Greater Horn of Africa Peace Building Project Draft Report Evaluation of USAID Programs in Burundi Search for Common Ground July & August 2004 Management Systems International 600 Water Street, SW, Washington, DC 20024 USA Contributors: Emery Brusset, MSI; Fabienne Hara, MSI Editors: Stephen Schwenke, MSI: Michael Lund, MSI EVALUATION OF USAID PROGRAMS IN BURUNDI ii TABLE OF CONTENTS EXECUTIVESUMMARY .......................................................................................................................................................................................................III ABBREVIATIONS ANDACRONYMS .................................................................................................................................................................................IX MAP OFBURUNDI................................................................................................................................................................................................................XI 1 BACKGROUND.........................................................................................................................................................................................................12 1.1 Evaluation Approach and Methodology ..........................................................12 1.2 Country Background ........................................................................................13 1.3 US Integrated Strategic Plan ...........................................................................18 2 BRIEF PROGRAM DESCRIPTION.......................................................................................................................................................................19 3 DESIGNAND PREPARATION: SEARCH FORCOMMONGROUND..............................................................................................................21 4 MANAGEMENT ANDIMPLEMENTATION: SEARCH FORCOMMONGROUND.........................................................................................25 5 IMPACTOFSEARCH FOR COMMONGROUND..............................................................................................................................................27 5.1 Providing Attractive Models That Capture the Imagination...........................27 5.2 New Forms of Interaction.................................................................................29 5.3 Capacities Created ...........................................................................................29 6 FINDINGS AND RECOMMENDATIONS FOR SEARCH................................................................................................................................31 ANNEXES..............................................................................................................................................................................................................................33 Annex I: List of Persons Met......................................................................................34 EVALUATION OF USAID PROGRAMS IN BURUNDI iii EXECUTIVE SUMMARY Introduction The evaluation was commissioned by USAID/REDSO in Nairobi to support a reassessment of programming priorities, in light of upcoming elections in Burundi, as planned in the Arusha Peace and Reconciliation Accord. For this field visit, Management Systems International (MSI) engaged the services of two consultants, Ms Fabienne Hara and Mr Emery Brusset, between July 7 and 20, 2004. They covered the work of the two organizations funded by REDSO in the field of conflict mitigation and prevention: Search for Common Ground, and Global Rights. The analysis was divided into two separate reports, each covering one of these institutions, although this Executive Summary is identical for both, and refers to both. Debriefings were held in Bujumbura and Nairobi, and the draft report presented on August 16, 2004. The findings are listed below as observations, following the three-tier structure foreseen in the Scope of Work. Observation N° 1: Design The evaluation finds that there is a small but significant mismatch between the objectives of the organizations. The USAID Intermediary Results (IRs) are focused on the institutional aspects of the political transition. Search for Common Ground’s program objectives date back to 1995, and give evidence of a gradual evolution over the years. This organization’s programs address a wide understanding of conflict, including the broad spectrum of societal vulnerabilities to conflict, as well as the risks of future conflict. The objectives of Global Rights are broader in the sense that they deal with NGO capacity, as well as forms of disputes common to and originated by the population. USAID’s and its partners’ objectives refer to different processes and timeframes. In recent months Search for Common Ground has gradually shifted its focus towards a community approach, broadly defined as being more attuned to specific cases of tension within the population. Global Rights seeks to phase out its capacity building program after 5 years. This evaluation report uses the NGOs’ stated objectives as its principal points of reference. Recommendation 1: • The IRs should be expanded or clarified by USAID so as to include the societal dimensions of conflict in the country, as these are key to the conflict (particularly because of the elections, and importance of land conflicts), and better reflect the programs funded. EVALUATION OF USAID PROGRAMS IN BURUNDI iv Observation N°2: Design In the design of objectives and weighting of activities (either written – as found in the agency literature, or oral – as very cogently communicated by the program directors) the two organizations do not make reference to any foundational analysis defining the drivers of the conflicts. The conceptual grounding that underpins their respective interventions therefore is not entirely clear. The objectives that the MSI Team discerned within the two organizations were not always clear. These objectives are frequently stated separately from one another, or are applied in a confusingly cross-cutting manner in some programming documents. As perceived by the MSI Team, there are two foundations common to both of the programs evaluated: (1) Expertise in certain conflict management and arbitration techniques, and (2) A wealth of personal contacts, rather than a deeper analysis of what causes the most risk of conflict. This leads to a reduced overall efficiency. It is not possible for the organizations to weigh the merits of the different components in terms other than continuity or non￾continuity with existing structures. Recommendation 2: • Both organizations should engage in their own long-range conflict assessments, leading to a priority ranking of different areas of interest and some indicators of success. • The triggers which the MSI Team of evaluators found for the peace process are: 1) demobilisation, disarmament and reintegration (DDR), 2) the repatriation of refugees, and 3) the poor quality of information about elections and the economy. All of these are areas where there will be significant and critical changes in the short term. Programs should address these issues more specifically. In particular Search for Common Ground should target the future demobilised soldiers and help them get information on DDR packages, as well as vocational training. Observation N°3: Management & Implementation There should be clear linkages between the stated objectives of each of the two organizations to their corresponding activities, and then on to the implementing structures of each. In the case of Global Rights (GR), these linkages are not clear. The GR program is divided into two: one stream addresses the grey zone between state law and customary law, while the other stream deals with facilitation of constitutional and legal issues in the Arusha Agreement. There is also evidence of a continual effort by GR to identify and strengthen local NGOs, but this has not yielded evidence of effective ongoing relationships. Search for Common Ground’s programs are structured in four components: 1) dealing with the promotion of a more benign/active media, 2) support to women in the communities, 3) assistance to victims of violence and rape, and 4) dialogue among high conflict risk youth groups. EVALUATION OF USAID PROGRAMS IN BURUNDI v The operational links between the different components of Search for Common Ground’s program are robust and effective at all levels, although it is not clear whether this is the result of the integration initiative. However these four components crowd out one another in the field, as the NGO’s logistical capacity is very limited for an outreach program (e.g. one vehicle per field office). Global Rights is run as two separate components, one centred in the communities in the field (where it is directly operational) and one addressed at legislative lobbying with other NGOs. The relationships of the two main organizations to their partners are heavily influenced by previous commitments. In some cases, the MSI Team noted that there was a sense of drift, as in GR’s program where legal clinics were phased out of certain areas with no clear rationale for doing so. The partners of the two NGOs include paralegal advisers (in the case of GR), a decreasing number of associations, and a divisive community of national civil society organizations. In the case of Search for Common Ground, this list of partners includes radio stations, and in particular Radio Isanganiro. Recommendation 3: • For each of these two NGOs, there should be a new planning workshop in which the coordination between components is clarified, based on a clearly articulated conflict assessment, and based on an analysis of the impact achieved (or on a persuasive argument describing the potential for impact). • The logistical capacity of both NGOs should be expanded with USAID funding support, to maintain and improve the efficiency of the programs. • There should be a better conceptualization of the links between objectives and partnerships, and a definition of the relevance of one to the other. The programs should, in this process, define partnerships and lasting institutional survival as one of the impacts of the programs. Observation N°4: Management & Implementation The political nature of the work and the highly polarised nature of the society have complicated the human resource management processes of Search for Common Ground, and to a lesser extent of GR. In both organizations, there is considerable uncertainty in some quarters concerning the criteria used in staff selection and promotion. The ratio of field staff to headquarters staff is good. Recommendation 4: • Both organizations, but particularly Search for Common Ground, should formalise and explain the human resource management structures that they operate on. The priority must be given to avoiding