney Verba. 1994. Designing Social Inquiry. Princeton: Princeton University. Kinzer, Stephen. 2008. A Thousand Hills and the Man who Dreamed It New York : Wiley. Informal repression, zero sum politics and late third wave Journal of Modern African Studies 38, no. 3: 383 405. Kriger, Norma J. 1992. Cambridge: Cambridge University Press. Krueger, Robert and Kathleen Krueger.2007 From Bloodshed to Hope in Burundi: Our Embassy Years during Genocide. Austin: University of Texas. Laely, Thomas. 1997. Journal of Modern African Studies 35 no. 1 : 695 716. La Renoveau 1 December 1993. Print Edition. Bujumbura, Burundi.
380 Lahai, John. 2010. Analysis of Peace and Conflict Review 4, no. 2: 1 17. Laitin, David.1992. Hegemony and Culture: Politics and Change among the Yoruba. Chicago: University of Chicago Press. Lake, David and Donald Rothchild. 1996. International Security 21, no. 2: 41 75. Lemarchand, Rene.1970. Rwanda and Burundi Pall Mall Press: London. -------------------------.1996 Burundi: Ethnocide as discourse and Practice New York: Cambridge. -------------------------.2006. African Affairs 106, no. 422: 1 20. ------------------------. 2009. The Dynamics of Conflict in Central Africa. Philadelphia: University of Pennsylvania Press. Licklider, Roy.1995 The consequences of negotiated settlements in civil wars 1945 American Political Science Review 89, no. 3:681 690. Lindberg, Staffan. 2006. Democracy and Elections in Africa. Baltimore: Johns Hopkins University Press. L i j phart, Arend. 1969. World Politics 21. Longman, Timothy. 1997. l Huxtable,eds. The African State at a Critical Juncture: Between disintegration and Reconfiguration Boulder: Lynne Rienners. ------------------------.1998. Proxy Targets: Civilians and the War in Burundi. New York: Human Rights Watch. -----------------------. 2001. Journal of Religion in Africa 31, no. 2: 163 186. ------------------------.2012. The New York Times June 29, 2012: A21. Luttwak, Edward. 199 Foreign Affairs 78, no. 4: 36 44.
381 Macquet, Jacques. 1961. The Premise of Inequality in Rwanda: A study of Political Relations in a Central African Kingdom. Oxford: Oxford University Press for the International African Institute. Mampilly, Zacariah. 2011. Rebel Rulers: Insurgent Governance and Civilian Life during War. Ithaca : C ornell University Press. Mann, Michael. 1984. The Autonomous Power of the State: Its Origins, Mechanisms and Results. European Sociological Archives. -------------------.1986. The Sources of Social Power: Volume 1, A History of Power from the Beginning to AD 1760 London: Cambridge University Press. --------------------.1993. The Sources of Social Power: Volume 2, The Rise of Classes and Nation S tates 1760 1914 London: Cambridge University Press. ---------------------.1999. New Left Review 235. ---------------------.2005. The Dark Side of Democracy: Explaining Ethnic Cleansing. Cambridge: Cambridge University Press. Mansfield Edward and Jack Snyder. 2005. Electing to Fight: Why Emerging Democracies go to War. Cambridge, MA: MIT Press. -------------------------------------2007. Journal of Democracy 18, no. 3: 5 10. ------------------------------------.2009. Pathways to War in Democratic Transitions. International Organization 63:381 90. Mason T. David et. al. 2011. International Studies Perspectives 12, no. 2: 171 189. Marysse, Stefaan and Filip Reyntjens. 2005. The Political Economy of the Great Lakes Region in Africa: The Pitfalls of Enforced Democracy and Globalization. New York: Palgrave Macmillan. Malkki, Liisa.1995. Purity and Exile; violence, memory and National Cosmology among refugees in Tanzania. Chicago: University of Chicago Press. Metelis, Claire. 2009. Inside Insurgency: Violence, Civilians and Revolutionary Group Behavior. New York: NYU Press.
382 Migdal, Joel. 1988. Strong Societies and Weak States: State Society Relations and Capabilities in the Third World. Princeton: Princeton University Press. Miguel, Edward et al. Journal of Political Economy 112, no. 4: 725 753. Moore, Barrington. 1966. Social Origins of Dictatorship and Democracy: Lord and Peasant in the Makin g of the Modern World. Boston: Beacon Press. Museveni, Yoweri. 1997. Sowing the Mustard Seed : The Struggle for Freedom and Democracy in Uganda. London: Macmillan Publishers. Mwenda, Andrew. 2007. Journal of Democracy 18, no. 3: 23 37. Ndikumana, Leonce. 1998. African Studies Review 41, no. 1: 29 47. June 2010, Reuters Available o nline < http://www.alertnet.org/thenews/newsdesk/LDE6501QL.htm > (accessed 21 June 2011). Newbury,Catherine. 1990 The Cohesion of Oppression: Clientship and Ethnicity in Rwanda, 1860 1960. New York: Columbia University Press ------------------------Africa Today 45, no. 1: 7 24. Newbury, David. 2001. International Journal of African Historical Studies 34, no. 2: 255 314. Conflict in Journ al of African Economies 9, no. 3: 370 409. ed African Guerrillas Oxford: James Currey: 91 106 Nindor era, Willy. 2008. CNDD Zeeuw, Jeroen, ed. From Soldiers to Politicians: Transforming Rebel Movements after Civil War. Boulder: Lynne Reiner. ---------------------. 2010. The CNDD FDD in Burundi: The path from armed to political struggle. Berghof Transitions Series No.10. Berlin:Berghof Foundation.
