33592 Paper No. 27 / June 2005 Conflict-Sensitive Development Assistance: The Case of Burundi Juana Brachet Howard Wolpe Summary Findings Since its independence in 1962, Burundi has been mired in an unending cycle of conflict. Successive Four years into the transition that began in November waves of violence have increased ethnic and regional 2001 some factors threaten to escalate Burundi's divisions, while deepening already extreme poverty. conflict, while others are working to dampen the Although Burundians traditionally did not mobilize conflict or have an uncertain impact. The large politically around their Hutu and Tutsi ethnic number of "on the brink" factors suggests that there identities, the chronic post-independence violence are still many opportunities for development inserted a mass dimension to what is fundamentally assistance, and traditional and Track II diplomacy to an elite-driven and manipulated conflict. Repeated assist in the consolidation of Burundi's peace inter-communal massacres have rendered the process. population susceptible to ethnically framed political appeals. The paper notes eight principles to guide development assistance: (i) "do no harm" particularly The still-evolving peace process begun in Arusha, to avoid reinforcing or triggering conflict causes; (ii) Tanzania in 1996 offers an opportunity for Burundi make peace dividends visible to the population; (iii) to break out of the zero-sum game that has include short-term issues, especially the restoration of characterized its political and social life since shortly security; (iv) limit the potential for mass after independence. Although active conflict has mobilization; (v) address the structural causes of diminished, peace and reconciliation remains conflict; (vi) address the perceptual and attitudinal fragile--as tragically demonstrated by the August legacy of the conflict; (vii) ensure that development 2004 massacre of Congolese Banyamulenge in a assistance is consistent and sustained; and (viii) refugee camp in Burundi and sporadic eruptions of consider the regional context. violence in Bujumbura Rurale province. In this context, the paper explores ways in which The paper reviews a number of development areas development assistance can contribute to the where there are opportunities to incorporate the consolidation of peace, framed within a above principles, including: the PRSP process; rural contextualized assessment of Burundi's conflict using development; infrastructure; security sector reform the Bank's Conflict Analysis Framework. and demobilization and reintegration of ex- combatants; land tenure; employment generation; The conflict analysis finds ethnicity is only one of governance; and the social sectors. several cleavages in Burundian society. Although the ethnic divide is perceived to be more prominent In terms of project design and implementation, the because of four decades of manipulation of socio- paper emphasizes the need to: engage the elites on ethnic identities, there are also important clan, the side of peace; invest in bottom-up approaches, regional and class-based divides. These divides have especially through inclusive and community-driven been exacerbated by: differential social opportunities; interventions; include peace-building components in a history of violence and impunity; poor economic projects; carefully monitor and assess development performance, inequality and environmental stress; interventions in terms of explicit peace-building failed governance and institutions; and the spillover objectives and indicators; map and consider the effects of regional conflict among Burundi's multiplicity of variables that affect the peace process; neighbors. and complement regional stabilization efforts. SOCIAL DEVELOPMENT PAPERS Conflict Prevention & Reconstruction Paper No. 27 /June 2005 Conflict-Sensitive Development Assistance: The Case of Burundi Juana Brachet Howard Wolpe This Working Papers Series disseminates the findings of work in progress to encourage discussion and exchange of ideas on conflict and development issues. Papers in this series are not formal publications of the World Bank. This paper has not undergone the review accorded to official World Bank publications. The findings, interpretations and conclusions herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or its Executive Directors, or the governments they represent. The papers carry the names of the authors and should be cited accordingly. The series is edited by the Conflict Prevention and Reconstruction Unit in the Social Development Department of the Environmentally and Socially Sustainable Development Network for the World Bank. To request copies of the paper or for more information on the series, please contact the CPR Unit. Papers are also available on the CPR Unit's website. For additional copies of this paper, please contact: Conflict Prevention & Reconstruction Social Development Department The World Bank 1818 H Street, NW Washington, DC 20433 Fax: 202-522-3247 Web : http://www.worldbank.org/conflict (see "Publications" in the navigation menu) E-mail: cpr@worldbank.org Printed on Recycled Paper Table of Contents Acronyms ..................................................................................................................................................... ii Acknowledgements ..................................................................................................................................... iii INTRODUCTION AND BACKGROUND ................................................................................................. 1 I. CONFLICT ANALYSIS: SCARCITY, COMPETITION, AND THE MANIPULATION OF ETHNICITY (1965-PRESENT)............................................................................................................ 4 Main Causes of Conflict: Scarcity, Competition and the Manipulation of Ethnicity ............................ 5 Economic Structure and Performance: Competition in a Context of Extreme Poverty and Concentration of Power ......................................................................................................................... 8 External Forces: A Challenging Neighborhood, with Cross Border Feedback Effects........................ 9 II. CONFLICT DYNAMICS: STRUCTURAL CAUSES, TRIGGERS, AND FEEDBACK EFFECTS...........................................................................................................10 III. CONFLICT DYNAMICS TODAY: ON THE BRINK .................................................................... 12 Potentially De-Escalating Factors........................................................................................................ 12 Potentially Escalating Factors.............................................................................................................. 13 Several Factors Are "On the Brink" with Currently Uncertain Impact ............................................... 14 IV. AREAS IN WHICH DEVELOPMENT ASSISTANCE CAN SUPPORT THE TRANSITION AND OPERATIONAL ISSUES TO CONSIDER .............................................................................. 15 Development Assistance and Diplomacy in War-to-Peace Transitions............................................... 15 Considerations for the Development Agenda in Burundi.................................................................... 17 V. OPERATIONAL ISSUES TO CONSIDER IN PROJECT DESIGN, IMPLEMENTATION AND MONITORING AND EVALUATION...................................................................................... 21 Annex 1: Chronology ................................................................................................................................ 24 Bibliography............................................................................................................................................... 27 Boxes Box 1: Four Phases in the Peace Process, 1993-2004..............................................................2 Box 2: Causal Relationships, 1996-2004............................................................................12 Box 3: Government Priorities and Donor Support, 2004-2005...................................................20 Tables Table 1: Selected Economic and Social Indicators, 1990-2002....................................................9 Table 2: Conflicts After Independence..............................................................................10 Table 3: Causes that Development and Diplomacy/Track II Interventions Can Address.....................16 Table 4: Conflict Variables and Some Key Program and Project-Level Implications........................23 ii Acronyms AMIB African Mission in Burundi BLTP Burundi Leadership Training Program CNDD-FDD National Council for the Defence of Democracy/Forces for the Defence of Democracy DDR Disarmament, Demobilization and Reintegration DRC Democratic Republic of Congo DRRP Demobilization, Reinsertion and Reintegration Program EMGI Etat-major general intégré FAB Armed Forces of Burundi FNL Forces for National Liberation GDP Gross Domestic Product GNP Gross National Product HIV/AIDS Human Immunodeficiency Virus/Acquired Immune Deficiency Syndrome IMF International Monetary Fund I-PRSP Interim Poverty Reduction Strategy Paper MDRP Multi-Country Demobilization and Reintegration Program ONUB Opérations des Nations Unies au Burundi PRSP Poverty Reduction Strategy Paper UN United Nations UNDP United Nations Development Programme UNICEF United Nations Children's Fund UPRONA Parti de l'Unité pour le Progrès National iii Acknowledgements This paper has been prepared by Juana Brachet and Howard Wolpe. Juana Brachet, currently Senior Technical Advisor at the National Commission for Disarmament, Demobilization and Reintegration of the Government of the Democratic Republic of Congo, was previously a Conflict Specialist with the Conflict Prevention and Reconstruction Unit of the World Bank and the DRC Country Team. Howard Wolpe, formerly the U.S. Special Envoy to Africa's Great Lakes Region, was the originator of the Burundi Leadership Training Program, a capacity-building initiative launched with the support of the World Bank's Post-Conflict Fund. Currently he directs the Africa Program at the Woodrow Wilson International Center for Scholars. This paper draws on four initiatives. The first is a collaborative effort between the World Bank and the Permanent Secretariat for Economic and Social Reforms of the Government of Burundi on the Poverty Reduction Strategy Paper (PRSP), identifying ways the consultation process and the PRSP itself can address structural causes of the conflict and contribute to the consolidation of peace (2004). The second is the Burundi Leadership Training Program, a post-conflict reconstruction initiative launched by the Woodrow Wilson International Center for Scholars in October 2002 with the support of the World Bank's Post Conflict Fund. The third initiative is the PRSPs in Conflict-Affected Countries Project of the World Bank (Poverty Reduction Group and Conflict Prevention and Reconstruction Unit), which includes analysis of a set of conflicts and a review of the conflict-sensitivity of PRSPs. The fourth initiative is the conflict analysis exercise conducted in 2003 by Libère Ndabakwaje, Jean-Baptiste Mbonyingingo, and the International Fund for Agricultural Development. The authors would like to thank those who were involved with these four endeavors, as well as their many colleagues who reacted to earlier drafts of this paper. We are particularly indebted to Ian Bannon, René Lemarchand, Ana Paula Lopes, Steve McDonald, Eugene Nindorera, Katrina Sharkey, Alassane Sow, Maude Svensson, Achille Toto Same, Peter Uvin, Per Wam, Ingo Wiederhofer and Peter Woodrow. Juana Brachet Howard Wolpe CONFLICT-SENSITIVE DEVELOPMENT ASSISTANCE: THE CASE OF BURUNDI INTRODUCTION AND BACKGROUND "It must be possible for the people of Burundi to materially distinguish between the destructiveness of conflict and the benefits of peace" Nelson Mandela, Burundi peace talks mediator Since its independence in 1962, Burundi has been stuck in an unending cycle of conflict. Successive waves of violence--particularly in 1965-69, 1972, 1988, 1991 and 1993--have increased ethnic and regional divisions while deepening already extreme poverty. In the 43 years since Belgian rule ended, Burundi has witnessed an estimated 300-400,000 killed, and since 1993, about 1.3 million (16% of the population) have become internally displaced and refugees. GDP per capita had fallen to $110 in 2003, one of the lowest in Africa, and the proportion of people living below the poverty line had risen from 35% in 1992 to more than 60% in 2002. The chronic violence has had a direct impact on the agricultural sector, which accounts for 50% of Burundi's national product. The economic impact has been compounded by the fall of coffee prices and land erosion. This has made competition for control of the state an exceptionally high-stakes contest. While Burundians traditionally did not mobilize politically around their Hutu and Tutsi ethnic identities, the chronic post-independence violence gave a mass dimension to what is fundamentally an elite-driven and manipulated conflict. Repeated inter-communal massacres have rendered the population susceptible to ethnically framed political appeals. The international community's record in responding to Burundi's chronic violence is mixed. The post-independence assassinations, repeated episodes of inter-communal violence, and even the 1972 massacres were virtually ignored by the international community--a pattern which did not go unnoticed by either Burundi's victims or perpetrators. On the diplomatic front, since the assassination in 1993 of Burundi's first democratically-elected Hutu president, the Burundi peace process has passed through three phases, each with its own weaknesses and strengths (Box 1).1 On the security front, UN Secretary General Boutros Boutros-Ghali's attempt at military contingency planning in 1994 failed because of a lack of support from major powers, and because the Tutsi-dominated government resisted a peacekeeping force that it felt would threaten its control of the Burundian army, seen by Tutsis as the key instrument of Tutsi protection and control. In later years, however, this resistance was overcome, and in 2003 the African Mission in Burundi (AMIB), an African Union- sponsored peacekeeping force, was put in place to monitor the cease-fire agreements. In May 2004, UN Security Council Resolution 1545 transformed AMIB into a UN peacekeeping operation, ONUB. With the recent integration of the CNDD-FDD (Nkurunziza) into both the army and the transitional political institutions, security sector reform--which involves the reform and downsizing of the army, disbanding the gendarmerie and the gardiens de la paix, and strengthening of the national police, and is recognized by all parties as one of the most sensitive and critical keys to a stable Burundian future--is finally on the table, and the process of demobilizing, reinserting and reintegrating ex-combatants has begun. 