situations where constructive criticism and otherwise effective management interventions fall prey to disagreements and misunderstandings due to the lack of clearly articulated personnel policies and defined staff management principles. Such a policy framework is not easy to develop, as it must address concerns of equitable ethnic representation as well as merit based recruitment and advancement, but despite such challenges a clear framework is needed. EVALUATION OF USAID PROGRAMS IN BURUNDI vi Observation N° 5: Management Both organizations have managed to achieve an impressive continuity of presence, high rates of staff retention, and a high profile and significant relations with partners, the society, and the state in Burundi. This is less true for GR, which is still not connected to major civil society initiatives or lobbying activities related to the lawmaking process, even if it has a strong profile in the Provinces where it operates and attracts high profile staff. Overall the programs are based on the twin assets of 1) robust partnerships with institutions and people (such as facilitators, or partners) and 2) the application of conflict mitigation processes (enhancing dialogue, for example). This means that the content of the messages and types of conflicts to be addressed can be adjusted rapidly. Recommendation 5: • In proposing new program orientations and objectives, USAID would do well to avoid disturbing what appears to be an effective balance in the objectives and operations of both NGOs evaluated. Instead, USAID should take advantage of the relative lack of programmatic rigidity in both NGOs, allowing their respective programs to adjust in an evolutionary manner to changing conditions in Burundi. Observation N°6: Impact (Search for Common Ground) The media programs have had a significant positive impact in mitigating conflict, because of the quality of design that enabled Search for Common Ground (i.e. Studio Ijambo and Radio Isanganiro) to catch and hold the attention of the majority of the radio-listening population. Their productions have led to a new interaction between media and the political class over the period 2001-2004, and a universally recognized improvement in quality of reporting in the country as a whole. There has been little attention on the part of Search for Common Ground directed at the capacities it has created (both in terms of individual journalists and also radio stations), other than Radio Isanganiro. Recommendation 6: • This issue of capacity created needs in fact to be reviewed from a variety of perspectives, in particular sustainability, protection and guarantees of political immunity to manipulation (for example, advice in human resource management), and recommended actions to follow in cases of intimidation. Search for Common Ground’s Youth and Women’s components are much more localized (represented by a few communes and sharply defined groups) in their impact than the media interventions, but are nevertheless very deep. This notable depth is due to the relevance to the population of the issues covered, the problematic nature of information on the peace implementation (particularly elections), and the need to build new capacities to respond to conflicts as the society moves away from the traditional modes of mediation and arbitration. EVALUATION OF USAID PROGRAMS IN BURUNDI vii • USAID through Search for Common Ground should seek to bring more funding to the youth and women’s activities to extend their geographic scope. Although the impact of VOT was not measured, the past evaluation suggests it is responding to conflict rather than preventing it, and the MSI Team would recommend it be handed over to another service organization operating in the health sector. A review of the importance of Search for Common Ground’s Victims of Torture (VOT) activity should be carried out, and an assessment made of the degree to which the Search-VOT activity currently constrains the performance of the other programs. The conceptual link between conflict mitigation and assistance to victims of rape and torture is tenuous. Observation N° 7: Impact (Global Rights) GR has not been able to achieve the desired impact in the activity covered by the evaluation (support to the Truth and Reconciliation Commission) due to the weakness of the context that it seeks to influence (Parliament and civil society) and the necessity for the Burundian political parties to conclude political negotiations on the issue of impunity. In the paralegal assistance field it has a very effective but geographically decreasing impact: it is currently only present in two Communes, even though it has been offered funding by UNHCR in Muyinga an Kirundo Provinces. Recommendation 7: • USAID and GR should revisit the provision legal clinics as an original, low cost and effective form of influence on one of the main long-term causes of conflict in the society. It should revamp past projects and launch new ones. • In particular, an assessment of the potential role of the legal clinics to help solve land conflicts related to refugee repatriation and DDR process should be conducted. Also, an assessment of the state of civil society would be helpful. • GR needs to assess risk of duplication with the Bashingantahe institution, to clarify relationship with local administration, local judiciary and CNRS. Some of the Provinces could be funded through other donors. Both organizations have invested in monitoring of impact, particularly Search for Common Ground. This monitoring and evaluation has helped promote the discussion about objectives. Few evaluations carried out, however, include impact data, and surveys and information generated by others are still not well used (e.g. research done by OTI, and evaluation of the VOT program). • This burgeoning process of impact monitoring should be promoted further, and funded by USAID. The work carried out has led to a great degree of voluntary commitment and risk taking. • More attention needs to be paid to security issues, e.g. the risks accruing to Search for Common Ground trained “messengers of peace”. The activities of both organizations also lead to a high exposure to security risks for those who cooperate with them. This problem raises issues of duty of care, which needs to be thought about more carefully than is currently done (maybe by EVALUATION OF USAID PROGRAMS IN BURUNDI viii generating more visits to show that there is a commitment to the work and information flows). Observation N°8: Adjustments as Regards Elections It is possible for both organizations to engage in a more targeted program dealing with issues pertaining to the elections. These adjustments could easily capitalize on the existing assets, but there is a concurrent risk of weakening existing programmatic vitality and effectiveness, which would have an impact on continuity and efficiency. The VOT model has not been the most constructive in this sense. However both organizations remain well placed to deal with other conflict triggers Recommendation 8: • The GR program could be used to address in a deeper manner the issue of land, which, even if there is repatriation, will be a source of unease; • Search for Common Ground, but particularly GR, could support civil society networks and frame their actions to allow for an effective communication plan, particularly rumor management and the trickle-up of good information on the campaign in the hills • Agree with civil society on a “code of conduct” and an acceptable “campaign language” and start monitoring the pre-campaign in a systematic way as of now; • Search for Common Ground could contribute in the training of local election observers through its youth and women’s programs (see Kenya and Zimbabwe experience). This contribution should not lead, however, towards becoming an electoral assistance program (which should be supported instead by an independent organization, preferably from the region). EVALUATION OF USAID PROGRAMS IN BURUNDI ix ABBREVIATIONS AND ACRONYMS AIDS Acquired Immune Deficiency Syndrome APDH Association pour la paix et les Droits de l’Homme AU African Union BBC British Broadcasting Corporation CARE Cooperative for Assistance and Relief Everywhere CMC The Ceasefire Monitoring Commission CMM Conflict Mitigation and Management CNDD-FDD Centre pour le Défense de la démocratie-Forces pour la Défense de la Démocratie CNRS Commission Nationale pour le Réinsertion des Sinistrés CRS Catholic Relief Service DDR Demobilisation, Disarmament and Reintegration DDRRR Demobilisation, Disarmament, Rehabilitation and Reintegration DFID Department for International Development FAB Forces Armées Burundaises FDD Forces pour la Défense de la Démocratie FNL See PALIPEHUTU-FNL FRODEBU Front pour la Démocratie au Burundi G10 Group of 10 Tutsi parties participating in the Arusha process G7 Group of 7 Hutu parties participating in the Arusha process GR Global Rights (formerly known as International Law Group) ICRC International Committee of the Red Cross IR Intermediary Result IR 1 Intermediary Result- Peace Process and Transition Institutions Strengthened IR2 Intermediary Result – Increased Participation of Civil Society MSI Management Systems International NGO Non-Governmental Organisation OFDA Office for Foreign Disaster Assistance OTI Office for Transition Initiatives PALIPEHUTU- Parti pour la Libération du Peuple Hutu-Forces EVALUATION OF USAID PROGRAMS IN BURUNDI x FNL Nationales de Forces Nationales de Libération POC Joint operational plan for disarmament and demobilization PRM Population, Refugees and Migration PSI Population Services International REDSO Regional Economic Development Services Office RFI Radio France Internationale RPA Radio Publique Africaine RTNB Radio Télévision Nationale du Burundi Search Search for Common Ground SFCG Search for Common Ground SI Studio Ijambo SIDA Swedish International Development Cooperation Agency SRSG Special Representative of the Secretary General SWAA Society of Women of Africa against AIDS UN United Nations UNICEF United Nations Children's Fund UPRONA Union pour le Progrès National US United States (of America) USAID US Agency for International Development USD United States Dollars USG US Government VoA Voice of America VOT Victims of Torture (also seen as VoT) WFP World Food Programme WHO World Health Organisation EVALUATION OF USAID PROGRAMS IN BURUNDI xi MAP OF BURUNDI EVALUATION OF USAID PROGRAMS IN BURUNDI 12 1 BACKGROUND 1.1 Evaluation Approach and Methodology Because of US Embassy travel restrictions, and the short time in country, field visits only took place to Ngozi province (two communes of Ngozi and Nyamurenza). Both Search for Common ground and Global Rights have field offices (‘antennas’) in Ngozi, and these offices facilitated the MSI team’s visits to the communities in the hills, including transport and translation. The MSI team was grateful for this friendly assistance from both organizations and their field staff, and