383 Nonyzima Hermenegilde. 2004. Terres des Heros Non Chantes: du crime et de Bujumbura: Remesha. Nzongola Njalala, Georges. 2002. The Congo: From Leopold to Kabila: History of Congo. Lond on: Zed Books. Olson, Mancur. 1993. American Political Science Review 87, no. 3: 567 576. Ottoway, Marina. 1999. Joseph, ed. The State, Conflict, and Democracy in Africa Boulder: Lynne Reiner Palacios, Marco. 2006. Between Legitimacy and Violence Durham: Duke University Press. Pool, David 2001. Front. Oxford: James Currey. Posner, Daniel. 2005. Institutions and Ethnic Politics New York : Cambridge University Press. Prunier. Gerard. 1995. The Rwanda Crisis: History of a Genocide New York: Columbia University Press ------------------.1998. African Guerillas. London: James Currey. -------------------.2005. Darfur: The Ambiguous Genocide. Ithaca: Cornell University. --------------------.2009. Africa's World War: Congo, the Rwandan Genocide, and the Making of a Continental Catastrophe London: Oxford University Press. Putnam, Robert. 1993. Making Democracy Work: Civic Traditions in Modern Italy Princeton: Princeton University Press. Quinn, J. Michael, T. Davi International Interactions 33: 167 193. ACLED Armed Conflict Location and Journal of Peace Research 47, no 5: 1 10. Reed William Cyrus.1995. Issue: A Journal of Opinion 23, no. 2: 159 185.
384 Reno, William. 1999. Warlord Politics and African States. Boulder: Lynne Rieners Publishing. -----------------.2000. Greed and Grievance: Economic Agendas in Civil Wars, ed. Mats Berdal and David M. Malone. Bould er: Lynne Rienner: 43 68. ----------------Development and Change 33, no. 5: 837 858. ----------------. 2011. Warfare in Independent Africa. London: Cambridge University Press. Reyntjens, Filip.1993. Journal of Modern African Studies 31, no. 4: 563 583. -------------------. 2001. African Affairs 100, no. 399: 311 317. -------------------African Affairs 103: 177 210. -------------------. 2008. Africa Today 54, no. 3: 141 144. -------------------. 2009. The Great African War: Congo and Regional Geopolitics 1996 2006. New York: Cambridge University Press. -------------------. 2011. World: Governance in Post Ge African Affairs 110, no. 438:1 34. Richards, Paul ed 2005. No Peace, No War: An Anthropology of Contemporary Armed Conflicts Oxford: James Currey. Roeder, Phillip and D onald Rothchild. 2005. Sustainable Peace: Power and Democracy after Civil Wars. Ithaca: Cornell University Press. Rotberg, Richard. 2003. When States Fail: Causes and Consequences. New Haven: Yale University Press. Samii, Cyrus. 2012 se in Burundi." American Political Science Review forthcoming. ----------------.2010. Who Wants to Forgive and Forget?: Civilian Attitudes Toward Post conflict Justice and Truth in Burundi. Working Paper.
385 al Rebels: Neighboring States as Sanctuary for Rebel World Politics. 59(2): 217 242. -------------------. 2010. Rebels without Borders: Transnational Insurgencies in World Politics Ithaca: Cornell University Press. Scherrer, Christian. 2002. Genocide and Crisis in Central Africa: Conflict Roots, Mass Violence, and Regional War. Westport: Praeger. Schultz, Kenneth .2010. Conflict over Rebel Support in Civi International Organization 64: 281 312. Scott, James C. 1977. The Moral Economy of the Peasant : Rebellion and Subsistence in Southeast Asia. New Haven: Yale University. Seay, Laura. 2012. Foreign Policy Snyder, Jack. 2000. From Voting to Violence: Democratization and Nationalist Conflict New York: Norton. Sobek, David. 2009. The Causes of War. London: Cambridge University Press. Sriram, Chandra and Karin Wermester. 2003. From Promise to Practice: Stre ngthening UN Capacities for the Prevention of Violent Conflict. New York: International Peace Academy. Miller McCune. 2: No. 6. Staniland, Paul. 2012. Perspectives on Politics 10, no. 2: 243 264. StataCorp. 2011. Stata Statistical Software: Release 12 College Station, TX: StataCorp LP. Stiansen, Endre. 2004. Journal of African History 45, no. 2: 350 351. Storey Review of African Political Economy 89: 365 385. Straus, Scott. 2006. The Order of Genocide: Race, Power and War in Rwanda Ithaca: Cornell University Press.