1See Wolpe (2003). 2 Box 1: Four Phases in the Peace Process, 1993-2004 Between 1994 and 1996, the UN attempted to facilitate the negotiation of new power-sharing arrangements to stabilize Burundi after the 1993 wave of violence that followed the assassination of President Ndadaye and the 1994 genocide in neighboring Rwanda. This attempt was handicapped by the absence of a power willing to take the lead on either the diplomatic or, with the UN Secretary General's call for a contingent peacekeeping force unanswered, on the security front. Between early 1996 and late 1999, former Tanzanian President Nyerere facilitated a regionally sponsored peace process, with the support of the wider international community. The peace process unfolded in three different but increasingly linked arenas, Arusha, Rome, and Bujumbura. Nyerere's facilitation was weakened by the limited trust the facilitator put in the negotiating parties, and vice-versa, and by the multiplicity of the regional sponsors of the Arusha process. In particular, Tutsi elements perceived the facilitator as biased in favor of Hutus; moreover, by acting not only as facilitator but also as enforcer of the regionally imposed sanctions regime, the neutrality and credibility of the Arusha process was called into question. In addition, there was competition between what the main Tutsi interlocutor, President Buyoya, viewed as a domestic peace process that he would guide, and Nyerere's regionally sponsored process. Peace negotiations were further complicated by strains between Nyerere's facilitation team and the Rome-based Community of Sant' Egidio that hosted the first secret talks between the government and the CNDD rebel movement. Nonetheless, Nyerere and his team succeeded in moving the Burundians toward agreement on some of the protocols that would eventually find their way into the final Arusha Accord. Between the end of 1999 and August 2000, following Nyerere's death, former South African President Mandela assumed the lead role in the process that culminated in the Arusha Agreement signed by 19 negotiating parties. The Agreement provided a blueprint for power sharing and the reform of key institutions, and prescribed a 36-month transitional period commencing November 1, 2001. However, the Arusha agreement was flawed in two key respects: it lacked the support of the two principal armed groups, and it was silent on the key issues of who would lead the transition and how security sector reform would be accomplished. Moreover, several signatories recorded formal reservations. In a post-Arusha process that at times resembled an arbitration, Mandela and the regional leaders persuaded the Burundians to divide the transition into two equal parts, with President Buyoya to lead the Government for the first 18 months and with his Hutu vice-president, Domitien Ndayizeye, to take over the presidency for the second half of the transition. Since 2000, one of the two main rebel groups absent from the Arusha talks, the CNDD-FDD (Nkurunziza), has agreed to a cease fire (December 2002) and to participate in the transitional government (November 2003), while the FNL-Palipehutu (Rwasa) until recently remained intransigent and, in Bujumbura Rurale, where the FNL operates, continued military operations. In April 2003, the prescribed hand-off from Buyoya to Ndayizeye occurred without incident. However, continued disputation over the power-sharing arrangements following the transition forced a delay in the holding of Burundi's first post-Arusha elections. On February 28, 2005, the public referendum to adopt the new Constitution was finally held, with over 90% of Burundi's 3 million registered voters casting "yes" votes. Under a new electoral timetable, communal elections were held on June 3, 2005; legislative elections will be held on July 4, 2005; senatorial elections will take place on July 29, and the president will be selected by the National Assembly and the Senate on August 19; local commune elections will follow shortly thereafter. In April 2005, the FNL declared its intention to stop fighting and its readiness to enter unconditional negotiations with the transitional government. 3 On the humanitarian and development front,2 international engagement has been marked by an on- and-off approach. Between 1992 and 2002, international financial assistance fell from an annual average of nearly $300 million to an annual average of less than $100 million, largely as a result of the resumption of violent conflict in 1993. Donors pledged substantial amounts of development assistance after the conclusion of the Arusha Accord in 2000. However, a significant proportion of this assistance has not been disbursed, in part because of persistent (albeit diminishing) insecurity and political uncertainties. In an effort to galvanize a more robust international effort, the Government of Belgium and UNDP hosted a Partners Forum for Burundi in January 2004, where donors pledged $1.032 billion in support of Burundi and development assistance has resumed. The Arusha peace process, for all of its imperfections, yielded a window of opportunity for Burundi to step out of the zero-sum game that has characterized its political life since shortly after independence. A transitional constitution was adopted in October 2001; a 36-month-transitional government was launched in November 2001, based on the principle of power sharing between the country's two main ethnic groups; and a transitional parliament was installed in January 2002. President Buyoya, a Tutsi, led the first 18 months of the transitional government. In May 2003, the transitional Hutu vice-president, Domitien Ndayizeye, ascended to the presidency without incident, while Alphonse Kadege, a Tutsi, assumed the vice-presidency. Following a series of difficult and extended negotiations, the transitional government signed new peace and cease-fire agreements with all but one of the armed groups that had not signed the Arusha agreement. In November 2003, the CNDD-FDD (Nkurunziza), the most significant of the rebel forces, agreed to participate in the transitional government. The new post-Arusha peace agreements resulted in an expanded Transitional Government inclusive of the new signatories. Fighting has halted in all but one province, Bujumbura Rurale, where an intransigent FNL-Palipehutu (Rwasa) continues to launch attacks and, until very recently, to take a hard line in opposition to negotiations and the Transitional Government. When it became clear that the originally mandated 36-month transitional timetable would not be met, and the referendum on the new constitution would be delayed, it was agreed that the draft institution would become the "interim constitution" pending its ratification. Subsequently, when the President--under new authority granted in the interim constitution--dismissed the Tutsi vice-president, with whom he had a difficult personal relationship, and appointed another Tutsi UPRONA party leader in his place, his decision was accepted with barely a murmur. While Burundians appear generally to be tired of war and committed to the negotiated resolution of outstanding issues, the August 2004 massacre of Congolese Banyamulenge in a refugee camp in Burundi, in which the FNL participated, was a tragic reminder of the fragility of the peace process and, in particular, of the impact on Burundi of the continuing volatility of the eastern Democratic Republic of Congo (DRC)­Burundi border region. In addition, while there is evidence of considerable effort being made by both the Burundian Armed Forces (FAB) and rebel military commanders to build a newly independent and cohesive national army and police force, the process of security sector reform carries with it potentially unsettling risks and uncertainties. In addition, continued disputes over the new constitution suggest that some Tutsi elements are not wholly reconciled to the new power-sharing arrangements. 2 In this paper, the authors do not separate humanitarian and development assistance. Although humanitarian aid is traditionally life-saving and needs-based, and development assistance has longer-term objectives, from a programmatic point of view the distinction is increasingly artificial: (i) post-Cold War "complex emergencies," such as conflicts characterized by chronic waves of violence and instability rather than sharply differentiated pre-conflict, conflict and post-conflict phases, make the short-term vs. long-term distinction less relevant; (ii) humanitarian actors have almost all started to integrate longer-term objectives in their programming (e.g., the International Committee of the Red Cross distributing seeds and tools in Afghanistan for over 5 years); and (iii) in post-crisis contexts, what is usually seen as development assistance has a direct humanitarian impact (e.g., the rebuilding of roads, which directly improves food security). 4 In this context of a diminution of active conflict but continuing political uncertainty, how can development assistance contribute to the consolidation of peace? Although peace opens opportunities for development, development only leads to peace under certain conditions. Well-conceived development assistance can address some of the underlying causes of the conflict. If appropriate macroeconomic policies are in place, development-assisted improvement in the material conditions of communities can reduce economically inspired conflicts, for example, over land. Moreover, bricks-and-mortar projects can also be a vehicle to address social issues, by strengthening inter-group trust and the capacity of communities to collaborate effectively and manage conflict without resort to violence. Such peace building development assistance, if it is to be effective, requires, first, the identification of the main causes and dynamics of the conflict to help inform donor engagement. For instance, beyond "doing no harm" by not reinforcing the structural inequalities that characterize Burundi, development assistance that would contribute to the war-to-peace transition requires an emphasis on inclusion, elite accountability, the decentralization of economic and political power, and support for the creation of economic opportunity outside the sphere of the state. Second, those specific areas in which development can and should support the transition should be identified. For instance, the PRSP process and large-scale macro and sector programs may provide opportunities to address some of the structural causes of the conflict. In addition, well-conceived external assistance should aid in the building of cohesion and managerial effectiveness of national elites and, at the community level, address the attitudinal legacy of four decades of conflict, and decrease the potential for ethnic mobilization and a return to violence. In this connection, development projects need to incorporate explicit and measurable peace-building objectives, based on the analysis of the conflict, and to define specific ways to reach these objectives. This paper, which draws on four initiatives,3 uses the Conflict Analysis Framework developed by the Conflict Prevention and Reconstruction Unit of the World Bank4 to identify the conflict dynamics in Burundi from its independence in 1962 to the present, with emphasis on the post-Arusha period, 2000- 2004. To analyze the Burundian conflict in a dynamic fashion, as well as to identify appropriate responses, this paper also makes a distinction between triggering causes and structural causes. Triggering causes are more proximate events or factors that cause a conflict to escalate. Structural causes refer to the underlying fundamental causes of a conflict. This paper first identifies the causes and dynamics of the conflict (Part I), then highlights critical areas in which development assistance can support the transition followed by some project-level recommendations. I. CONFLICT ANALYSIS: SCARCITY, COMPETITION, AND THE MANIPULATION OF ETHNICITY (1965-PRESENT) The Arusha Agreement recognizes that the conflict "...is fundamentally political, with extremely important ethnic dimensions; it stems from a struggle by the political class to accede to and/or remain in power" (Preamble, Protocol I). Using the Conflict Analysis Framework, this section reviews the most salient variables involved in understanding the Burundian conflict, with particular attention on those that 3These four initiatives are: (i) ongoing collaboration with Secretariat Permanent/REFES on the Poverty Reduction Strategy Paper; (ii) the Burundi Leadership Training Program, implemented by the Woodrow Wilson Center between 2003 and 2005; (iii) the World Bank's PRSPs in Conflict-Affected Countries Project; and (iv) the conflict analysis exercise conducted in 2003 by Libère Ndabakwaje, Jean-Baptiste Mbonyingingo, and the International Fund for Agricultural Development (IFAD), using the Bank's Conflict Analysis Framework (CAF). 4See www.worldbank.org/conflict. 