for their complete openness to the evaluation process. The visits included semi-structured interviews and direct observation, particularly of the legal clinics. Interviews were also carried out in Bujumbura with Barundi politicians, journalists, other agencies, and staff from the two NGOs. The list of persons met is included in Annex I. The results framework provided the points of reference on which the evaluation was to be based. These were drawn from the USAID and NGO documents, and when found of insufficient content, from interviews of staff. These results were analyzed at three levels, as provided for in the Scope of Work: 1) planning and design of the interventions, 2) management and implementation of the activities, and 3) impact achieved in terms of changes in the broader context (beyond the outputs of the programs). The design analysis segment was carried out primarily on the basis of documents given, supplemented with interviews with key personnel. Following the Scope of Work, reference was made to conflict analysis. The evaluation also refers to implicit theories of conflict and non-formal goals. The management analysis section was carried out by semi-structured interviews with a number of staff (approximately 10 individual interviews for each organization and three group interviews). The question of impact is considered very important, and the evaluation allocated a larger amount of time to it, mainly dedicated to semi-structured interviews and review of the relevant reports, seeking out specific indicators. The selection of the indicators was based on a methodology laid out and discussed with the client during the inception phase. For IR 1 (peace process and transition institutions) the evaluation followed one action by GR from initiation to its phasing out, rather than evaluating all activities. The evaluation proceeded by the identification of possible waste or on the contrary multiplier effects; gap analysis between the intended and achieved; constraints and how they were addressed; and an analysis of the quality and relevance of outcomes. For IR2 (civil society) the evaluation proceeded by splitting the interviews into four groups of population: 1) the political class, including some high level Burundi officials, and donors other than Search and GR; 2) farmers and persons who have EVALUATION OF USAID PROGRAMS IN BURUNDI 13 benefited from the community activities of both organizations; 3) staff from all organizations, which can be called “partners” by the fact that they implement similar activities; and 4) staff from both NGOs. Due to evaluation time constraints, the impact assessment relied on existing surveys in both organizations, and some surveys carried out by other organizations (for example baseline surveys by OTI) in related fields. The indicators for IR 2 required some degree of elaboration by the evaluators. An objective should be explicit, that is, it must clearly show why an intervention will be judged better or worse in meeting the objective. To be used in an evaluation, an objective should indicate a form or level of success at which an intervention will be considered good in terms of reaching this objective. As this quality was not found in the documents, we proposed to use three forms of implicit objectives to track direct contributions, drawn from language, which has been recurring in the preliminary interviews. These can be defined as: 1. To catch people’s imagination (short term impact): Has there been a propagation of new models of social interaction, which are attractive and are reproduced and emulated further within the society? What are these models, how attractive are they? Were they proposed at the right time, or allowed the organization to buy time while contradictory messages of incitation to violence and tension were being spread? 2. To create new modes of interaction (medium term impact): Have the activities allowed groups which did not previously have contact to talk to each other, either through political representatives, community to politicians, or between communities? Have these contacts been of a new nature? Were they noticed as important by a significant number of people? 3. Capacities (long term impact): What institutions have been created, for example new radio stations? What personnel have been trained and what skills have been improved that will allow the society to respond to any new upsurge in violence? How sustainable are these new capacities? 1.2 Country Background The Scope of Work asked that the evaluation begin with an analysis of “the evolving drivers and manifestations of the waning conflict, as well as the peace building processes in Burundi”. MSI here has focused on the events and trends that characterize current conditions in Burundi, rather than adopting a more historical perspective. In the last ten years, the Burundi context has been characterized by two dynamic processes that have run in parallel: the civil war that started in 1993 with the assassination of the first democratically elected Hutu president Melchior Ndadaye, and the peace process, which began officially in Arusha under the auspices of Julius Nyerere in June 1998. The Arusha process, which came after the failure of several internally negotiated power sharing agreements, has so far lasted six years. It has been negotiated in stages (Arusha agreement in 2000, agreement on the transitional arrangement in 2001, the ceasefire with Conseil national pour la défense de la démocratie – Forces de défense de la démocratie in 2003), and by several mediators.1 1 UNSRSG Ahmedou Ould Abdallah, Carter Center, Community of Sant Egidio, Julius Nyerere, the Regional Initiative on Burundi (chaired by President Musevenu), Nelson Mandela, Jacob Zuma. EVALUATION OF USAID PROGRAMS IN BURUNDI 14 The former AU – now UN – peacekeeping mission is providing the guaranty for the implementation of these agreements. The length of the process can be explained by the fact that the formal peace process was never fully inclusive of all parties, allowing them to “talk, and fight strategy” in order to gain a position of superiority. These constant re-alignments and the rebels’ late accession to the talks have given rise to internal divisions within the parties, alliances and counter-alliances. Burundi has hung between war and peace since 2000 as the Arusha agreement's implementation has become an endless cycle of re￾negotiations on the concerns of various parties. The current talks on the electoral system and post elections constitution, as well as the current attempts to include the FNL in the talks, are a good indication that the process has not yet been completed. To date, none of the fundamental reforms addressing the root causes of the conflict and included in the Arusha document have been implemented. In this context, the ‘peace process’ can be defined in two ways: either strictly as the negotiation and implementation of the Arusha agreement, or broadly as the framework that has gradually formalized the different stages of conflict transformation. Formal and informal achievements of the peace process The first principal result of the talks, the Arusha agreement, is a thorough and balanced reflection of the 20 parties’ wishes for political representation and access to political power, and is based on a series of unanimous resolutions. The three completed protocols (I, II, IV) establish a clear and ambitious program of action aimed at advancing the cause of reconciliation, democracy and reconstruction in Burundi. Protocol I, which focused on the nature of the conflict, includes the establishment of two justice mechanisms: an International Judicial Commission of Inquiry and a Truth and Reconciliation Commission. Protocol II, which dealt with democracy and good governance, lists power sharing principles and outlines a complex election system that would protect minority rights; Protocol IV of the Arusha Agreement provides a roadmap for economic aspects of the post-conflict period. The two major sub-agreements missing in Arusha, a transitional power sharing agreement (an item of Protocol II) and a cease-fire and security sector reform program (Protocol III), were reached much later. In July 2001, the two main political parties, FRODEBU and UPRONA, agreed on a three year transition period with a rotating presidency, which led to the return of exiled FRODEBU politicians to Burundi. The composition of the National Assembly was modified and the Senate was set up to include more members from the various coalitions of political parties (called G7 for the Hutu and G10 for the Tutsi parties). On November 16, 2003, the transitional government led by President Buyoya signed a landmark ceasefire agreement with the party of Jean-Pierre Nkurunziza. This complemented the ceasefire reached earlier in 2002 with two minor rebel groups (the CNDD-FDD faction led by Jean-Bosco Ndayikengurukiye and the PALIPEHUTU￾FNL faction led by Alain Mugabarabona). This agreement anticipates full integration of the current Burundi army and the FDD. Following the signing, an African Union EVALUATION OF USAID PROGRAMS IN BURUNDI 15 force with South African, Ethiopian and Mozambican troops was deployed in the spring 2003, and replaced a year later by a UN peacekeeping force, authorized on June 1, 2004. Role of media in the process The inclusion of the rebels in the process would not have been possible without the active participation of the media in broadening the political debate after the signing of the Arusha agreement in 2000. For the first two years of the Arusha process, the media failed to fulfil its duty to inform the people about the state of negotiations. The Buyoya government argued that restrictions were necessary to avoid ethnic mobilization and strong reactions from certain constituencies. In particular, it accused FRODEBU of civil disobedience and of mobilizing the Hutu population against it. In fact the media were being used as a pretext to control the information that people received about the talks. Control of the media has indeed been a deeply entrenched obsession shared by all Burundian politicians. In 1996 suspension of the freedom of press followed a three￾year period during which the media of various parties had encouraged violence by spreading messages of ethnic hatred. When Nelson Mandela , in his capacity as Mediator, sought to include the rebels in the peace process, he asked President Buyoya to comply with rebel pre-conditions. These pre-conditions included the release of all political prisoners regardless of their crimes; the restoration of the rights of political parties; and the closing of military population ‘regroupment’ camps. Mandela also demanded that freedom of the press be restored. He thought that a radical change of policy towards the press was required in order to prepare the people for a peace agreement and to pave the way for the return of exiled Hutu politicians. The media had to start supporting the process of