386 Stearns, Jason K. 2011. Dancing in the Glory of Monsters. The Collapse of the Congo and the Great War of Africa London: Perseus Books. Sull ivan Daniel. 200 5 Journal of Modern African Studies 43, no. 1: 75 95. Thie s, Cameron. 2009. World Politics 61, no. 4: 623 669. Thomson, Susan. 2009. Resisting Reconciliation: State Power and Everyday Life in Post Genocide Rwanda PhD research Dalhousie University. Tilly, Charles. Charles, ed. The Formation of National States in Western Europe. Princeton: Princeton University Press. ----------------.1978. From Mobilization to Revolution New York: McGraw Hill. Tirone David and B Savun. 2011. Foreign aid, democratization, and civil conflict: how American Journal of Political Science 55 no. 2: 233 246. Toft, Monica Duffy. 2010 a International Security. 34, no 3: 7 36 Toft, Monica Duffy. 2010b. Securing the Peace: The Durable Settlement of Civil Wars Princeton: Princeton University Press. Sharing: African Affairs 104: 375 398. Turner, Thomas. 2007. The Congo Wars: Conflict, Myth and Reality. London: Zed Books. Turner Thomas and Crawford Young. 1985. The Rise and Decline of the Zairian State. Madison: University of Wisconsin. Accessed 4 November 2011. Available online at < https://www.cia.gov/library/publications/the world factbook/geos/ug.html >. Ulfelder, Jay. Foreign Policy
387 United Nations High Committee on Refugees ( UNHCR ) 2004. Child Soldiers Global Report 2004. New York: United Nations. University of Gothenburg. 2009. The Quality of Government Cross Section Data version 17 June 2009. Gothenburg: University of Gothenburg. Uppsala Conflict Data Program, UCDP Conflict Encyclopedia. Accessed 2011/06/21. Available online< www.ucdp.uu.se/database >. UCDP dataset One Sided Violence v. 1.4 2012. 1989 2011. Uppsala: UCDP/PRIO. Uvi n, Peter. 1998. Aiding Violence: The Development Enterprise in Rwanda West Hartford: Kumarian Press. ---------------Census and Identity: The Politics of Race, Ethnicity and Language in National Censuses Kertzer, David and Dominique Arel, eds. Cambridge: New York: 148 175. -----------------. 2009. Van de Walle, Nicolas. 2001. African Economies and the Politics of Permanent Crisis: 1979 1999. Cambridge: Cambridge University Press. Walter, Barbara. 1997. International Organization 51, no. 3: 335 364. Walter, Barbara. 2002. Committing to Peace: The Successful Settlement of Civil Wars. Princeton: Princeton University Press. Recurring Civil Journal of Conflict Resolution 41, No. 3: 371 388. Watt, Nigel. 2008. Burundi: The Biography of a Small African Country. New York: Columbia University Press. Journal of Conflict Resolution 49: 589 624 ----------------------. 2007. Inside Rebellion: The Politics of Insurgent Violence. New York: Cambridge. Weinstein, Warren. 1976. Journal of Modern African Studies 14, no. 1: 161 163.
388 Wickham Crowley, Timothy. 1987. Guerrillas and Revolution in Latin America: A Comparative Study of Insurgents and Regimes since 1956. Princeton: Princeton University Press. Wilkinson, Steven. 2004. Votes and Violence : Electoral Competition and Ethnic Riots in India New York: Cambridge University Press. Wood, Elizabeth. 2003. Global Governance 9, no. 2: 247 260. Journal of Peace Research 47, no 5: 60 1 614. World Bank. 2007. World Bank Development Indicators through year 2006 Washington, D.C.: World Bank. World Bank. 2009. The Worldwide Governance Indicators (WGI) project 1996 2008 Washington, D.C.: World Bank. Young, Crawford. 1973. Patrimonialism and Political Change in the Congo By Jean Journal of Politics 35, no. 2: 506 508. -------------------.1994. The African State in Colonial Perspective New Haven: Yale University Press. Zartman, William. 1995. Collapsed States : The Disintegration and Restoration of Legitimate Authorit y Boulder: Lynne Reiner
389 BIOGRAPHICAL SKETCH Cara Eugenia Jones grew up on the Gulf Coast. She studied at the University of Texas at Dallas, earning a Bachelor of Arts degree in government and politics During graduate school, she studied abroad and worked at MS TCDC in Arusha, Tanzania, and Bujumbura, Burundi. She ear ned a Master of Arts degree in political s cience and a graduate certificate in African Studies from the University of Florida in August 2010. She then returned to Burundi for a year of field study under a Fulbright Hays Doctoral Dissertation Research Abroad fellowship from May 2010 to July 2011. Cara has worked on political asylum cases in the United States for Great Lakes refugees and served as ves and works in Grinnell, Iowa, as an instructor of political science at Grinnell College.