5 are responsible for the escalation or de-escalation of tensions, and to the interplay of these variables. The implications for development assistance are based on the conclusions of this section. Main Causes of Conflict: Scarcity, Competition and the Manipulation of Ethnicity Social, ethnic and economic cleavages and regional imbalances: ethnicity is only one of several cleavages in Burundian society. Social divides in Burundi are both vertical and horizontal, and are of multiple kinds. Four divides predominate: ethnic, clan-based, regional, and class-based (urban elites vs. rural masses). Although the ethnic divide is perceived to be prominent because of four decades of manipulation of socio-ethnic identities in the post-independence struggle for power, these four divisions have been consistent throughout independent Burundi's history and remain in place to this day. The four divisions often overlap. Although historically ethnic divisions were tempered by clan- and region-based loyalties, the distinction between Burundian "haves" and "have nots" often coincides with the ethnic (Tutsi vs. Hutu and Twa) and regional (e.g., Bururi vs. non-Bururi) distinctions. Although Burundi's ethnic composition is similar to that of Rwanda (85% Hutu, 14% Tutsi, 1% Twa), and the two countries' fates seem to be inextricably linked, Burundi is distinctly different from Rwanda in that, prior to Burundian independence in 1962, the traditional Tutsi-Hutu distinction was not the basis of political mobilization. In Burundi, Hutu and Tutsi were united by their common allegiance to, and identification with, the monarchy. Members of the ganwa aristocracy served as the provincial governors, ruling the various areas of Burundi in the name of the King. Political competition occurred between the princely clans that collectively constituted the ganwa, and the contenders for monarchical power drew upon both Tutsis and Hutus. Partly as a consequence of this historical pattern, in the colonial period the Belgians in Burundi relied far more on solidarities and divides within the clan system ("abaganwaization") than on the Tutsi/Hutu distinction. In Rwanda, by contrast, the Belgians focused on the Tutsi/Hutu divide, elevated the standing and power of the Tutsi, and effectively ruled through them. While the Hutu-Tutsi divide was, to some extent, hardened under colonial rule, it was tempered by a complex system of crosscutting social cleavages. First, Burundi society also includes clan-based divides: vertical groupings or castes (for example, the two distinct Tutsi castes: Hima--lower, and Banyaruguru-- higher), and horizontal divides within castes among lineages. Second, these castes and lineages were grounded in territories, with the third most salient line of political mobilization after ethnicity and clan being regional and economic differences (e.g., the dominance of Bururi province since the 1960s, which contributed three presidents and much of the high military command).5 In addition, today there are deep divisions within each ethnic group between "moderates" and "hard liners," and divides between Burundian elites and the mass of the population (educated vs. uneducated, rural vs. urban, urban employed vs. urban unemployed). Further complicating the contemporary Burundian social mosaic, women face legal and societal discrimination, including discriminatory inheritance laws and credit practices, and domestic violence is thought to be widespread. In rural areas, there are tensions between returning internally displaced persons and refugees and their communities of origin, stemming from conflicts over land (compounded by speculation and irregularities in the allocation of land). Significantly too, members of Burundi's Twa minority have always been marginalized economically, socially, and politically, with many living in isolation, without access to social services. While not to the same extent as in Rwanda, Belgian colonialism in Burundi did elevate to some degree the significance of the Tutsi/Hutu ethnic distinction. Beginning in 1965 with the assassination of Prime 5 For social indicators by province, see Interim Strategic Framework for Accelerating Economic Growth and Reducing Poverty (interim PRSP), November 2003, Republic of Burundi, and Burundi Poverty Note, Prospects for Social Protection in a Crisis Economy, For example, Karuzi province had three times the percentage of households under the poverty line as did Cankuzo province in 1999. 6 Minister Ngendandumwe by a Rwandese Tutsi refugee in Burundi, the Hutu-Tutsi divide became the dominant lens through which Burundian political conflict came to be viewed. Subsequent events in neighboring Rwanda further crystallized the Burundian conflict around ethnicity. Growing Burundian Hutu resentment of post-independence Tutsi dominance was fueled by the situation in Rwanda, while Burundi's Tutsi minority began to fear the extension of the Rwandan "Hutu power" revolution to Burundi. The 1965 assassination, the mass killings that followed, subsequent assassinations and inter- communal massacres (1965-69, 1972, 1988, 1991 and 1993) greatly intensified fears on both sides, and prompted a spiral of violence that is only now starting to recede with the Arusha transition. These fears were fueled by Burundi's political elites, both Hutu and Tutsi, who almost always put the Hutu-Tutsi divide forward as the justification for coups, assassinations and insurgencies. Ethnic appeals became a powerful force for legitimization and mass mobilization. For instance, Captain Micombero, a non-aristocratic Tutsi who in 1966 declared the abolition of the monarchy and established a military dictatorship, was a low caste Tutsi who used his "Tutsiness" to garner support from high caste Tutsi. He "used ethnicity for reasons which were far from being abstractly ideological. In fact, the new military course reflected a marginalization of the old elites and a rise of new groups of parvenus. The old ganwa elite became irrelevant, thus removing a buffer between Tutsi and Hutu." (Prunier 1994) Differential social opportunities: exclusion and inequity in a context of extreme poverty and resource scarcity. Differential social opportunities have significantly exacerbated the ethnic, social, economic and regional divides described above. The cleavages have not by themselves led to violent conflict. But in the context of deepening economic depression, the concentration of economic and political power in the hands of a narrow urban elite, and conflicts related to access to land, widening social disparities have over time dramatically increased the potential for violent conflict. Ultimately, the most destabilizing factor has been the exclusion of Hutus from virtually all of Burundi's key political, economic and social institutions--a practice that began in 1965 and reached an unprecedented level after the 1972 massacres targeting educated Hutus. Social capital: weak bridges, strong bonds. Bridging (or inter-group) social capital was traditionally important, including across clan and ethnic boundaries. Indeed, Burundi is one of very few contemporary African states with a national identity that preceded European colonialism: "the monarchical system had succeeded since the 17th century through a subtle interplay of alliances with all the ethnic groups as well as with some clans thus establishing a genuine Nation-State." (Gahama 2002) However, this bridging social capital was heavily depleted in the course of 40 years of violent conflict. Growing fear and mistrust between Hutus and Tutsis, deepening divides among regions, between Bujumbura and the provinces, between the "political class" and the mass of the population, and the politicization of the abashingantahe institution seriously eroded Burundi's traditional capacity to resolve conflicts non- violently.6 By contrast, bonding (intra-group) social capital grew over the 1965-94 period, but manifested itself in mainly negative ways. The manipulation of ethnicity, and the successive waves of violence, hardened "us versus them" rhetoric and provided powerful means of in-group mobilization, as demonstrated in the massive killings of 1965, 1972, and 1993. However, in the post-Arusha period, hardliners in both the Hutu and Tutsi camps appear to have lost much of their leverage, with both sides concluding that military victory is unattainable and that negotiated political agreements are required. The renewed contemplation of peaceful co-existence has--at least for the moment--slowed the group identity-building dynamics at work since the 1960s, and increased opportunities for inter-group cooperation. Although this peace- building dynamic has not significantly impacted group identities, inter-group elite exchanges are 6These divides are only structural conditions. Some, not all, have been triggered by political mobilization. 7 widening through newly inclusive transitional political institutions and also through Track II initiatives such as the Burundi Leadership Training Program. Burundian politics today continue to be dominated by the Tutsi/Hutu divide, Burundian leaders have yet to tackle the structural inequities that underlie the Burundi conflict, there are still hardliners in both camps who remain potential spoilers, and the prospective elections could increase ethnic tensions. Yet, political discourse is increasingly focused on the objective of inter-ethnic cohabitation and power sharing. While there are still major hurdles to overcome, Burundians now appear committed to effecting the compromises required for a durable peace and sustainable economic recovery. A history of violence and impunity. Chronic waves of pre-emptive violence and revenge killings, together with the absence of any sense of judicial impartiality and legal accountability, have produced a culture of impunity, deeply ingrained inter-ethnic grievances, and mutual fear and mistrust. The proliferation of small weapons throughout the population is yet another source of fear and tension. Many citizens have lost confidence in the judicial system's ability to provide even basic protection and assume that the courts are corrupt, lack independence, and are incapable of providing impartial adjudication. Weaknesses in the formal judicial system have only been partially compensated by traditional dispute resolution mechanisms, such as the abashingantahe; the politicization of this institution in recent years has caused it to lose much (though not all) of its traditional moral power. The perceived absence of justice and accountability for acts of violence has given rise both to pre-emptive murders (where people are driven to strike before they themselves are targeted) and to revenge killings (in the belief that there is no other way justice will be done). Establishing the rule of law and credible institutions of justice are essential requirements for a durable Burundian peace. Governance and political institutions: a poor track record. Burundi's history of political assassinations, coups and uprisings is testimony to failed governance and institutional weakness. While there have been positive historical moments--such as the ethnically inclusive independence government formed by Rwagasore, and the 1993 democratic transition that brought to power Burundi's first democratically elected Hutu president--these moments have been short-lived. Authoritarianism, corruption and nepotism have shaped institutions and relations between government and citizens. While the ongoing transition appears to be moving in a more positive direction, both in terms of power sharing and in terms of the development of legitimate mechanisms for conflict management, at least at the national level participation and accountability mechanisms remain undeveloped and uncertain. There is a huge chasm between the country's national leaders and Burundi's grassroots. Decades of conflict and a brutal civil war that took a terrible toll on the civilian population have yielded a deeply alienated and cynical population that views its leaders as self-serving, corrupt, and unresponsive. The rebuilding of public confidence in the institutions of government will take time and considerable effort. In this regard, one distinctly positive development has been the emergence over the past decade of a more vibrant civil society--consisting of human rights groups, dynamic youth and women organizations, active church institutions, and grassroots non-governmental organizations focused on reconciliation and community development. Also noteworthy is the enthusiastic reception given by the leaders of Burundian political parties to recent Burundi Leadership Training Program (BLTP) workshops designed to build their capacity to work cooperatively and to effectively manage the first post-transition elections. Slowly but surely, a more participant-based, bottom-up political ethic is taking root in what has, historically, been an authoritarian political system. Further reinforcing this democratizing trend has been the emergence of new, relatively independent, media outlets. Until recently, Burundian media were tightly controlled by the state. When the state monopoly on the media was lifted in 1992, there was an explosion of highly partisan and ethnic-based media that occasionally became the vehicles of hate speech (e.g., Radio Rutomorangingo), albeit not the 8 scale of what transpired in Rwanda. Now, however, there is a growing use of independent radio as a peace-building tool. One such example is Radio Publique Africaine, offering programs on a broad number of social issues and recognized as a symbol of Hutu/Tutsi reconciliation. Human security: the cycle of conflict and poverty. Human security, including physical, food, social and environment security, has been precarious for decades. The lack of human security has deepened poverty and increased the potential for ethnically based mass mobilization. The restoration of physical security, for the most part accomplished now in 16 of Burundi's 17 provinces, is the first condition for economic recovery. While the army and most of the armed rebel groups have begun a program of disarmament, demobilization, reinsertion and reintegration, the rebel FNL continues to launch attacks within Bujumbura Rurale. Moreover, one legacy of the decade-long civil war is a heavily armed civilian population, in the form of Tutsi and Hutu militias and hill-based self-defense groups. Two tasks are central to the goal of a stable, enduring Burundian peace: the successful reform of the national army and police, to ensure that they serve as cohesive, professional, independent security forces, serving equally all elements of the population and symbolic of the unity of the nation; and the disarmament of the civilian population. While there is a long road still ahead, there has been some encouraging forward movement. Particularly noteworthy have been BLTP training programs requested by key military and police officials, as a means of building the collaborative and management capacities of their newly-integrated high commands. While it will take considerable time and effort to build public confidence in the professionalism and impartiality of Burundian security forces, the foresight and courage displayed by these leaders are cause for optimism. Economic Structure and Performance: Competition in a Context of Extreme Poverty and Concentration of Power Economic growth and inequalities. One of the poorest countries in the world at the time of its independence, in recent years Burundi's agriculturally-based economy has been further impacted by chronic violence and massive displacements. Burundi's average economic growth rate of 4% before 1993 turned negative over the rest of the decade, with per capita GDP reduced by half between 1993 and 2001. At a rate of 2.4% over 1980-2000 (6% in urban areas). Burundi's population has been expanding far more quickly than its economy, resulting in a declining GDP per capita. The population explosion has also led to the division of already scarce land into smaller sized plots and to further environmental