national reconciliation and reconstruction of the country through a debate that would be as broad as possible. Gradually, radio was made accessible to all parties involved in the conflict. The FDD and the FNL appointed spokespersons started being interviewed by Burundian journalists by mid 2001. The content of the Arusha agreement was also finally publicized though the media and the actions of some NGOs. Today all private radio stations in Burundi try to give a fair share of time to politicians from all sides in their broadcasted debates. Ceasefire and DDRR Since the signing of the comprehensive ceasefire agreement between the transitional government and CNDD-FDD rebels headed by Jean-Pierre Nkurunziza, both sides have demonstrated total respect for the cessation of hostilities. Bujumbura Rural is the only province where members of the PALIPETHUTU-FNL (FNL) still clash with government forces, which now include members of the FDD. FNL, the sole remaining rebel group in the field, has only recently demonstrated its willingness to join the talks. It has been seriously weakened by the operations of EVALUATION OF USAID PROGRAMS IN BURUNDI 16 forces under the new integrated high command of the Burundi army (FAB) and the FDD and by sanctions imposed by regional leaders on the movement in June 2004. An integrated military high command responsible for carrying out the reform of the army has been working since January 2004 on a plan to integrate former FDD rebels. The Ceasefire Monitoring Commission (CMC) has proposed a joint operational plan (POC) for disarmament and demobilization. Both sides have demonstrated willingness to implement part of the plan by separately disengaging and assembling their forces and respecting the cessation of hostilities. The process is running out of steam, however, because of lack of commitment and insufficient funds to carry out the actual integration. The main issue that remains to be solved is the conflict between two different conceptions of the DDRR process. International donors, including the World Bank, put an emphasis on disarmament and demobilization of the rebels, while the Global Cease Fire Agreement talks about “integration” of both troops into a new national army first. Furthermore, many issues remain unsolved. The Forces Armées Burundaises can hardly be considered as a unified body. Many officers are now openly rebelling against their superiors and accusing them of misinforming them about the demobilization and disarmament process. They claim that they refuse to remain hostages to the 1993 “putschist officers”, now that the war is over, and that they want to negotiate their own future. For example they want to negotiate their demobilization package directly, as it is very likely that most ex-combatants will want to buy a piece of land and build their own house in the region of origin. Repatriation and land A permanent suspension of hostilities in Burundi and the prospect of elections carry the risk that a great many people who were uprooted will rush home to a country not yet prepared to receive them. Only by means of completing thorough advance preparations will it be possible to repatriate approximately one million Burundians. Both the transitional government and the international community2 however have paid too little attention to the land question that this repatriation involves. Burundi has experienced two main waves of refugees. The first was in 1972, when genocidal acts of the army against the Hutu elite led approximately 300,000 people to flee, mostly to Tanzania. In 1993 the assassination of President Ndadaye and the massacres that followed started another round of flight and displacement. After ten years of war, over 500,000 persons are estimated to be in the refugee camps in western Tanzania. Another 300,000 persons are thought to be dispersed across Tanzania. There are approximately 280,000 persons permanently displaced persons in Burundi itself, living in 226 registered places. Moreover, every month 100,000 people on average became temporarily displaced as a result of the ongoing fighting. To one degree or another all these refugees and displaced persons have been the victims of land expropriation. The 1972 Hutu refugees were deprived virtually systematically of their goods and lands in the fertile Imbo plain by the Micombero and Bagaza regimes. The 1993 refugees were less often the victims of expropriation - but this does not mean that their return will be any easier. As with displaced Tutsis 2 See ICG reports, 2004. EVALUATION OF USAID PROGRAMS IN BURUNDI 17 currently living on the edges of the cities, many were victims of profiteers who benefited from the absence, or death, of the legal owners either to seize land or to sell it at a profit. War-profiteers have also manipulated and encouraged rampant administrative and political corruption to appropriate the estates that could have been used to help with the resettlement of refugees. The Burundi government and the international community have thus far failed to recognize the scale of the problems they will face with the return and resettlement of refugees and displaced persons. There is a precedent in Burundi’s history for what the poorly prepared return of refugees can mean for political stability, however. Following the election victory of FRODEBU in 1993 some 50,000 refugees from 1972 returned spontaneously. Their arrival was handled badly by the newly installed government, which was trapped between the necessity of returning to the refugees what the former regime had stolen from them and the fear of the Tutsis that they would be the losers. It was, in part, the demonstrations of expropriated Tutsi families that led to the coup d’état and the assassination of President Ndadaye on 21 October 1993. The resettlement issue (with varying degrees of intensity depending on the areas) will be an ongoing source of tension during the transition process and could become an issue during the elections. Those disappointed by the peace process are likely to use every opportunity to block reforms in the first few months after a definitive cease￾fire. There is scope for both Hutus and Tutsis to engage in a political war over the restitution of land to refugees and displaced persons and over payment of reparations and compensation to expropriated or resettled families. The foreseeable disappointment of a large number of refugees who will be unable to recover their property offers ideal political opportunities for the opponents of the process, and could place the entire transition in jeopardy. The urgent requirement in this situation is to defuse the land conflict trigger through the creation of an innovative transitional judicial process designed exclusively for land management. One example would be to adapt the traditional institution of Bashingantahe, and work with the local administration and CNRS to help implement the resettlement process. These new structures should help elaborate the precise rules governing: a) the resettlement of refugees and displaced persons on their property; b) compensation criteria for expropriated families; c) resettlement criteria for estate lands; d) the amount of reparations for those unable to return to their former property; and e) establishment of the principle whereby a widow has full rights to the property of her deceased husband. Elections Arusha November 1, 2004 as the deadline for elections, and tensions have been growing in the lead-up to this new phase in the peace process. At the last regional summit on Burundi on June 5, 2004, the transitional government proposed rescheduling the elections to October 2005. Regional leaders rejected this ploy, insisting that those conditions that were already agreed on be respected, to avoid an institutional vacuum and the reopening of the negotiations. They called for “last chance” consultations, which were held in Pretoria on July 21, 2004. Progress was made but the parties have yet to agree on post elections power sharing. EVALUATION OF USAID PROGRAMS IN BURUNDI 18 The main issue of the talks was to harmonize the Arusha agreement, signed by UPRONA and FRODEBU, with the demands of the FDD rebels on the elections, post elections constitution and power sharing arrangement. The FDD have now fully accepted the Arusha measures, the principle of ethnic quotas, an indirect election of the President through Parliament, and the existence of a Senate based on ethnic quotas. However, UPRONA, fearing to score miserably in the elections, but still retaining instruments of power like the army command and the economy, seeks to guarantee a 40% share of posts in the post elections government during at least five years. Some of its leaders, who have been involved in past crimes, also want to guarantee their immunity from prosecutions through the elections. UPRONA, mainly representing Buyoya’s constituency of Tutsis from the southern region of Bururi, is now attempting to reach an alliance with other Tutsi parties. The FDD, more confident in a ballot victory, wish to remain free of appointing Tutsis of their choice. The FDD have started an official campaign in the countryside and actively recruit new members for the movement. Meanwhile, FRODEBU is threatened by a continuing loss of members to the FDD. The party is eager to see an agreement on the constitution, as well as the cantonment of troops to start its own campaign on the ground. Frustrated by the failure of the Pretoria talks, it has now called for the resignation of the UPRONA vice president Alphonse Kadege. In a situation of unfinished negotiations on power sharing, the population is left in the dark about the date of the elections and the type of electoral systems that will prevail. And while the informal electoral campaign has already started, no preparation is being made for local elections, which are supposed to happen first according to Arusha. Voter registration and education also have not yet begun. Moreover, most refugees have not returned, the political prisoners have not been released and the FNL is still not officially part of the process. The armed rebels are free to move among the local population, and the army has not started returning to barracks or handing in heavy weapons. It is most likely that elections will take place in a situation where the two separate armies still will coexist with the UN peacekeepers. 1.3 US Integrated Strategic Plan The US Government has deployed resources from a variety of agencies to ensure that Burundi becomes a “peaceful, reconciled and equitable country that supports individual prosperity and national development” (USG Strategic Plan), underpinned by a goal of transition to peace and socio-economic recovery. This goal is to be achieved through support to the provision of basic social services from REDSO and PVC (through Population Services International, UNICEF, WHO, Family Health International), Food security (WFP, Africare, CARE, CRS and World Vision) from FFP, and good governance. Additional NGO activities are also funded by OFDA and PRM to alleviate suffering, facilitate repatriation and reintegration, and prevent unnecessary deaths in the population. Some funding for local NGOs also comes from the US Embassy Democracy and Human Rights Fund. EVALUATION OF USAID PROGRAMS IN BURUNDI 19 The good governance strategic objective is defined by two Intermediary Results (IR1 and IR2), ensured by the work of two NGOs, Search for Common Ground, and Global Rights (formerly known as International Law Group). Some funding is also earmarked from the CMM office in Washington for Search. The governance component of the USG strategy is the object of the present evaluation. OTI also has been implementing an accelerated program to promote citizens’ participation and peace education. This may soon shift to closer cooperation with the two REDSO￾funded NGOs, but has not been covered by this evaluation. IR1 is defined in the following terms: “transition institutions and the peace process strengthened”. The peace process here is understood in a broader sense of the Arusha agreement and the overall process to achieve its terms. The activities supported are consequently the contributions by NGOs to a greater public support and interest, even to a strengthened a constituency for peace, as well as targeted support to all institutions that are critical to the peace process. The implementation of the agreement would result in the root causes of the conflict being addressed, and is consequently seen as a valid priority. IR 2 is defined as: “increased participation of civil society”. The USG Integrated Strategic Plan for Burundi (2003-2005) does not describe civil society in detail, but points out that the population has been denied the opportunity to “provide meaningful inputs into the structure of the governing bodies or receiving the benefits that their economies have produced”. It also mentions significant community initiatives to organize against violence and support the peace process. In practice the evaluation finds that the range covered by civil society extends from nationwide structures such as the Catholic Church commissions to rural agricultural cooperatives in rural communes, often handed over from previous development efforts. USAID management structures have had a reduced scope. Due to the instability in the country (for example a general evacuation of the capital took place in July 2003, where only essential Embassy staff remained in country) and the pervasive insecurity, there is only a Limited Presence Office on the ground, which is supported by a team based in REDSO in Nairobi. 2 BRIEF PROGRAM DESCRIPTION Search for Common Ground opened its presence in the country in 1995, initiating with the launch of Studio Ijambo (“speaking the truth”). This was rapidly followed by the Women’s Peace Centre in 1996, whose intention was to allow women to meet across the divides of their society through associations. In 2001 the Youth Program was launched, as well as the Integration Initiative. Studio Ijambo journalists contributed to the creation of Radio Isanganiro, which, even though it is an independent association, is exclusively funded from Search. In 2003 the USAID funded Victims of Torture Project was also launched, along with other components run in close coordination by other NGOs3 , for which Search was given the lead. By 3 Global Rights , Ligue Iteka, THARS (Trauma Healing and Reconciliation Services) and APDH (Association pour la Défense des Droits Humains). Help is provided by other organisations, such as SWAA. EVALUATION OF USAID PROGRAMS IN BURUNDI 20 the end of 2003 the Integration Initiative had closed down, and is consequently not covered by this evaluation4 . The program was established when conditions were not conducive at all to a peace￾building programme. Violence was widespread in all provinces, and access to all areas was entirely dictated by the security situation. The organization concentrated on the media as a method of outreach. There were two obligatory forms of cooperation in the early years, one with RTNB and one with the Ministry of Defence, whose priorities had to be accepted. The organization however benefited from the presence of the Special Representative of the Secretary General (SRSG) Mr Ould Abdallah, who had been facilitating dialogue between FRODEBU and UPRONA and pursuing a policy of inclusion since the beginning of the crisis in 1993. There was a direct synergy with the negotiations between the two main parties (in September 1994 there was an agreement to share power) and the SRSG strategy was to strengthen the moderate elements, based on an alliance, marginalizing the extremes. The personal contact established at that time helped launch the program, and gave Search the neutral legitimacy it has since enjoyed. Search came to operate in a highly strategic area, where there was a state monopoly, with a single party line. There were no peace-building media programs, just at the time when the role of the media in Rwanda was highly visible and foreign resolution to act on this was high. To maximise the opportunities, it was decided to operate in a pragmatic manner. The Search objectives are relatively unformulated, but clearly predate those of USAID, which are aimed at the Arusha Accord. From 1995 to 2003 inclusive, the program received an average of USD 2 million from USAID per year5 . Funding has also been received from other sources, although in lesser amounts (SIDA and DFID). The focus of Search’s work is on collaborative methods of resolving conflicts. These are defined in the broadest of terms, and even though some members of staff understand this in terms of violent conflict, this is not the goal chosen by the organization as a whole (which states in its documents that it aims to promote reconciliation at all levels of society). Since 2004 two cross-cutting themes have been given for all programmes: the reconciliation of divided communities, and dialogue within Burundi and beyond. These categories are taken to be intuitively understood, but revolve around coexistence and communication, and in practice staff report that they have difficulty in defining what these themes include or exclude. The organization has provided a number of ancillary objectives in support of its overall goal, which will be reviewed in the evaluative sections. The evaluators have however found it easier to retrace these objectives on the basis of the four program components, which run in relative autonomy from one another. The target population determines these. The objectives for each component as described below are based on interviews with the program staff, in particular the directors. 4 It was however included in the external evaluation by Mr Amr Abdalla et al. in 2001. 5 USAID Modification of Assistance, 21 October 2003 EVALUATION OF USAID PROGRAMS IN BURUNDI 21 1. The media program includes Studio Ijambo (producing material to be used by all media outlets in Burundi) and since November 2002 supports Radio Isanganiro, which is the most widely listened to national broadcasting radio in the country. The staff of both bodies (some 50 people) described their goal as the promotion of a style of journalism, the “crossroads style” (“Carrefour”), based on the ideal that “all can gather, without any exclusion, without dehumanising stereotypes, and seek solutions”6 . 2. The Women’s Center is in fact an association of 315 women7 trained to project a pacifying influence in communities, who operate as a network of volunteer resident facilitators (in one province, Makamba, this includes exceptionally some men), supported by the Centre to deal with problems and promote reconciliation events. Search support teams are located in three, but the network covers seven Provinces. 3. The Youth project is designed to reduce violence. This is achieved by approaching groups of young men who are particularly vulnerable to mobilisation into organised violent action, especially university students, and young men from the suburbs (which are traditionally poorer). They are co-opted into events and debates, where sensitive issues and the theme of peace are tackled openly. This is the smaller component of the Search program, using only eight staff members, present in Ngozi, Ruyigi and Bujumbura Provinces. 4. Victims of Torture’s aim is to create an organised system of referrals to assist victims of physical and mental violence. 90% of the cases treated (through hospitality, advice, and material assistance in particular transportation when required for clinical reasons) are rape victims, but the program seeks to cover all forms of trauma. Staff estimate that 60% of all Barundi have been exposed to trauma, and there is a need to reconstitute the social fabric. The organization is based in Bujumbura, where its offices host Studio Ijambo and Radio Isanganiro, as well as the staff of the project components. Offices are established in four provinces including Bujumbura, Makamba, Ruyigi, and Ngozi. The provincial antenna pool the logistical resources for each project component, in particular the single vehicle each one has. The evaluation team was able to observe many instances of symbiotic interaction between projects (such as journalists training women facilitators on rumour management), even though the field offices were not headed by a single person to ensure geographic coherence between the components. 3 DESIGN AND PREPARATION: SEARCH FOR COMMON GROUND The outcomes to be achieved by Search are set out in some detail in the many documents which frame the work of the organisation, in particular “Program Goal and Objectives SFCG” 8 , the grant application documents, and a variety of framework documents in various stages of evolution, such as “Perspectives 2004” or “Log Frame: Reconciling Divided Communities – Programmatic Domains”. 6 Interview with Studio Ijambo journalists. 7 In July 2004. 