degradation. In addition, Burundi's small, open economy was affected by a 66% decrease in international aid between 1996 and 2003; a sharp decline in the prices of coffee and tea, its two main exports; the embargo imposed by six neighboring countries (1996-99); substantial increases in military expenditures; higher costs associated with serving Burundi's external debt; high inflation (40% in 1998); and by a 20% devaluation of the Burundian Franc in August 2002. All combined to shatter the economy. Of all the factors contributing to the depressed state of the Burundian economy, none was more important than the civil war--which was the direct consequence of a massive failure of governance. Burundi's failing economy, the country's rapid population growth, and the concentration of economic and political power in the hands of a small sub-set of the Tutsi minority, have combined not only to severely limit socioeconomic aspirations for Tutsi and Hutu alike, but also to make Burundi's ethnic and regional cleavages more acute. For the distinction between society's "haves" and "have nots" has tended to coincide with these cleavages and make even more significant the link between dominant economic 9 interests and the control of the state machinery.7 For Burundians the state has come to represent access to wealth and most opportunities for upward economic mobility have been centered in the public sector. Access to public and private sector jobs, however, is conditioned by access to education, which in recent years has been highly biased in ethnic and regional terms. Moreover, Burundians understand quite well that the economic mismanagement of the country has contributed both to the depressed state of the economy and to the country's severe economic and social inequities. While national income was declining, military outlays were expanding--to the severe detriment of urgently required social expenditures. Household poverty doubled during the war years.8 Conflict-induced poverty. Burundi has experienced a tragic cycle of poverty and conflict--poverty made the population more receptive to ethnic-based mobilization, and conflict gradually destroyed human, physical, and social capital. The conflict also created a growing population of vulnerable people, including refugees, internally displaced persons and disenfranchised youth. It is estimated that war and inter-communal massacres have claimed over 300,000 Burundian lives since the country's independence; another 800,000 became refugees in neighboring countries; while some 700,000 were internally displaced. Significantly, however, not all of Burundi has been equally affected by the violence and ethnic polarization; consequently, local political dynamics vary across different regions. Competition over scarce natural resources has always been an additional source of tension in Burundi, an agricultural society with 92% of the population living in rural areas. The diminution of available productive land, combined with the massive return of refugees and displaced persons, will almost certainly produce significant new tensions and conflicts over land. Table 1: Selected Economic and Social Indicators, 1990-2002 Indicators 1990 1992 1994 1996 1998 2000 2002 Real GDP na 2.7 -3.8 -8.3 -1.0 2.1 2.7 GNP pc (current US$) 214.4 179.6 170.8 131.8 129.3 119.1 110.0 Population (m) 5.3 5.6 5.9 6.0 6.3 6.7 7.0 % Urban 6.3 6.8 7.2 7.5 7.8 8.2 9.0 % Rural 93.7 93.2 92.8 92.5 92.2 91.8 91.0 Population below poverty line (%) 34.9 33.5 47.8 60.7 65.6 68.7 68.0 Population density (persons/sq. km) 211.4 224.1 234.6 243.1 251.6 266.2 280.8 Arable land as % of land area 58.8 56.7 60.0 60.0 54.5 55.0 55.0 Labor force as % of total population 45.0 44.4 44.1 44.2 44.4 44.6 49.0 Public sector as % of structured sector 36.7 36.8 36.6 36.2 37.1 36.9 33.7 Private sector as % of structured sector 35.2 35.1 35.2 35.4 34.9 34.9 35.0 Source: Government of Burundi, Permanent Secretariat for Monitoring Social and Economic Reforms. External Forces: A Challenging Neighborhood, with Cross Border Feedback Effects The Great Lakes Region is embroiled in national conflicts that have regional dimensions. DRC, Rwanda and Burundi have been staging arenas for each other's conflicts and hosts for each other's refugees. As a consequence, the three neighbors feed a regional conflict psychology revolving around constructed 7In 1985, less than 20% of the student population of the National University was Hutu; one commune of Bururi province, Mugamba, accounted for 15% of the 6,000 students of the University of Burundi. In addition, most of the educated Hutu population was massacred in 1972, or fled to neighboring countries. It is not surprising, therefore, that the gaps between Hutu and Tutsi are large in terms of both employment and income (Nkurunziza and Ngaruko 2002). 8It should be noted that inequalities in Burundi are also viewed in non-ethnic terms, with public resentment being expressed toward those who seem relatively well off, regardless of their ethnicity. Bujumbura-based members of the "political class" are particularly subject to such attacks. 10 identities, with malleable ethnic solidarities such as among Congolese Banyamulenge (Hutu and Tutsi Congolese of Rwandan origin), Rwandan and Burundian Tutsi. Military and political developments in Burundi and Rwanda have had particularly important feedback effects. Thus, the "Hutu power" revolution in Rwanda both gave inspiration to Burundi Hutus who felt excluded from key institutions in the post-independence period, while at the same time, Burundian Tutsis came to fear they might experience the same violent fate as their Rwandan ethnic brethren. As for the DRC, it came to serve as a rear base and safe haven for both Rwandan and Burundian Hutu rebel groups. Burundian and Rwandan fighters allied themselves with the Congolese army resisting the armed intervention of Uganda and Rwanda in the DRC. They received training, arms, and funding in payment for their contribution to the DRC war effort. Still another regional state, Tanzania, has been directly impacted by the Burundian conflict. Hundreds of thousands of refugees (7% of the population of Burundi in 1999) fled to Tanzania, and the Hutu armed rebellion drew many of its recruits in recent years from the Tanzanian refugee camps. In addition, many Tanzanians sympathized with the Hutu struggle against Tutsi hegemony, and some gave at least tacit support to Hutu armed groups. II. CONFLICT DYNAMICS: STRUCTURAL CAUSES, TRIGGERS, AND FEEDBACK EFFECTS The relationships among these variables are essential to understanding the overall dynamics of the conflict. One way of comprehending these relationships is by distinguishing between structural and triggering causes of conflict, looking at discrete events over a long period. Table 2, depicting key Burundian events over the past three decades, illustrates this approach: Table 2: Conflicts after Independence Year Duration Deaths Refugees Region affected 1961 Assassination of Prince Louis Rwagasore 1963 1 month n.a. n.a. Localized conflicts 1965 Assassination of Prime Minister Ngendadumwe 1965 Hutus win parliamentary majority but King refuses to appoint a Hutu Prime Minister; Hutu police coup attempt suppressed 1965 1 month 5,000-25,000 n.a. Center (Muramvya) 1966 1966: Coup and abolition of monarchy 1966 1 month n.a. n.a. Localized conflicts 1969 1 month n.a. n.a. Localized conflicts 1972 Killings of educated Hutus 1972 Assassination of Ntare V 1972 4 months 150,000 300,000 Whole country 1973 1 month n.a 300,000 Localized conflicts 1976 Bagaza coup 1976 1 month n.a. n.a. Localized conflicts 1987 1 month n.a. n.a. Localized conflicts 1987 Buyoya Coup 1988 2 months 5,000-25,000 20,000 - 60,000 North (Ngozi, Kirundo) 11 1991 Hutu local uprisings and Tutsi Army repression 1991 1 month 1,500-5,000 40,000 Cibitoke, Bubanza, Bujumbura 1992 1 month n.a. n.a. Cibitoke, Bubanza 1993 Assassination of President Ndadaye 1993 96 months 150,000-200,000 250,000 - 800,000 Whole country n.a. = not available Sources: Minorities at Risk Project, University of Maryland; Human Rights Watch, Economist Intelligence Unit, United States Institute of Peace; International Crisis Group; and Nkurunziza (2001). Looking at the discrete events in Table 2, the triggering causes for each of the main onslaught of massive violence have been: · Elite-inspired political violence, including four political assassinations, the 1972 massacre, three coups or attempted coups; · Mass mobilization along ethnic lines; and · Events in the sub-region that have also offered opportunities for the unleashing of violence (violence in Rwanda, war within the DRC, generalized cross-border refugee flows and violence). The structural causes of each of these violent episodes have been essentially the same, though varying in relative importance: · Structural social, ethnic, regional and economic cleavages and inequity in access to social services and employment, in a context of extreme poverty and concentration of power; · The cumulative legacy of violence, impunity and failed governance and political institutions; and · The deepening cycle of conflict and poverty and threats to security, which make the population, not the elites, the principal victims of the conflict and increase the potential for mass mobilization. In sum, the struggle for power in a context of extreme poverty and scarcity of resources has featured elite manipulation of ethnicity as an instrument of mass mobilization. Box 2 illustrates the many feedback effects which have produced a particularly vicious conflict cycle in Burundi. 12 Box 2: Causal Relationships, 1996-2004 Extreme poverty and Grievances resource scarcity Colonial divide and rule Ethnic divisions Hutu power revolution Mass mobilization in Rwanda along ethnic lines Mass violence (inter- communal massacres, genocide, civil war) Unequal access to services and opportunities Post-independence Failed institutions and high stakes struggle for use of political violence control of the state Limited impact of external pressure for better governance III. CONFLICT DYNAMICS TODAY: ON THE BRINK Four years into the transition that began in November 2001 and is scheduled to end in 2005, some factors are working to dampen the conflict, while others continue to threaten its escalation. In some instances, the impact of conflict-related variables is as yet uncertain. Potentially De-Escalating Factors · War fatigue and receding political violence, with 16 out of 17 provinces currently free from violent conflict, a pattern reinforced by the initiation of security sector reform and the Demobilization, Reinsertion and Reintegration (DRR) Program; · The fact that the war has meant a serious decline in both income and security for almost everyone, so that most Tutsi and Hutu are effectively now "in the same economic boat"; · Initiatives to bridge ethnic cleavages at the central level, through power-sharing and confidence- building; 13 · Training programs initiated by both the army and the new independent police force to strengthen the cohesion and collaborative capacity of the high commands of the two institutions; · The emergence of independent media and of ethnically mixed civil society organizations; and a large number of community-level initiatives, focused on leadership training, reconciliation, and the management of conflict;9 · The apparent marginalization of many Tutsi hardliners; · The significant ONUB international peacekeeping presence, apparently welcomed by most Burundians, and the donor community's involvement in post-conflict reconstruction; · The efforts of key Burundian political figures to continue power-sharing negotiations, even after parliamentary ratification of the interim constitution that was just approved by public referendum; · Joint initiatives taken by leaders of Burundian political parties, including their participation in workshops designed to strengthen their cohesion, their successful collaboration in drafting an electoral code of conduct, and their issuance of a press communiqué to assure the public that they were collectively committed to work for elections without violence or intimidation; and · The successful management of both the Constitutional referendum and the commune elections; in the case of the latter, aside from six communes which in which voting was delayed by violence, close to 75% of the population turned out in what were viewed by most observers as free and fair elections. Potentially Escalating Factors · The upcoming national elections carry the risk of escalating ethnic tensions. Both Hutu and Tutsi parties may be tempted to portray themselves as the stronger advocate of their region or their ethnic or clan group's interests, thereby giving a sharply regional, ethnic or clan cast to the party competition. Moreover, in a country in which violent solutions and impunity have become the norm, and in which the loss of state power is perceived to have devastating consequences, elections may well be an invitation to further violence;10 · Extreme land scarcity, compounded by the return of refugees and internally displaced persons since the end of 2003; · The tendency of aid organizations to unintentionally exacerbate ethnic and class cleavages by the uneven geographic location of services and opportunities; · The resistance of some Tutsi elements who may fear a diminished economic status or are anxious about how unresolved justice issues may be handled by a future government, or who, with the 1993 inter-communal massacres in mind, may fear that their very survival is in danger if Tutsis lose control of the army and the state; · Continued impunity and the absence of any serious agenda to put an end to the cycle of violence, the absence of accountability for violent acts, and the resulting legitimization of the use of force; · The continued intransigence and violence of the FNL, the one armed group that remains outside the peace process; 9One particularly interesting initiative is a two-year pilot Community Based Leadership Program being undertaken in Ruyigi and Gitega Provinces, with the financial support of the Office of Transition Initiatives (USAID). With the goal of advancing the community reintegration of refugees, displaced persons and ex-combatants, the CBLP combines three principal elements: leadership training of local leaders, small grants for community development, and vocational training centers. This initiative involves a partnership between PADCO; African Strategic Initiatives, a local Burundian NGO; the Wilson Center; and the BLTP. Search for Common Ground also has played a major role in developing among grassroots communities the skills required for non-violent conflict resolution. 10It is time that donors rethink their conventional preoccupation with elections as a central mechanism of political transition. Indeed, it could be argued that donor pressure for expedited elections in 1992-93 may well have contributed to the tragedy that subsequently unfolded. 14 · Destabilizing trends in DRC that could generate further cross-border violence and refugee flows, such as occurred in August 2004 with the influx of Banyamulenge from South Kivu in DRC and their subsequent massacre in Burundi; · A possibly negative Rwandan regime reaction to a successful Burundian transition and to the emergence of a Hutu majority government on its border;11 and · The proliferation of small arms that pose a continued threat to peace, given that mass mobilization is still possible, and increased acts of domestic banditry and criminal (non-political) violence. Several Factors Are "On the Brink" with Currently Uncertain Impact · The highly sensitive Security Sector Reform process, essential to both the creation of integrated command and control structures and the restoration of the legitimacy of the army and police, carries evident risks that must be well-managed. Close attention must be paid, for example, to the economic anxieties and frustrations of ex-combatants, who fear the consequences of demobilization or who may have unrealistic expectations about what they might secure from the DRR Program. · Measures to improve equity of governance and political institutions and forms of consociational democracy, including the protection of minority rights, are being publicly debated; however, no consensus has yet been reached and popular participation in this debate has been extremely limited. Without initiatives to strengthen Burundi's conflict management capacity, to build trust among key leaders, to strengthen collaborative decision-making capacities across ethnic and political boundaries, and to address the stark divide between urban elites and the rural mass of the population, the structural inequities underlying the Burundian conflict will not begin to be addressed. · Ethnic and regional cleavages are being openly discussed, but there has been little programmatic effort to address underlying issues, and baseline statistics capturing ethnic and regional disparities have never been compiled. · In the framework of consultations on the Poverty Reduction Strategy, there are some attempts to improve vertical accountability, but there is still room for strengthened public participation; · There is not yet a nationally owned program to reform the judicial system and to establish accountability for the war crimes and human rights abuses that have been committed over the past four decades. · Economic growth resumed in 2001 (2.1% percent in 2001 and 4.5% in 2002) but declined to 1% in 2003. While there is anecdotal evidence of strengthened economic performance in 2004--in part the result of the enlarged international peacekeeping presence--population growth has exceeded economic growth rates and economic growth has had little impact on the poorest of Burundi's population. Burundi remains one of the poorest nations in the world, and the country's extreme poverty remains one of the most significant contextual realities impacting political dynamics in Burundi. · Burundi's future cannot be divorced from unpredictable developments in post-genocide Rwanda and in the still extremely fragile peace process in DRC. In the post-Arusha period, extreme poverty, environmental scarcity, a continuing high stakes struggle for control of the state, and continuing unequal access to services and opportunities remain of concern, as do unresolved issues of justice and accountability. In the context of the political transition and raised expectations on the part of the population, the status quo itself carries a risk of conflict escalation. The 11On the other hand, Burundian success might conceivably encourage the Rwandans to re-evaluate their rather more authoritarian post-conflict political strategy that seeks to suppress all discussion of ethnic political differences. 15 large number of "on the brink" factors suggests that there are still many opportunities for development assistance and traditional and Track II diplomacy to assist in the consolidation of Burundi's peace process. Notable targets of opportunity include: current highly unequal access to social services, employment and productive resources; the lack of opportunities in rural areas and the urgent need to strengthen the agricultural sector; the glaring absence of will to break Burundi's history of violence and impunity which leaves the use of violence as a possible recourse; the lack of rooted, nationally-driven justice and reconciliation processes; and the limited nature of vertical accountability and public participation. In addition, a concerted international effort to stabilize the broader region will be an important contribution to Burundi's own future peace and stability. IV. AREAS IN WHICH DEVELOPMENT ASSISTANCE CAN SUPPORT THE TRANSITION AND OPERATIONAL ISSUES TO CONSIDER Development reports in the months preceding the 1994 genocide in neighboring Rwanda were colored by unrealistic assessments of the intentions of belligerents and underestimated both the need for deep reforms to address structural inequities in Rwanda and the multiple risks of violence triggers on the political and security fronts. Although in many ways the situation in Burundi is radically different, its structural inequities are only beginning to be measured and addressed, and there remains a risk of renewed violent conflict, especially in the context of the first post-transition elections. Several scenarios are in fact possible: (i) renewed violent conflict, (ii) continuation of chronic conflict with periods of peace interspersed with violent episodes, (iii) cosmetic stabilization "from above" which ignores remaining fundamental inequities, and (iv) consolidation of the transition by a serious effort to address these inequities. Given the importance of development assistance in Burundi (24% of GNP in 2002)12 and its potential role in addressing Burundi's structural inequities, but also some short-term, triggering causes of violent conflict, it is essential to look at the role of development assistance vis-à-vis these possible scenarios. This section identifies the implications of the analysis of the Burundian conflict for the development agenda, both for prioritizing programs that will help in consolidating Burundi's transition, and for sensitizing donors to key operational issues at the project level. Development Assistance and Diplomacy in War-to-Peace Transitions Development assistance and diplomacy are complementary approaches in war-to-peace transitions. Table 3 lists typical factors of conflict,13 distinguishing those that development assistance has traditionally been able or expected to address, and those that have traditionally been tackled through conventional and Track II diplomacy. The table aims to provide an overall picture and does not distinguish among interventions by bilateral donors, multilateral institutions, UN agencies, or NGOs; however, it should be noted that there are significant differences in the respective mandates of these entities and none has a mandate to tackle all causes of conflict. The table takes into account recent innovative developments such as the shift in focus from post-conflict reconstruction to conflict prevention and reconstruction, and new approaches that have been developed in recent years, such as the community-based integration of displaced persons and ex-combatants, new training modalities designed to enhance the collaborative capacities of formerly belligerent parties, new financial instruments (e.g., grant programs for small, quick and high-impact projects; the Multi-Donor Regional Trust Fund for DDR), and new partnerships between conventional diplomatic entities and non- governmental organizations. 12World Bank World Development Indicators database, August 2004. 13Adapted from the Conflict Analysis Framework: www.worldbank.org/conflict. 16 Table 3: Causes that Development and Diplomacy/Track II Interventions Can Address Cause Development assistance Diplomacy/Track II Social, economic, ethnic and Public expenditure reviews, monitoring and Inter-group elite facilitation and mediation regional cleavages evaluation using peace and conflict indicators; investments targeted to assist disadvantaged groups; leadership training, with particular emphasis on techniques of conflict management and mitigation Differential social Sector programs with explicit social equity Negotiation with elites opportunities (e.g. objectives education, health) Bridging/bonding social Bricks-and-mortar inter-group projects; peace- Training in conflict management and capital, group identity- building media projects; promotion of fair and mitigation; building & myth-making professional media History of violence and Training in conflict management and mitigation; Training in conflict management and impunity support for judicial system reform and capacity- mitigation; International Criminal Court; building; support for truth and reconciliation support for truth and reconciliation processes processes Governance and institutions Budget support, capacity-building training Negotiation with elites initiatives The rule of law and the Budget support, capacity-building training Negotiation with elites judicial system initiatives Links between government Participatory requirements in development Negotiation with elites and citizens programs Human rights Support for human rights advocacy groups, Negotiation with elites; ending impunity support for judicial system reform and capacity- (governance; war crimes tribunal) building, support for security sector reform Militarization of society and Financial and technical support for Facilitating negotiation of regional arms small arms proliferation demobilization and reintegration programs control regimes; continued conditionality of IFI programs requiring reduction in military expenditures Economic structure and Support for community-based development and Facilitating the negotiation of new regional performance social protection programs; expansion of the economic compacts and institutions Heavily Indebted Poor Countries initiative; promotion of economic diversification and transformation industries; support for employment projects; technical assistance for land reform Environment and natural Natural resource management Negotiation of regional environmental resources (including land) compacts regarding water and other resources External forces: Regional Development programs, with clear incentives for International arms embargoes; the negotiation conflicts; Role of kindred regional cooperation and integration; support of a new regional security architecture; the groups outside country; Role and technical assistance for regional trade and engagement of international monitors and of Diasporas investment agreements; facilitation of regional peacekeepers networking among professional and other social sectors; facilitating diaspora involvement with development projects There is a wide range of areas in which development assistance can help address some of the sources of conflict. The challenge is to ensure that the development agenda is well prioritized and sharply focused. Prioritizing a few initiatives that have high pay-off is preferable to attempting to address all issues simultaneously; a scatter-gun approach to development assistance is likely to yield many initiatives of lesser impact. A dynamic conflict analysis can greatly assist in prioritizing development assistance by focusing attention on these key questions: Ø Whatarethestructuralandtriggeringcausesofconflict? Ø Which development and/or diplomatic tools are best suited to addressing these causes? How can development and diplomatic tools complement each other to address these causes of conflict by applying different strategies and tools, with different mandates and time frames? In particular, in 17 their analysis development experts and diplomats should go beyond the factors they each traditionally consider. For instance, although diplomacy best handles spoiler management, development planners should include the role of spoilers in their own analysis of risks to the projects they are contemplating. Ø Whattypeofdevelopmentprojectscancontributenotonlytoeconomicreconstructionbutalso to the consolidation of the peace process and long-term political stability? Ø Whatinitiativescanmosteffectivelyaddresstheunderlyingstructuralcausesoftheconflict? Ø As initiatives that support stabilization may in some instances contradict initiatives to address structural causes, what is the proper sequencing of development and/or diplomatic initiatives? Ø What is the required time-frame for the effective implementation of a sustainable development agenda? It has been reported that countries emerging from conflict have a 44% chance of relapsing into violence during the five years following the signing of peace accords; yet, post- conflict aid, often flooding in at the first sign of a negotiated peace agreement (providing as much as 95% of GDP), typically declines sharply to a trickle during the following five years (Collier et al. 2003). Ø How should each project be designed, implemented, monitored and evaluated to ensure that peace-building objectives are reached? With appropriate attention to such operational issues, a bricks-and-mortar project can contribute to the consolidation of peace. Considerations for the Development Agenda in Burundi The Arusha-sanctioned transition has opened a window of opportunity for the international community to help address some of the structural causes of the Burundian conflict. Moreover, the development community is assigned a role in overseeing the implementation of the Arusha Agreement, through the UN-led Implementation Monitoring Committee and the Bujumbura-based donor coordination unit for the implementation of Protocol IV of the Arusha Agreement. Given the analysis of the causes and dynamics of the conflict in Part I, an effective donor development strategy might consist of the following: · First, "do no harm" and in particular avoid reinforcing structural or triggering causes of the conflict. Care must be taken to ensure that aid flows and programs do not have the unintended consequence of perpetuating the structural inequalities of the past. The risk of aid being captured by the elites and increasing the already exceptional concentration of economic and political power is high. There is also some evidence that geographic areas that have better infrastructure and more experienced staff in place may be "favored" locations