8 This document which is currently distributed publicly is undated and no author is given. EVALUATION OF USAID PROGRAMS IN BURUNDI 22 These outcomes are variously described as single: reduce ethnic conflict and promote reconciliation, or as threefold: reconciling divided communities; dialogue - public conversation in Burundi and beyond; victims of torture (presumably VoT needs met). The phrasing varies slightly: the log frame provides two goals: facilitate the reconciliation of communities divided by war, and advance the efforts of reconciling communities to realize a shared vision for the future. On the other hand the USAID Intermediary Results elaborated in 2002-2003 are more focused on the institutional aspects of the political transition, mentioning specifically Transitional Government institutions and publicising important information.9 The USAID Intermediary Result 2 concerning civil society emphasises the need to build structured independent participation in the national dialogue, mentioning specifically grass roots truth and reconciliation meetings, and independent radio. Search’s objectives evolved gradually from 1995, and with an underlying focus on developing new modes of interaction among the broadest number of people (broadly described as recognising differences, seeking commonalities) spanning the society’s cleavages. While there is no document describing why one project component or other were selected other than as creative initiatives, the organization covers a very wide understanding of the term “conflict” to include all societal vulnerability to conflict and the risks of future conflict. One example of positive conflict management mentioned by the organisation is the resolution of the student strike at the University of Ngozi. Media formed the original core and was followed by more direct work with social groups: women, then youth, then victims of violence and rape. While this is predominantly related to events (The Heroes Summit, for example, where people who protected members of the other ethnic groups are awarded special mention), in recent months Search has gradually shifted its focus toward a community approach, broadly defined as activities requiring more prolonged contact with the same population group. This leads in fact to a dual focus of the program, one based on direct personal interaction within the population, the other focused on media and addressing political decision makers or diffusion of information on issues which are relevant across the nation. This has not however been articulated consciously by Search. The stated outcomes are formulated in unverifiable terms. It is for example not possible to decide, on the basis of the objectives, why a chosen intervention will be judged better or worse in meeting the objective, and to what extent the objective was truly met -- or example that reconciliation has truly taken place, even when looking at the intended related objectives (“promote inclusion of alienated groups”, or “identify and/or create opportunities for diverse groups to work together”). Subsequent efforts at rationalisation have not been able to overcome this hurdle. The evaluation team was given to see iterations of the Logical Framework devised over 2003-2004, which defined cross-cutting programmatic domains designed to enhance coherence. The stated purpose toward the dual goal of reconciliation and a shared 9 USG Integrated Strategic Plan for Burundi, 2003-2005. EVALUATION OF USAID PROGRAMS IN BURUNDI 23 vision included for example “strengthen the ability of communities to increase their solidarity”. The outputs listed include 50 communities coming together to assess the causes of the divisions between them, and the indicator is the number of cases where “truth is met with mercy”. These results were not related to any individual project component but shared amongst all of them, potentially leading to a decreased degree of responsibility for outcomes. In the design of activities, according to the responses given very cogently by the program directors and staff, the organizations implicitly refer to their own analysis of the drivers of the conflict, in particular the monopoly of information and arbitration exercised by state organs, their neglect of key issues (e.g., land tenure conflicts ), or the need to empower key groups to create a more polycentric society. However, these or some related underlying concepts are not clearly articulated, and strategies, when they are derived from an analysis of context, refer to empirical data about a gap to be filled. A good example of this is found in the January 2004 Concept Paper on elections10 , which proposes objectives, purposes, and activities which draw on existing program structures. The objectives found reflect more the two assets developed by the organization over time: Its specialised expertise in certain conflict management and arbitration techniques, and the wealth of personal contacts and trust of certain individuals, rather than a deeper analysis of what causes the most risk of conflict. The approach is a pragmatic adjustment of the content of activities, more faithfully reflecting the issues of the moment than a clearly elaborated strategy of intervention. Activities such as organizing “round tables” are emblematic of this openness. This flexibility however leads to a reduced overall efficiency. It is not possible for the organizations to weigh the merits of the different components in terms other than continuity or non-continuity with existing activities, and, unavoidably, staffing structures. Although it is important in such an unstable environment to preserve continuity in program approaches, the many areas of involvement of Search in Burundi do not represent a coordinated approach. It is not possible to say whether more or less resources should be given to one activity over another, other than on the basis of what had been done in the past. The main meaning given to internal coordination concerns logistics and office use, rather than a sequential strategy deploying one activity alongside another. The program follows the standard monitoring structure of many NGO programs, specifically including periodic reports from the project components, and annual Program updates linked to the renewal of grants on the part of USAID. This is based on activities, and references are made to the program goal and objectives, as well as to the new activities and orientations chosen. There is no use of indicators, and the monitoring relies on narrative descriptions of important events and the way in which it affects the population. The radio programs have engaged in focus group research concerning the content and attractiveness of the programs11 , including the way in which it contributes to reconciliation, according to the above mentioned survey. 10 Facilitate the Electoral Process in Burundi: Concept Paper, Search for Common Ground in Burundi, January 2004. 11 For example: Evaluation des emissions ‘Inama N’Ingingo’ et Ramutswa Iwanyu’ produites par le EVALUATION OF USAID PROGRAMS IN BURUNDI 24 Although the texts mention the desire to create a line position for evaluation, and the intention to carry out external evaluations every two years, this has not taken place. Two evaluations were carried out over the life of the program (1999 and 2001), which relied on interviews and discussions with persons who had been involved in the projects, and did not apply standard evaluation criteria. Some of the program design documents mention impact indicators such as can be gathered from Knowledge￾Attitude-Behavior surveys, but none have been conducted. The sector of conflict resolution activity is relatively new to evaluation, and the commissioning of evaluations is relatively onerous in terms of direct and indirect costs. These are the reasons given by the organization for the modest results achieved so far in monitoring and evaluation. There is however a clear interest in the subject, partly based on the notion that the quality of the program needs to be captured in a verifiable form, and be better communicated. The level of technical competence of the program staff is high, even in very innovative areas such as “rumor management”, that is,counter-acting the destructive influence of false information transmitted informally because of its sensationalism. This is due to a high emphasis given to training, as well as a high level of staff retention, at least for national staff (the Country Director position tends to rotate on a one to two year basis). The quality of content of the productions of Studio Ijambo is mentioned in an OTI commissioned survey of radio broadcasts as the main reason people like to listen to them.12 The evaluation did not collect any salient evidence concerning technical support on the part of USAID, either positive or negative. The increasing willingness of USAID staff to visit the field, and the continuity of funding, have contributed to the quality of dialogue. Quality of planning, management of constraints, multiplier effects achieved amongst components, and with other USAID activities. Search operates in a mode of optimal ambiguity as regards its concepts and goals. The non-definition of terms such as “community” (which can mean people working in the same place as well as people living on the same hill) and the nature of the divisions to be healed are partly there out of design. It allows the organization to be more reactive to the diversity of situations. The strength of the program lies instead in its most constant elements, namely the structures put in place for the radio station and Studio Ijambo, and the volunteers working in the suburbs, schools, and countryside. These are highly recognisable among the beneficiary population. Interestingly the content of the work they do, in terms of the target group and the subjects to be treated, can be altered considerably over time. The exposure that the NGO staff and persons related to its work have to the most burning issues prevailing in the country at any particular time is very high. The journalists, and the facilitator volunteers, are naturally tuned to seizing the Studio Ijambo, 2002. 12 Baseline Assessment, Gitega and Ruyigi Provinces, June 2004, OTI/PADCO EVALUATION OF USAID PROGRAMS IN BURUNDI 25 undercurrents and addressing problems as they emerge, as for example the unofficial campaign which has started in the countryside. However the evaluation observed the possibility of more operational responsiveness to the situation. As a ceasefire is signed, and there is practically no military violence in the entire country, and while party recruitment through intimidation is increasing in the rural areas because of the elections, new project activities could be undertaken. This would include in particular more work in new areas (which have been off limits in the past because of fighting), or more focus on designing information concerning unfamiliar concepts like Municipal and Provincial elections. There is a good transfer of resources between the networks on the ground, and the training provided by specialists. There is also a high degree of horizontal transfer of information between different organizations the closer one comes to the communities. For example Search, CARE and GR work in the same area in Ngozi, and their village level volunteers are in fact in many cases the same people. This has not led to poaching of staff in terms of higher incentives paid by one organization or the other (although the evaluation was informed that tensions had developed concerning this issue in Ruyigi as OTI was gradually fielding volunteers). However there is an underlying pull in Burundi toward providing material benefits for humanitarian reasons. It is not clear from an objectives point of view, why Search moved into the torture victim sphere, and then from there to assistance to AIDS victims. The agency has been providing transport to Bujumbura for those needing anti-retroviral drugs, and financial assistance for the procurement of drugs in the hospitals. This has placed it in competition with ICRC, which is providing medicine for patients in hospitals. This leads to absorption of staff energy and logistical resources in areas which may not be the most productive in terms of peace-building. Although it can be argued that rape victims and former detainees are an important part of the healing of the country, it is not clear why their needs are better addressed by an organisation primarily focused on media and the prevention of conflict. 