for aid headquarters and services, unintentionally exacerbating local ethnic and class tensions. In the context of the pending elections, donors must also take into account the risk of aid being used for electoral purposes and fueling competition between groups; in addition, donor assistance should be directed to training initiatives that bring together political party leaders, that focus on the development of negotiations, communications, and collaborative decision-making skills, and that can help strengthen their personal relationships. The focus should be on encouraging good "process" and on promoting an inclusive forum, rather than on the extension of assistance to specific parties or groups. · Second, while the conjunction of diplomatic, security and development initiatives can help consolidate the return of peace, development assistance can increase its sustainability by making the dividends of peace visible to the population. Improving the equitable access to 18 social services, fostering economic growth and diversification, stimulating employment--all can calm the political climate and deepen public confidence in the peace process. Public sector reform, including anti-corruption and transparency initiatives, can be a means of addressing the high level of concentration of power in the hands of a few families. Elite accountability, at least at the local level, may be further strengthened through the expansion of community-driven development projects. At the same time, initiatives to encourage private sector development can also assist in the dispersion of power from the state sector. · Third, challenging the view that development assistance is limited to medium-to-long term objectives, development assistance in Burundi can address some issues that are critical in the short-term. For instance, bilateral and multilateral development assistance can significantly contribute to the restoration of security through support for the demobilization and reintegration of ex-combatants (as through the Multi-Country Demobilization and Reintegration Program for the Great Lakes region),14 and through the provision of technical assistance for security sector reform. Indeed, security must be seen as the most immediate priority--because it is security that will create the possibility for the economic reconstruction and social reconciliation work to begin. It is security that brings with it the first dividends of peace: the opening of markets, the expansion of agricultural production, the resumption of trade, the emergence of new investment opportunities--all become possible with a secure environment. In this regard, the UN mission in Burundi, ONUB, has the critical mission of making certain the ceasefire is respected and that the security conditions permit the safe provision of humanitarian aid. · Fourth, development assistance can limit the potential for mass mobilization in both rural and urban areas by targeting youth,15 by focusing on labor-intensive work projects and medium-term employment generation, and by investing in training programs to strengthen community-level conflict management capacities. · Fifth, development assistance can address some of the structural causes of the conflict, by: o Actively promoting inclusion, and by incorporating measurable objectives of equitable distribution of jobs, resources, and access to social services, across ethnic groups and regions--at all levels of government; o Decisively putting governance on top of the agenda, and encouraging participant-based processes, seeking to counter the deeply ingrained elite-driven decision-making that, historically, has yielded the pattern of exclusion and inequities that have given rise to the Burundian conflict; and o Seeking to reverse the pattern of state-centered economic concentration that favored a small sub-set of the Tutsi minority (primarily from Bururi) at the expense of most Tutsi and the majority Hutu. In particular, the creation of economic opportunities outside the state is essential to defuse the tension surrounding competition for control of the machinery of government, and systematic efforts must be made to overcome the social effects of the discriminatory and exclusionary policies that have disadvantaged the Hutu majority. · Sixth, development assistance must give particular attention to the perceptual and attitudinal legacy of the Burundian conflict, and incorporate capacity-building initiatives designed to assist Burundians restore their sense of national community, rebuild relationships of trust, and develop a new consensus on how power should be shared and how decisions should be made. Unless Burundians are able to move away from the zero-sum paradigm induced by their war and by earlier poor governance, in which each side has come to see its survival as requiring 14http://www.mdrp.org 15The population under 15 years old was 47% in 2002. 19 the defeat of the other, to a new shared recognition of common interests and interdependence, Burundians will be unlikely to seriously address the structural inequities underpinning their conflict. Capacity-building initiatives that are focused on strengthening collaborative decision- making skills--both among elites and within grassroots communities--together with leadership training designed to make elites more democratically accountable, are the best means of inculcating a culture of non-violent conflict resolution. Absent attention to this "process" dimension in development programs, peace will not be sustainable, and bricks-and-mortar investments will mean little. In addition, to address continuing memories and fears of collective violence, a priority development objective must be to assist Burundians build an inclusive, reformed legal system, and to put in place institutions designed to facilitate both accountability and reconciliation. Particular attention must be paid to the need for constitutional and judicial protections for minorities. · Seventh, development assistance needs to be consistent and sustained. Although the current trends are, for the most part, positive, the Burundian peace process must still be considered fragile. Internal fiscal resources are very limited, and continued and extensive external support will be essential if urgent social issues are to be addressed and a return to conflict avoided. The typical donor response to post-conflict situations, which sees a sharp decline of foreign assistance before peace is consolidated, must be resisted (Collier 2003). The efforts of Burundians to overcome decades of traumatic conflict and inter-communal violence will require not only considerable courage and skill; it will also require a sustained partnership with the international community. · Eighth, a donor development strategy for Burundi must also take into consideration Burundi's regional context. A sustainable Burundian peace and the country's economic reconstruction will depend both upon the political stabilization of the Great Lakes region and upon Burundi's integration into a broader regional economic framework. As in the MDRP for the nine countries of the Greater Great Lakes sub-region (World Bank 2002), the donor community can play a major role in encouraging all regional states to recognize and act upon their common interests in building a new regional security architecture and a regionally integrated economic framework. The Government's development strategy, including the PRSP process, and donor pledges of assistance following the January 2004 donor meeting in Brussels offers many opportunities to address the above elements. The following review of development initiatives currently in process is illustrative: · Rural development. Current projects seek to target the coffee and the tea sectors by improving product quality, productivity, and marketing. Agricultural diversification is an important objective, with significant potentials to be realized in fishing, livestock, off-farm processing and value-added crops such as vegetables and flowers. · Infrastructure. Now that physical security is gradually being restored, the development of basic infrastructure, including secondary roads, would improve food security and access to health services. However, there must be regional balance in the distribution of infrastructure projects: capital-centric, urban-biased projects risk aggravating some of the structural imbalances described in Part I. · Security sector reform and demobilization and reintegration. Security sector reform, supported by ONUB and key international partners, is ongoing, together with the DRR Program started in mid-2004 with World Bank and MDRP support. These programs are critical to the goals of 20 building a united, professional, apolitical army and an independent national police force, and to the integration of ex-combatants into Burundian society. Box 3: Examples of Ways the PRSP Can Contribute to the Consolidation of Peace I. The PRSP consultation process can help consolidate peace by: · Institutionalizing the participation of marginalized groups (ethnic minorities, the poor, the displaced, women, youth, neglected regions) including by use of targets and quorums; · Giving these marginalized groups the means and space to express their priorities, e.g., through the provision of training on advocacy and leadership skills, or through the use of focus groups drawn from marginalized populations; · Addressing critical questions relevant to peace consolidation, such as (i) the differential positions of key groups (defined by ethnicity or region) with respect to access to social services, to jobs in the public and private sectors, to access to capital; and (ii) the extent to which community-based development is inclusive and local authorities are accountable; and · Linking with other conflict-sensitive processes (e.g., community-based reconciliation programs, the DRR program, sector-specific and group-specific initiatives targeting such groups as the displaced and ex- combatants), and explicitly identifying synergies and complementarities that can contribute to the consolidation of peace. II. The PRSP content can help consolidate peace by: · Recognizing the factors of conflict (root and triggering causes, including ethnic disparities) and identifying the relationships between peace and development; · Addressing some of the root causes of the conflict through: (i) programs to reduce destabilizing structural imbalances; (ii) inclusive decision-making processes; (iii) conflict resolution mechanisms and training initiatives directed at advancing inter-group reconciliation and collaboration, both locally and nationally; and (iv) economic strategies that are directed at inclusive growth; · Tackling some of the triggering causes for conflict through initiatives targeting disenfranchised urban youth and ex-combatants; and · Developing indicators and collecting statistics that will permit measurement of the extent to which development is proceeding in an equitable, inclusive fashion--considered from the standpoint of ethnicity, region and gender. · The Poverty Reduction Strategy Paper process and large-scale macro and sector program. Macroeconomic stabilization has improved since the installation of the Transitional Government in November 2001, with the implementation of key fiscal, monetary, and trade measures.16 The PRSP process, macro programs17 and large scale sector or multi-sector programs and projects18 also provide opportunities to address structural causes of the conflict, including the ethnic and regional disparities in the distribution of resources and access to social services (see Box 3). However, the weakness of Burundian civil society--particularly the absence of effective advocacy on behalf of the rural population--means that this potential will not be realized absent broad-based leadership training, designed to strengthen the capacity for effective participation in PRSP and related processes. If such training is not undertaken, the participative element of such processes will remain illusory. 16Inflation, which averaged 24% in 2000, fell to less than 9% in 2002. Economic growth resumed in 2001 and reached 5.4% in 2004 (Source : World Bank) 17Such as the Economic Management Support Project and the Economic Recovery Credit Project. 18Such as the Agricultural Rehabilitation and Support Project, the Transport Infrastructure Rehabilitation Project, and the Social Action Projects. 21 · Conflicts over land. Old tensions over land tenure are increasingly acute--the result both of widespread soil erosion and of the massive return to already land-short communities of refugees and internally displaced persons. The Commission Nationale de Réhabilitation des Sinistrés and current programs for vulnerable groups recognize the social and political difficulties inherent in the reintegration of these elements in their communities of origin, and in the additional pressures that are being placed on already limited or nonexistent infrastructure. Investment in community capacity to manage such tensions is critical. · Employment generation. Given the predominance of the agricultural sector, employment generation requires a focus on rural development and the development of sector-specific strategies. From a peace consolidation perspective, public works and employment generation projects targeting youth have particular merit. These projects are also an opportunity to help rebuild the social fabric, in particular by training initiatives that foster community decision- making and inter-group cooperation. · Governance. The Government has launched an ambitious program to improve economic performance and fiscal management. The implementation of these reforms and further transformation, such as through privatization and decentralization, will require determined efforts to overcome vested interests and strengthen the capacity of reformers both in government and in civil society. · Social sectors. Social indicators are alarming. Social sector projects must be sensitive to the need to redress ethnic and regional imbalances and must incorporate indicators to measure the distributional impact of development assistance. V. OPERATIONAL ISSUES TO CONSIDER IN PROJECT DESIGN, IMPLEMENTATION AND MONITORING AND EVALUATION It is not enough to simply allocate funds to the areas identified as priorities and assume that they will ultimately impact the cause of conflict to which they are directed. Change, even if on balance it is positive, often creates losers as well as winners. Peace consolidation also depends on how projects are designed, implemented and monitored, for development assistance may well fuel conflict even as it seeks to address its causes. The analysis of the Burundian conflict in Part I carries a number of implications for project design and implementation. q Development assistance has a role to play in engaging the elites on the side of peace. Conventional peace-building initiatives tend to target the fringes of the population that already are in favor of peace, or at least not structurally against peace, such as women and children; yet such initiatives often fail to engage the elites who tend to be the drivers of national conflict. Creatively structured development assistance should seek to provide incentives for institutional reform and peace-building--both within the government and within civil society--and should assist polarized leaders in rebuilding their relationships and strengthening their capacity to work collaboratively across the lines of ethnic and political division. Financial and technical assistance can also provide incentives for structural reforms, increase the focus on pro-poor growth, and encourage equitable access to social services. If well conceived and rigorously monitored, it can help leverage reform even in the face of entrenched interests. q Development projects also need to invest in a bottom-up approach to strengthen the peace constituency and hold the elites at central and local level accountable. Well-conceived community- 22 driven reconstruction consists of both product and process. On the one hand, a road, a well or a school represents a material response to the immediate community needs. On the other hand, a bottom-up approach to community development also involves a process of inclusive decision-making that will help an ethnically mixed community make further decisions, identify other needs, and peacefully resolve emerging conflicts. Absent this process dimension, development assistance may amount to delivering outputs that may not be sustainable. Broadening inter-group interactions-- through inclusive decision-making processes--increases a society's resilience to conflict.19 Unfortunately, community-driven projects in divided societies do not always pay sufficient attention to marginalized ethnic groups, or to other vulnerable groups such as the poor, the displaced, women, and youth. Donor development strategies must ensure that historically subordinated groups have the means and the space to express their own priorities. If traditional institutions do not permit such participation, new vehicles and mechanisms will be required. q Development assistance should prioritize projects that not only provide bricks-and-mortar but also are peace-building in their effects. Peace-building decoupled from concrete economic reconstruction will be less sustainable than a fully integrated, holistic approach to community development. Many conventional development initiatives are either directed at the building of social capital, without reference to material development, or they are primarily bricks-and-mortar projects that pay scant attention to the need to build social capacity. There is no reason why bricks-and- mortar projects cannot simultaneously serve as vehicles to bring Burundians together, and to increase Burundi's managerial and leadership capacity. In this way, the ability of Burundians to address the underlying structural inequities that have given rise to the conflict will be greatly strengthened (Barron et al. 2004). q The impact of donor-assisted development must be carefully monitored, and assessed in terms of explicit peace-building objectives and indicators. Establishing a baseline is an opportunity to take a hard look at the present status of different groups (broken down by ethnicity, region, rural/urban, etc.) with regards to access to social services; to the distribution of jobs in the public, private and security sectors; and to land. Also, if peace building is an explicit objective of post-conflict reconstruction in Burundi, the indicators of success need to measure factors such as inter-group social capital,20 perceptions of youth, degree of security, and the extent of gender equity. Peace and Conflict Impact Assessments are a useful methodology to measure the impact of development projects on dimensions of peace.21 Finally, in the case of countrywide programs, the regional distribution of program beneficiaries must be monitored to ensure overall balance. q Development assistance strategies must take into account the multiplicity of variables that condition peace processes, and be flexible enough to adapt to the changing environment, including opportunities and setbacks in the peace process. The multiple factors and interventions that impact on peace-building (security, conventional diplomacy, Track II initiatives, private sector activities, population displacement, humanitarian and development assistance, etc.) must be mapped and their inter-relationships must be understood. Technical and political obstacles to peace must be distinguished, and available tools must be matched to the problems requiring resolution. Where can development assistance have its greatest impact? Where must the tools of traditional diplomacy be brought to bear? The links--synergies, complementarities, contradictions, appropriate sequencing-- between development projects and other conflict-sensitive events (e.g., truth and reconciliation processes, initiatives of the Commission Nationale de Réhabilitation des Sinistrés, preparation for national elections), are usually not explicitly identified. The required coordination can take the form 19See for example Colletta and Cullen (2000), Barron et al. (2004), and Cliffe et al. (2003). 20Colletta and Cullen (2000), Grootaert (1998), and Grootaert and van Bastelaer (2001). 21For example, see DFID, World Bank, and USAID Peace and Conflict Impact Assessments. 23 of: (i) joint needs assessments and appeals, such as at the January 2004 donor meeting in Brussels; (ii) the sharing of conflict assessment methodologies and applications; and (iii) a collaborative effort of donors and Burundians to develop a joint vision of conditions, incentives and types of assistance that will reduce the potential for conflict. Table 4: Conflict Variables and Some Key Program and Project-level Implications Conflict Variable Some Implications for the Burundi Some Implications for Project Design Development Portfolio and Implementation Social, ethnic, regional and Establishing mechanisms to coordinate Defining explicit objectives and indicators of economic cleavages security/diplomatic/Track progress in addressing structural inequities. II/humanitarian/development initiatives in Using bricks-and-mortar projects as a vehicle order to address structural inequities. to strengthen inter-group cohesion. Exclusion and inequity in a context Targeting disadvantaged sectors; avoiding Emphasizing inclusive and participant-based of extreme poverty and resource unintended reinforcement of structural decision-making processes. scarcity imbalances. Social capital: weak bridges, strong Prioritization of leadership training at all Focusing on inter-group cooperation in project bonds levels, to strengthen collaborative capacity. conception and implementation. Recognizing importance of "process" in project development. A history of violence and impunity Prioritizing security and justice reform, Emphasizing inclusive and participant-based support for truth and reconciliation decision-making processes processes. A poor track record in governance Prioritizing leadership training, with Promoting bottom-up development (elite and political institutions emphasis on mechanisms to establish accountability, through inclusive, participant- accountability and transparency. Focus on based processes). decentralization, local governance Human security: A cycle of conflict Prioritizing programs that strengthen human Emphasizing projects that will yield concrete and poverty security (physical, food, social) and security improvement (e.g., decrease the potential for mass mobilization professionalization of army and police by generating a concrete peace dividend for structures, projects to increase agricultural the broader population productivity, arms buy-back programs) Economic competition in a context Encouraging economic decentralization and Targeting disadvantaged groups, encouraging of extreme poverty and privatization; addressing educational private sector options concentration of power inequities A challenging neighborhood, with Supporting regional initiatives (cross border Facilitating regional networking, technical cross-border feedback effects repatriations and DRR, curbing illegal assistance for regional integration and weapons trade, regional economic and cooperation security compacts) q Recovery in Burundi and the consolidation of peace requires stabilization of the sub-region. Development can support projects addressing cross-border issues, such as cross-border weapons trade and the presence of foreign armed groups in Burundi, Rwanda or DRC. One such ongoing initiative is the Demobilization and Reintegration Program, which was developed and is being implemented in the framework of the Great Lakes Multi-Country Demobilization and Reintegration Program. As noted above, diplomatic interventions and external technical assistance can play a significant role in moving regional states toward greater economic cooperation, in facilitating inter-state networking in a number of institutional spheres, and in encouraging the construction of a new regional security architecture. 24 Annex 1: Chronology22 1100-1400s: Hutu people settle in the region, imposing their language and culture on the original inhabitants, the Batwa. Tutsi settlers establish themselves as feudal rulers. 1890: The Tutsi kingdom of Urundi and neighboring Ruanda (Rwanda) incorporated into German East Africa. 1916: Belgians occupy the area. 1923: Belgium granted League of Nations mandate to administer Ruanda-Urundi. 1959: Influx of Tutsi refugees from Rwanda following the "Toussaint rwandais," a Hutu insurgency against Tutsi domination in Rwanda. 1961: September - Parliamentary elections, nationalist and ethnically mixed party UPRONA wins 82% percent of the vote. 1961: October - Assassination of Prince Louis Rwagasore, leader of UPRONA who was to become the first Prime Minister of independent Burundi. Independence 1962: Urundi is separated from Ruanda-Urundi, becomes Burundi and is given independence as a monarchy under King Mwambutsa IV. 1965: King Mwambutsa bypasses the results of the May elections, which had given the Hutu 23 out of a total of 33 seats in the National Assembly, and appoints his private secretary, Léopold Biha, as Prime Minister; attempted coup by Hutu police led by the head of the gendarmerie is brutally suppressed, with 2,500-5,000 Hutus killed. 1966 July: Mwambutsa is deposed by his son, Ntare V. 1966 November: Micombero stages a second coup; orders the repression of Hutus and the execution of the leading Hutu leader of the period, Paul Mirerekano; declares the monarchy abolished and pronounces himself President. 1972: Between 125,000 and 150,000 mainly educated Hutus are killed after Ntare V is killed, supposedly by Hutus. 1976: Micombero is deposed in a military coup and is replaced by Jean-Baptiste Bagaza as President. 1981: A new constitution, providing for a national assembly, is adopted. 1987: President Bagaza is deposed in a bloodless coup led by Pierre Buyoya. 1988: Hutu peasants kill Tutsis in two communes; between 5,000 and 20,000 Hutus are massacred by the army in retaliation and 60,000 refugees flee to Rwanda. Dashed hopes 1992: New constitution providing for a multiparty system is adopted in a referendum. 1993 June: Melchior Ndadaye becomes Burundi's first democratically elected Hutu President. 1993 October: Elements of the army assassinate Ndadaye, provoking the massacre of tens of thousands of Tutsis and Army retaliatory killings of an even larger number of Hutus, the internal displacement of 700,000 thousand, and the flight of 600,000 refugees. 1994: Parliament appoints Cyprien Ntaryamira, a Hutu, as President; Ntaryamira is killed in a plane crash together with his Rwandan counterpart; more ethnic violence and refugees fleeing to Rwanda; parliament speaker Sylvestre Ntibantunganya appointed President. 1995: Renewed ethnic violence in the capital, Bujumbura, marked by ethnic cleansing of Hutus and the assassination of Hutu parliamentarians. 22Adapted from http://news.bbc.co.uk/1/hi/world/africa/country_profiles/1068873.stm (May 2004) 25 1996: Former Tanzanian President Julius Nyerere is appointed Burundi peace process facilitator by the OAU and UN, launching the regionally sponsored Arusha process. Pierre Buyoya Returns to Power 1996: Pierre Buyoya returns to power in a second coup, deposing Ntibantunganya and suspending the constitution. 1998: An "internal partnership" between Frodebu and Uprona parliamentarians results in an agreement between Buyoya and the National Assembly on a new transitional constitution under which Buyoya is formally sworn in as President. 1999: Julius Nyerere dies and former South African president Nelson Mandela takes over as Burundi peace process facilitator. 2000: Arusha Agreement signed by 19 delegations--including 17 political parties, the Government and the National Assembly. Some parties register their reservations to some provisions. The CNDD-FDD (Nkurunziza) and the FNL reject the Arusha Accord. Mandela attempts to launch a separate negotiating track with the two armed groups, and mediates an agreement that the 36-month transitional period will be divided into two parts--the first to be led by a Tutsi (Buyoya), the second by a Hutu (Domitien Ndayizeye). It is agreed that neither will stand for the presidency when the post-transition government is formed. 2001 January: Ceasefire talks between the Government and the largest armed group, the CNDD-FDD (Nkurunziza), commence. 2001: Two failed attempts to depose Pierre Buyoya by coup. The Transition 2001 October: Installation of transitional government under which Hutu and Tutsi leaders share power (Arusha Agreement). CNDD-FDD (Nkurunziza) and the FNL continue to reject a cease-fire and fighting intensifies. 2002 December: Government and the CNDD-FDD sign a ceasefire agreement. 2003 January: renewed fighting between the Government and CNDD-FDD. 2003 May 1: Domitien Ndayizeye, a Hutu, succeeds Pierre Buyoya as President without incident. Alphonse Kadege--a Tutsi from the UPRONA party--is designated the Vice-President. 