4 MANAGEMENT AND IMPLEMENTATION: SEARCH FOR COMMON GROUND Partners: delegation of management decisions, consultation in design and execution, participation in implementation, contextualisation of intervention, integration of gender, and general coordination with partners. Search for Common Ground Burundi operates closely with partner agencies, one of which in particular was created to implement the new style of journalism sought: Radio Isanganiro (via Association Ijambo, which provides the governing board for the radio). These relationships stipulate that the orientation of the partners respect the objectives of Search, defined in the sub-contractual agreement between Search and Association Ijambo, for example, the promotion of dialogue, peace and reconciliation. EVALUATION OF USAID PROGRAMS IN BURUNDI 26 Search supports in financial terms (apart from Izanganiro) Radio Bonesha, three theatre groups (Gezaho and Bamagana), and three NGOs active in the human rights and victim support sphere (Ligue Iteka, Trauma Healing and Reconciliation Services, and Association pour la Promotion des droits de l’Homme). The total annual amount in 2003 for these institutions was USD 400,00013 . In addition to this it has provided equipment to RPA (a radio station founded by a former journalist of Studio Ijambo but which is struggling financially), and RTNB. It cooperates also closely with SWAA in the victim support sphere in Ngozi, and with Jamaa in Bujumbura14 . More importantly Studio Ijambo provides for free its productions to all the independent radio stations in the country. There is no overall capacity building policy as such within the organization. Small associations are widespread in Burundi, and many of them survive on very small budgets. Some of them engage in programs, which are broadly compatible with Search’s objectives. The relationships of Search with its partners are however heavily influenced by previous commitments, and by contacts on the ground. A document entitled “Les forces stimulantes” dated February 200315 lists six assets, including capacity building. This document drafted at a staff workshop shows that the question of how capacity building is done is the only one to provide no answer. It is for example, not clear why Radio Bonesha (which has been targeted for support by OTI) and not RPA receive funding from Search, or why Radio Isanganiro receives such disproportionate funding (one quarter of the total given to partners in 2003). The sub-contractual agreement with Association Ijambo / Radio Isanganiro includes a gradual phasing out of funds over three years, but at the same time there is no particular support given to ensuring a new donor base, or even to justify why such a rapid pace of decreasing funding. Radios in Burundi depend entirely on government or donor funding (at this point in time all independent radios and a large part of the public radio are funded by foreign agencies), and in fact the perception that Radio Isanganiro (to which the population surveyed in the OTI study referred as “the American radio”) is fully funded by USAID has weakened its fundraising potential. By fulfilling the key needs in Burundi for good content of radio programs, and providing proper conflict resolution techniques through community facilitators, an important contribution was made by Search to the conflict situation. It is however undeniable that the generation of a fabric of civil society in the country to continue 13 Liste des associations avec lesquelles Search a un contrat de partenariat actuellement . Mai 2004, Search for Common ground. 14 A non exhaustive list of the organizations with which Search cooperates on specific activities was provided by Search to the evaluation : Association pour la Paix et les Droits de l’Homme (APDH), CAFOB, Dushirehamwe, ACORD Burundi, Réseau Femmes et Paix, Centre Indépendant de Recherche et d’Initiatives pour le Dialogue (CIRID), Ligue Africaine pour la Non Violence Active (LANOVA), Association des Jeunes pour la Non Violence Active (AJNA), Association des Guides du Burundi (AGB), Union pour la Promotion des Rasta (UPR), Club Génies en Herbe (GEH), Association pour la Promotion de la Fille Burundaise (APFB), Association JAMAA, Association des Jeunes Imboneza, Association des Etudiants de Rumuri (ASSER), Club Nouvel Horizon, Association des Jeunes pour la Paix sans Epée, Club Bâtissons la Paix 15 Planification stratégique, synthèse. Search for Common Ground au BURUNDI Février 2003. EVALUATION OF USAID PROGRAMS IN BURUNDI 27 transmitting the messages is also fundamental. This is not reflected in the objectives of Search, in spite of frequent references to strengthening the capacity of the society to cope better with conflict. The approach chosen by Search has been to achieve a good representation of a diversity of groups, including ethnic balance (although it does not arguably reflect the proportions in the population), and gender balance. The preponderance of women in the organization as a whole ensures that the more difficult aspects (in cultural terms) to treat in public fora, such as rape, are not neglected, but well covered. The closely woven networks of family and acquaintances in Burundi ensure that information flows freely to and from the organization. The organisational structure of the NGO is clear, and tasks are well distributed between the different components (even to the point where no organigram is required, but levels and specializations are reflected in the telephone list). Interviews reflect an impressive continuity of presence within the organization, with high rates of staff retention (probably enhanced by the poor state of the labour market for qualified personnel outside the NGO world). The evaluators were however struck by the relatively high level of discontent of staff in the organisation, particularly as regards incentives and promotions. In a highly polarised society it is inevitable that the conflicts be reflected within the offices, even in an organization, which prides itself on its resolution skills. However the institution opens itself to more suspicion and rumour by not having clearly recognizable standards of performance, and criteria for recruitment. These could be placed in highly visible locations, and give less ground to a pervasive sense of injustice in what is after all, in the context of the country, a privileged working environment. Overall the programs are based on the twin assets of partnerships with institutions and people (such as facilitators, or partners) and on conflict mitigation processes (enhancing dialogue for example). The skills of the staff in the Women’s Centre and the Youth Centre are honed to identifying so-called leaders, which are in Barundi culture those who can see afar and whose judgment is respected. In some instances potential trouble makers are gathered together (for example at a school in Makamba where anxieties had been roused after grenades were mysteriously thrown) and discussions organized to defuse tensions before they get out of hand. Such activities require great sensitivity to group dynamics and the political and cultural context, which Search personnel have clearly achieved. 5 IMPACT OF SEARCH FOR COMMON GROUND 5.1 Providing Attractive Models That Capture the Imagination The media programmes have had a significant impact on the conflict, because of the quality of design and the relevance of the content, which enabled Search (i.e. Studio Ijambo and Radio Isanganiro) to catch and hold the attention of the majority of the radio-listening population. EVALUATION OF USAID PROGRAMS IN BURUNDI 28 An OTI survey carried out in two eastern provinces report that focus group participants unanimously identified radio and meetings with local administrative authorities as the two primary sources of information on current events. In both target provinces, Radio Isanganiro won acclaim as the most popular radio station, both for its entertainment and for its dissemination of timely, accurate and detailed news information. In Gitega Province, RPA ranked second in listernership. In Ruyigi Province, Tanzania-based Radio Kwizera ranked second for its coverage of issues related to repatriates. All of the groups covered by OTI which had participated in media outreach were visited physically by Radio Isanganiro as part of its outreach. Community members uniformly expressed interest in participating in media outreach activities. When asked what they would say to the media, several groups said that they would bear witness for the rest of the country on how to cohabitate and how to have peace and security. Persons also said they would share stories on how individuals hid people of another ethnic persuasion during the crisis. Another survey of listenership shared by Search with the evaluators16 shows that nationally among sex workers national radio in Kirundi is listened to by 18.5% of the population, followed by Isanganiro (10.1%), then RPA (9.3%), Bonesha (5.4%), RTNB in French (2.5%), Radio Kigali (1,1%), and BBC and Radio France Internationale (both 0.3%). Voice of America had no listeners. These groups mentioned Bonesha, RPA and Isanganiro as having the highest level of trustworthiness, in that order. Among teenagers the proportion of listenership shot up to 27% nationally, with national Kirundi radio gaining only 19.6% in second position. Yet another survey17 shows that 37% of youth (mostly males) have a radio, and 60% of them mentioned as reasons for listening to the radio that it contributes to reconciliation (55% are concerned about AIDS, 55% see it as educational, and only 16% as a means of getting information). This last group preferred RPA (55.3%) to Radio Isanganiro (32.3%). Search’s youth and women’s project components are much more localised in their impact than media (a few communes in four provinces and sharply defined groups attached to particular institutions), but verifiably very deep. The evaluators were given to attend a training session of 60 volunteer facilitators in Ngozi, where participation had cost personally to some of the women (one day of travel in difficult conditions, handing over of domestic responsibilities), and where learning lasted well into the evening, by personal choice. This is probably because of the relevance of the issues covered; the problematic yet important nature for the population of information concerning the peace implementation (particularly elections); and the perceived need to build new capacities to respond to conflicts, as the society moves away from the traditional modes of arbitration. 