2003 November: President Ndayizeye and FDD leader Pierre Nkurunziza sign a final agreement to end hostilities at summit of African leaders in Tanzania. Nkurunziza and other FDD members are given ministerial posts, The smaller Hutu rebel group, the Forces for National Liberation (FNL), continues to launch attacks in Bujumbura Rurale and Bujumbura suburbs. Thirty-seven military commanders from the Burundian army and all armed groups except for the FNL participate in a BLTP-led six-day Nairobi workshop in collaborative decision-making, as preparation for implementation of the cease-fire agreement, DDR, and unification of the army. 2004 February - A six-day BLTP workshop in collaborative decision-making is held for the members of the Joint Ceasefire Commission. 2004 May: The largest former rebel group, the Forces for the Defence of Democracy (FDD), temporarily withdraws from participation in Cabinet meetings and in the National Assembly, claiming that the Government had failed to implement certain promises. FDD leader Hussein Rajabu says the decision does not mean a return to war. The members of the Etat-Major Général Integré (EMGI), comprised of members of the FAB and of the CNDD-FDD (Nkurunziza), together with representatives of the five smaller armed groups, participate in a six-day BLTP-led retreat, focusing on collaborative decision- making. This workshop leads to the EMGI's resolution of one of the most sensitive issues blocking the launch of the DRR Program: the definition of "combatant." 2004 August: FNL claims responsibility for a massacre of over 150 Congolese refugees at Gatumba. 26 2004 September: National Assembly and Senate, acting jointly, approve draft constitution to be presented for public referendum. 2004 September: UPRONA leaders express opposition to the draft constitution, announce intention to challenge its legality in the Constitutional Court. 2004 October: The President removes the draft constitution from the consideration of the Constitutional Court, on the grounds that the Court had not acted within the stipulated 15 day review period. 2004 October: Regional Summit in Nairobi endorses the Electoral Commission's recommendation that the scheduled elections be delayed for technical reasons, calls for a six-month extension of the Transitional period, and affirms that the recently approved draft constitution should serve as Burundi's new interim constitution pending its ratification by public referendum. The Burundian Parliament subsequently enacts legislation to implement these Summit recommendations. 2004 October: Several Tutsi parties drop their opposition to the country's interim constitution, affirm their intention to remain within the transitional institutions, but call for continued negotiations on prospective power-sharing arrangements. 2004 November: Leaders of 31 of Burundi's 33 political parties convene for six-day BLTP workshop. Agree on recommendations to be made to Independent Electoral Commission concerning the Electoral Code of Conduct. Request further training and opportunities to work collaboratively. 2004 December: DRR Program formally launched. 2004 December: President Ndayizeye dismisses Vice-President Kadege without incident, appoints Frederic Ngenzebuhorg, an UPRONA-endorsed candidate, to succeed him. Scheduled referendum on interim constitution is further delayed for technical reasons. 2004 December: Burundian political party leaders participate in second BLTP workshop on collaborative decision-making, issue joint communiqué to assure the public of their collective determination to work to ensure that the elections are free of violence and intimidation. 2004 December: President indicates interest in possible amendments to the interim constitution that would permit direct election of the President and would remove the prohibitions on either of the transitional Presidents from standing for the post-transition presidency. He is opposed by key members of his own FRODEBU party, by the CNDD-FDD, and by many of the smaller Hutu and Tutsi political parties. 2005 January: The Electoral Commission sets February 28, 2005 as the date for the constitutional referendum. 2005 January: A coalition of 20 political parties, half Tutsi and half Hutu, issue joint statement opposing amendments to the interim constitution prior to the public referendum. Facilitator Zuma visits Bujumbura, reaffirms region's view that the negotiated Arusha and post-Arusha agreements must be respected. 2005 February: President announces that there will be no change in the constitutional provisions that had generated controversy, and calls for quick action on the Electoral Law. 2005 February: Inaugural BLTP workshop on leadership and collaborative decision-making is held for the high command of the new national police force. 2005 February 28: Referendum on the new constitution is held, with over 90% of Burundi's 3 million registered voters casting "yes" votes. 2005 March 11: National Assembly approves new Electoral Law. 2005 April: The Tanzanian government secures an agreement from the FNL to cease fighting and enter unconditional negotiations with the Burundian government. 27 Bibliography Anderson, Mary B. 1999. Do No Harm. How Aid Can Support Peace or War. First ed. London: Lynne Rienner Publishers. Barron, Patrick, Rachel Diprose, David Madden, Claire Smith and Michael Woolcock. 2003. "Do Participatory Development Projects Help Villagers Manage Conflicts?" CPR Working Paper No. 9. Washington D.C.: World Bank. Beleli, Ozsel. 2001. Aid as a Peacemaker: A View from Burundi. Bowen, Michael and Kay Miller. 1973. "No Samaritan: The U.S. and Burundi." Africa Report 18: 32-39. Bowen, Michael, Gary Freeman, and Kay Miller. 1972. Passing By: The United States and Genocide in Burundi. New York: Carnegie Endowment for International Peace. Brachet, Juana and Maude Svensson. 2004. Poverty Reduction Strategies in Conflict-affected Countries: Burundi, July (internal World Bank paper). Cullen, Michelle. 2003. Exclusion and Development in Conflict-Affected Countries, September (unpublished). Chrétien, Jean-Pierre. 1997. Le Défi de l'ethnisme. Rwanda et Burundi: 1990-1996. Paris: Karthala. ------. 1993. Burundi: l'Histoire retrouvée. Paris: Karthala. ------. 1994. "Burundi: pogromes sur les collines." Esprit (July) Cliffe, Sarah, Scott Guggenheim and Markus Kostner. 2003. "Community-Driven Reconstruction as an Instrument in War-to-Peace Transitions." CPR Working Paper No. 7. Washington D.C.: World Bank. Colletta, Nat and Michelle Cullen. 2000. Violent Conflict and the Transformation of Social Capital. Lessons from Cambodia, Rwanda, Guatemala and Somalia. Washington, D.C.: World Bank. ------. 2000. The Nexus between Violent Conflict, Social Capital and Social Cohesion: Case Studies from Cambodia, and Rwanda. Washington, D.C.: World Bank. Colletta, Nat and Taies Nezam. 1999. From Reconstruction to Reconciliation. Washington, D.C.: World Bank. Collier, Paul. 1999. "On the Economic Consequences of War". Oxford Economic Papers, Vol. 51, pp.168-83. ------. 2003. "Aid, Policy and Peace: Reducing the Risk of Civil Conflict." CPR Dissemination Notes No.9. Washington, D.C.: World Bank. Collier, Paul, Lani Elliot, Håvard Hegre, Anke Hoeffler, Marta Reynal-Querol, and Nicholas Sambanis. 2003. Breaking the Conflict Trap. Civil War and Development Policy. Washington D.C.: World Bank 28 Commission Nationale, Burundi. 1989. Rapport de la Commission Nationale chargée d'étudier la question de l'unité nationale. Bujumbura. DFID. 2004. Conducting Conflict Assessments: Guidance Notes. London: DFID. Edlin, John and Colin Legum. 1994. "Burundi: The Land that Lost Its Head." The New African, February. Gahama, Joseph. 1983. Le Burundi sous administration belge. Paris: Karthala. ------. 2002. Conflict Prevention, Management and Resolution in Burundi. DPMF. Addis Ababa: Development Policy Management Forum. Grootaert, Christiaan. 1998. "Social Capital: The Missing Link?" Washington, D.C.: World Bank. Grootaert, Christiaan and Thierry van Bastelaer. 2001. Understanding and Measuring Social Capital: A Synthesis of Findings and Recommendation from the Social Capital Initiative. Washington, D.C.: World Bank. Hakizimana, Déo. 1991. Burundi: le non-dit. Vernier: Éditions Remesha. Harroy, Jean-Paul. 1987. Burundi (1955-1962). Brussels: Hayez. Human Rights Watch/Fédération Internationale des Droits de l'Homme. 2003. Rapport de la Commission Internationale d'Enquête sur les Violations des Droits de l'Homme au Burundi depuis le 23 Octobre 1993. Brussels. July 1994 International Crisis Group. 2003. A Framework for Responsible Aid to Burundi. Africa Report N°57, 21 February. International Crisis Group. 2004. End of the Transition in Burundi: The Home Stretch. Africa Report Nº81, 05 July. ------. 2004. Elections in Burundi: The Peace Wager. Africa Report NºXX, 09 December. Lemarchand, René. 1970. Rwanda and Burundi New York: Praeger. ------. 1992."Burundi: The Politics of Ethnic Amnesia." Genocide Watch. Helen Fein, ed. New Haven: Yale University Press, pp. 70-86. ------. 1996. Burundi: Ethnic Conflict and Genocide. Washington, DC: Woodrow Wilson Center Press. Lemarchand, René and David Martin. 1974. Selective Genocide in Burundi. Report No. 20. London: Minority Rights Group. Meisler, Stanley. 1976. "Holocaust in Burundi, 1972." Case Studies on Human Rights and Fundamental Freedoms: A World Survey, vol. 5, ed. Willem A. Veenhoven. The Hague: Martinus Nijhoff. Muscat, Robert. 2002. Investing in Peace: How Development Aid Can Prevent or Promote Conflict. ME Sharpe. 29 Ndabakwaje, Libère and Jean-Baptiste Mbonyingingo. 2003. Cadre d'analyse du conflit (internal World Bank paper). Nkurunziza, Janvier. 2001. "Economic Determinants of Violent Conflict in Burundi." Bonn: Deutsche Stiftung fur Internationale Entwicklung. Nkurunziza, Janvier and Floribert Ngaruko. 2002. Explaining Growth in Burundi: 1960-2000. Centre for the Study of African Economies, Oxford. Pabanel, Jean-Pierre. 1988. `Statistiques tribales au Burundi en 1986.' Politique Africaine. No. 43, December. Percival, Valerie and Thomas Homer-Dixon. 1995. Environmental Scarcity and Violent Conflict: The Case of Rwanda. Washington, D.C.: American Association for the Advancement of Science, University of Toronto. Prunier, Gérard. 1994. "Burundi: A Manageable Crisis? " United Kingdom: Writenet. Reyntjens, Filip. 1994. L'Afrique des Grands Lacs en crise. Paris: Karthala. ------. 1995. Burundi: Breaking the Cycle of Violence. London: Minority Rights Group. ------. 2000. Burundi: Prospects for Peace. London: Minority Groups International. Stedman, Stephen. 1997. "Spoiler Problems in Peace Processes." International Security, v. 22, #2. Swedish International Development Cooperation Agency. 2004. A Strategic Conflict Analysis for the Great Lakes Region. Stockholm: SIDA. Uvin, Peter. 1996. Development, Aid and Conflict: Reflections from the Case of Rwanda. United Nations University World Institute for Development Economics Research--Research For Action. Watson, Catherine. 1993. Transition in Burundi: The Context for a Homecoming. Washington, D.C.: U.S. Committee for Refugees. Weissman, Stephen. 1998. Preventing Genocide in Burundi: Lessons from International Diplomacy Washington, D.C.: United States Institute of Peace. Weinstein, Warren and Robert Schrire. 1976. Political Conflict and Ethnic Strategies: A Case Study of Burundi. Syracuse: Syracuse University Press. Wolpe, Howard. 2003. "Burundi: Facilitation in a Regionally Sponsored Peace Process." Unpublished manuscript prepared for a United States Institute of Peace conference on best facilitation practices. Wolpe, Howard et al. 2004. "Rebuilding Peace and State Capacity in War-torn Burundi." The Roundtable, Vol. 93, No. 375, 457-467. World Bank. 2002. Greater Great Lakes Regional Strategy for Demobilization and Reintegration. Report No. 23869-AFR. March. 30 ------. 1998 Operations Evaluation Department. World Bank's Experience with Post-Conflict Reconstruction. Washington, D.C.: World Bank. ------. 2002. Holst Peace Fund ­ Supporting Development in the West Bank and Gaza. ------. 2002. West Bank and Gaza: An Evaluation of Bank Assistance. WB Country Case Studies: ------. 2000. Bosnia and Herzegovina: Post-Conflict Reconstruction, (A. Kreimer, R. Muscat, A. Elwan and M. Arnold). ------. 2000. El Salvador: Post-Conflict Reconstruction, (J. Eriksson, A. Kreimer and M. Arnold). ------. 2000.Uganda: Post-Conflict Reconstruction, (A. Kreimer, P. Collier, C. Scott and M. Arnold). Country Assistance Evaluations: ------. 1997. Mozambique (Report No.17209, December 1997). ------. 2000. Cambodia (Report No. SecM2000-708, December 2000). ------. 2001. Sri Lanka (Report No. 21771, January 2001). ------. 2002. Guatemala (Report No. 25212-GT, December 2002). ------. 2002. Haiti (Report No.23637, February 2002). CONFLICT PREVENTION AND RECONSTRUCTION UNIT--WORKING PAPER SERIES No. Title Authors Date 12 Social Capital and Survival: Prospects for Community-Driven Paul Richards April 2004 Development in Post-Conflict Sierra Leone Khadija Bah James Vincent 13 MIGA's Experience in Conflict-Affected Countries: The Case John Bray June 2004 of Bosnia and Herzegovina 14 The Devil in the Demographics: The Effect of Youth Bulges Henrik Urdal July 2004 on Domestic Armed Conflict, 1950-2000 15 Practical Guide to Multilateral Needs Assessments in Post- Uwe Kievelitz August 2004 Conflict Situations Et al. 16 The Private Sector's Role in the Provision of Infrastructure in Jordan Schwartz August 2004 Post-Conflict Countries: Patterns and Policy Options Shelly Hahn Ian Bannon 17 Within and Beyond Borders: An Independent Review of Post- Swarna Rajagopalan October 2004 Conflict Fund Support to Refugees and the Internally Displaced 18 Drugs and Development in Afghanistan William Byrd December 2004 Christopher Ward 19 Understanding Local Level Conflict Pathways in Developing Patrick Barron December 2004 Countries: Theory, Evidence and Implications from Indonesia Claire Q. Smith Michael Woolcock 20 Addressing Gender in Conflict and Post-Conflict Situations in Sonia Margallo January 2005 the Philippines 21 Community Cohesion in Liberia: A Post-War Rapid Social Paul Richards, et al. January 2005 Assessment 22 International Companies and Post-Conflict Reconstruction: John Bray February 2005 Cross-Sectoral Comparisons 23 The Roots of Violence and Prospects for Reconciliation: A Claire Q. Smith February 2005 Case Study of Ethnic Conflict in Central Kalimantan, Indonesia 24 The Mindanao Conflict in the Philippines: Roots, Costs, and Salvatore Schiavo- February 2005 Potential Peace Dividend Campo Mary Judd 25 Counting Conflicts: Using Newspaper Reports to Understand Patrick Barron May 2005 Violence in Indonesia Joanne Sharpe 26 Young Men and the Construction of Masculinity in Sub- Gary Barker June 2005 Saharan Africa: Implications for HIV/AIDS, Conflict and Christine Ricardo Violence 27 Conflict-Sensitive Development Assistance: The Case of Juana Brachet June 2005 Burundi Howard Wolpe Also published as Social Development Paper No. 87