16 Annexe 5: Niveau d’écoute des stations de radio au Burundi (2003 et 2004) 17 Sondage: Préférences radiophoniques des jeunes, 250 jeunes des quartiers nords, Nduwabike / Search for Commong Ground, août 2003. EVALUATION OF USAID PROGRAMS IN BURUNDI 29 Some respondents who were given to observe the events organised by the women facilitators expressed scepticism concerning the sincerity of the emotions and gestures expressed, for example in cases of public forgiveness for crimes committed. The evaluators would be cautious, in light of the very polite culture of Burundi, to conclude from declarations about the depth of change. However the fact that issues were mentioned publicly at all was a measure of deep seated changes, partly caused by the work of Search. 5.2 New Forms of Interaction The productions have led to a new interaction between media and the political class over the period 2001-2004 and a universally recognised improvement in quality of reporting in the country as a whole. This point has been brought up consistently in the course of the evaluation, by respondents who indicate the fact that most independent￾minded journalists have worked for or benefited from Studio Ijambo productions, and that the current media landscape and style is inspired by Studio Ijambo. There is of course also an influence coming from the fact that political opposition is now legitimate and institutionalised, contributing to more open debate. But this could not take place without the cooperation of journalists. Search has had an impact on the way the journalistic profession is perceived by the political class and by Burundi society. As mentioned in the introduction, media have had a reputation of bias in Burundi society that has always relied more on RFI, BBC, VOA to get news than on its own media. Moreover, every Burundian remembers how radio was used to incite the population to murder Tutsis in 1993 and Hutu politicians between 1994 and 1996, and how Radio Mille Collines called for genocide in 1994 in Rwanda. Journalists have now demonstrated that they can responsibly and fairly manage information and the training they have received has increased their professionalism. Also, politicians now call for the possibility of including opponents in public debates, as a way of convincing their audience (which would otherwise trust the message less). 5.3 Capacities Created Search has created an environment that has allowed individual talents to blossom. Alexis Sinduhije and Agnes Nindorerra have both won awards and grants from Harvard University. Jean Marie Gasana has been recruited by Fondation Hirondelle in Arusha, International Crisis group in Nairobi and Institute for Security Studies in Pretoria. Jeanine ?, ex journalist at SI, has become the director of Radio Isanganiro, Alexis of RPA. Search has become an institution in Burundi employing more than 100 people. The organization has a commitment to see the conflict through all its process of transformation, which is a very valid one. However, in contrast to GR, which is very concerned about its exit strategy, Search doesn’t seem to have one. The approach taken is that continued response to tensions takes the place of any handing over. There has been little attention on the part of Search to the capacities it has created, in the form of support to individual journalists, but also radio stations, other than Radio Isanganiro. The previous sections have covered the issue of an absence of a policy EVALUATION OF USAID PROGRAMS IN BURUNDI 30 guiding the allocation of resources. The fact that none of the radio stations, nor networks of facilitators or groups of youth show any guarantee of sustainability once the Search program phases out is an object of real concern. EVALUATION OF USAID PROGRAMS IN BURUNDI 31 6 FINDINGS AND RECOMMENDATIONS FOR SEARCH Recommendation 1: The IR should be adjusted by USAID to include the societal dimensions of conflict in the country, as these are key to the conflict (particularly because of the elections, and importance of land conflicts), and better reflect the programs funded. Recommendation 2: A long-range conflict assessment should be carried out (either by Search, or by others but with the results made accessible to Search), leading to a re-evaluation of the current Search priority ranking of different areas of program interest, and possibly to a revision of some of the indicators of program success. This conflict assessment should be based on the implementation of the peace agreement, and should articulate those causes and drivers of conflict that an organization like Search has the potential to ameliorate. The triggers which the evaluators found for the peace process are: 1) Demobilisation, Disarmament and Reintegration; 2) the repatriation of refugees; and 3) the poor quality of information about elections and the economy. All three are areas where there will be much critical change in the short term. Programs should address these issues more specifically. In particular they should target the future demobilised soldiers and help them get information on DDR packages, as well as vocational training. Recommendation 3: There should be a new planning workshop in which the coordination between components is clarified, based on the conflict assessment, and an analysis of the impact achieved, or the potential for impact. There should also be an increase in the logistical capacity, funded by USAID, to maintain the efficiency of the programs. There should be a better conceptualisation of the link between objectives and partnerships, and a definition of the relevance of one to the other. The programs should in this process define partnerships and lasting institutional survival as one of the impacts of the programs. Recommendation 4: Both organisations, but particularly Search, would need to formalise and explain the structures they run. The priority must be given to avoiding providing an excuse for criticisms about unclear staff management principles, rather than squaring the circle between representativity and merit based recruitment. Recommendation 5: While the notion of strengthening the relevance of the program by increasing the election support dimension, it is crucial for USAID not to upset the balance created by proposing new program structures, but take advantage of the relative indetermination and flexibility of programs to adjust to evolving reality. The precedent of VOT should not be reproduced here. EVALUATION OF USAID PROGRAMS IN BURUNDI 32 Recommendation 6: The new focus on elections needs in fact to be reviewed from a variety of perspectives, in particular sustainability, protection and guarantees of political immunity to manipulation (for example advice in human resource management), demarches in cases of intimidation. Recommendation7: Search should seek to bring more funding to the youth and women’s activities to extend their geographic scope. Although the impact of VOT was not measured, the past evaluation suggests it is responding to conflict rather than preventing it, and we would recommend it be handed over to another service organisation operating in the health sector. A review of the importance of Search-VOT should be carried out, which currently constrains the performance of the other programmes. The conceptual link between conflict mitigation and assistance to victims of rape is tenuous. Recommendation 8: More attention needs to be paid to security of Search trained “messengers of peace”. The activities of both organisations lead to a high exposure to security risks for those who cooperate with them. This raises issues of duty of care, which need to be thought about more carefully than is currently done (maybe by generating more visits to show that there is a commitment to the work and information flows). Recommendation 9: This burgeoning process of impact monitoring should be promoted further, and funded by USAID. This would allow Search (and USAID) to map in a more deliberate way the manner in which the program relates to the drivers of the conflict. Recommendation 10: Search should agree with GR and some other organizations in civil society on a “code of conduct” and an acceptable “campaign language”, and start monitoring the pre campaign in a systematic way as of now. Search could contribute in the training of local election observers through its youth and women’s programs (see Kenya and Zimbabwe experience). However this should not lead to an electoral assistance program, which should be instead supported by an independent organisation, preferably from the region. EVALUATION OF USAID PROGRAMS IN BURUNDI 33 ANNEXES EVALUATION OF USAID PROGRAMS IN BURUNDI 34 Annex I: List of Persons Met Mr Francis ROLT, Deputy Program Director, Brussels, covering media in Africa, Search for Common Ground Cheyanne Church, Director, Institutional Learning and Research, Search for Common Ground Ozong Agborsanganya, Director Subsaharan Africa Programmes, Search for Common Ground Mr Robert LUNEBURG, USAID Coordinator, Bujumbura Mr Alex LASKARIS, Deputy Chief of Mission, US Embassy, Bujumbura Mrs Danielle LUSTIG, Country Director/Representative, Search for Common Ground Mr Stéphane MORA, Deputy Representative, Search for Common Ground Ms Lena SLACHMUIJLDER, Director, Studio Ijambo, Search for Common Ground Ms Francine NZIBAREGA, Director of the Studio Ijambo Foundation Ms Odette KWIZERA, Journalist, Studio Ijambo, Search for Common Ground Ms Francine GAHIMBARE, Journalist, Studio Ijambo, Search for Common Ground Mr Désiré NDAGIJIMANA, Journalist, Studio Ijambo, Search for Common Ground Mr Adrien SIDAYGAYA, Journalist, Studio Ijambo, Search for Common Ground Mr Aloys NIYOYITA, Journalist, Studio Ijambo, Search for Common Ground Ms Zura NYANDWI, VOT, Search for Common Ground Ngozi Ms Marie Vianney GAHUNGU, Women’s Centre, Search for Common Ground Ngozi Ms Dominique BUTOYI, Coordinator Centre for Family Development Women facilitators, Ngozi Mr Jean IHOTORIHIGWA, Bureau of Diocesan Development, Ngozi 5 students, Association pour la Paix et les Droits de l’Homme, Ngozi Ms Sylvana INAMAHORO, Coordinator Society for Women Against AIDS in Africa Mr Jean Bernard COUPPE, Health Delegate, International Committee of the Red Cross Mr Alexis Sinduhije, Director of RPA Mr Steven SMITH, Senior Regional Conflict Management Advisor, USAID/REDSO Mr Emmanuel BULAMATARI, Youth Project Director, Search for Common Ground Ms Leanne BAYER, Chief of Party, PADCO/OTI/USAID Mr John RIGBY, OTI/USAID Coordinator a.i. Ms Marie Goreth NAHIMANA, Program Assistant, UNHCR EVALUATION OF USAID PROGRAMS IN BURUNDI 35 Ms Perpétue KANYANGE, Director, Women’s Peace Center Ms Jeanine ?, Director, Radio Isanganiro, Search for Common Ground Mr Abdul Aziz THIOYE, Director, VOT, Search for Common Ground Mr Tony JACKSON, Great Lakes Policy Adviser EVALUATION OF USAID PROGRAMS IN BURUNDI 36