Report No. 32587 Toward a Conflict-Sensitive Poverty Reduction Strategy Lessons from a Retrospective Analysis June 30, 2005 SDV ESSD Document of the World Bank ©2005 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail:pubrights@worldbank.org. ACRONYMS AND ABBREVIATIONS APC All People's Congress (Sierra Leone) BiH Bosnia and Herzegovina CAF Conflict Analysis Framework CBO Community-based Organization CPN Unity Centre Party (Nepal) CSO Civil Society Organization DEPAC Development Partnership Committee (Sierra Leone) DfID Department for International Development (UK) DWG Donor Working Group on the Peace Process (Sri Lanka) GNI Gross National Income HIPC Highly Indebted Poor Countries IDP Internally Displaced Person IFI International Financial Institution IMF International Monetary Fund I-PRSP Interim Poverty Reduction Strategy Paper LICUS Low Income Country Under Stress LTTE Liberation Tamil Tigers of Eelam (Sri Lanka) NCHR National Commission for Human Rights (Rwanda) NGO Non-governmental Organization NPFL National Patriotic Front (Sierra Leone) NPRS National Poverty Reduction Strategy (Cambodia) NRS National Recovery Strategy (Sierra Leone) NURC National Unity and Reconciliation Commission (Rwanda) PRA Participatory Rural Appraisal PDR People's Democratic Republic (Laos) PRS Poverty Reduction Strategy PRSP Poverty Reduction Strategy Paper RUF Revolutionary United Front (Sierra Leone) SAPRI Structural Adjustment Participation Review (Ghana) UNAMSIL United Nation's Mission to Sierra Leone UNDP United Nations Development Programme UXO Unexploded Ordinance Vice President Ian Johnson Sector Director Steen Lau Jřrgensen Sector Manager Caroline M. Kende-Robb Task Team Leaders Per Egil Wam/Katrina Sharkey Foreword In what ways have Poverty Reduction Strategies (PRS) in conflict-affected countries taken their particular contexts into account? To what extent have an assessment of the sources of conflict and the ways they interact with poverty informed the strategies? How did groups that were socially excluded or especially conflict-affected participate in the process? How did the countries plan to address sources of conflict and deal with the destructive consequences of violence? How did governments organize the PRSP preparations in divided societies, and how did international donors engage with the process? These are among the key questions discussed in this report, based on the lessons from PRSP retrospective case studies in Bosnia-Herzegovina, Burundi, Cambodia, Chad, Georgia, Nepal, Rwanda, Sierra Leone, and Sri Lanka. The report analyzes the key experiences in these nine countries as they developed their first PRSP. This report constitutes the first year's product of a three-year program that aims to contribute to effective poverty reduction strategies in conflict-affected countries. The program is conducted by the World Bank's departments for Social Development (SDV) and Poverty Reduction (PRMPR), and is funded by the UK Department for International Development (DFID) and the World Bank. It includes the following components: Retrospective study in 9 countries Analysis and dissemination of country lessons Development of appropriate measures for a conflict-sensitive PRSP Support of conflict-sensitivity measures in selected countries Monitoring of country experiences Feedback and dissemination of experiences The report was written by Shonali Sardesai, Katrina Sharkey and Per Wam, with contributions from Sigrun Aasland, Deborah Davis, and Mark Mattner. The report also benefited from inputs and advice from Duncan Overfield and Robert Wilkinson (DFID), and the World Bank peer reviewers: Bhuvan Bhatnagar, Maria Correia, Brendan Horton, and Linda Van Gelder. The work has been guided by Ian Bannon and Caroline Kende-Robb (SDV). The retrospective case studies were conducted by the following team: Sigrun Aasland (Chad), Niclas During (Burundi), Lindsay Judge (Bosnia Herzegovina, Georgia, Nepal), Svetlana Luca (Cambodia, Nepal), Victoria Salinas (Sierra Leone), Shonali Sardesai (Sierra Leone, Sri Lanka), Katrina Sharkey (Chad), Maude Svensson (Burundi, Rwanda) and Per Wam (Cambodia). The team received inputs and help from a large number of individuals, including government officials, NGO staff, donor representatives, local consultants, and DFID, UN, and World Bank staff working with the countries covered by the retrospective study. Discussions on the emerging lessons covered by the report took place during the April 2005 ESSD and PREM weeks at the World Bank in Washington DC; in May 2005 at the Conflict Prevention and Reconstruction Network meeting in Stockholm, and attended by donor and international organizations active in conflict-related work; and at PRS-conflict seminars at DFID in London and at the Ministry of Foreign Affairs in Oslo. These discussions contributed to the report. Table of Contents SUMMARY..................................................................................................................................................7 Inter-relationship between Conflict and Poverty........................................................................ 7 Findings and Lessons Learned.................................................................................................... 8 Participation............................................................................................................................ 8 Poverty Diagnostic................................................................................................................ 10 Policy Actions....................................................................................................................... 12 Institutional Arrangements.................................................................................................... 14 Donor Behavior..................................................................................................................... 15 Key Issues for Conflict-Sensitive PRS Development............................................................... 17 INTRODUCTION ......................................................................................................................................18 Organization of the Report........................................................................................................ 18 SECTION I: CONCEPTUAL FRAMEWORK AND APPROACH..........................................................19 Interrelationship between Conflict and Poverty ....................................................................... 19 Approach................................................................................................................................... 21 What does Conflict Sensitivity entail?...................................................................................... 22 SECTION II: CONFLICT FACTORS .......................................................................................................24 Country-Specific Findings........................................................................................................ 24 Burundi ................................................................................................................................. 24 Cambodia.............................................................................................................................. 25 Chad...................................................................................................................................... 25 Georgia.................................................................................................................................. 26 Nepal..................................................................................................................................... 26 Rwanda ................................................................................................................................. 27 Sierra Leone.......................................................................................................................... 27 Cross-Cutting Conflict Findings............................................................................................... 29 Governance Challenges ........................................................................................................ 29 Economic Performance......................................................................................................... 29 In-country Regional Disparities............................................................................................ 30 Social Divisions .................................................................................................................... 30 Access to Land and Other Resources.................................................................................... 31 Militarized Society................................................................................................................ 31 External Impacts ................................................................................................................... 32 Potential Responses to Conflict Factors ................................................................................... 32 SECTION III: KEY LESSONS..................................................................................................................36 What is Realistic? ..................................................................................................................... 36 Participation.......................................................................................................................... 36 Poverty Diagnostic................................................................................................................ 42 Policy Actions....................................................................................................................... 44 Institutional Arrangements.................................................................................................... 49 Donor Behavior..................................................................................................................... 52 SECTION IV: TOWARD A CONFLICT-SENSITIVE PRSP...................................................................59 Conflict-Sensitive PRSP ­ The Way Forward.......................................................................... 59 Participation.......................................................................................................................... 60 Poverty Diagnostic................................................................................................................ 62 5 Policy Actions....................................................................................................................... 63 Institutional Arrangements.................................................................................................... 65 Donor Behavior..................................................................................................................... 66 Next Steps................................................................................................................................. 67 BIBLIOGRAPHY.......................................................................................................................................68 Table 1: Sample Cases Classified...............................................................................................................21 Table 2: Potential Responses to Conflict Factors .......................................................................................33 Box 1: DefiningConflict-affected Countries and Conflict ....................................................................19 Box 2: Defining Conflict Factors (Factors of Conflict) ........................................................................22 Box 3: Bosnia and Herzegovina PRSP Participatory Process...............................................................39 Box 4: Rwanda: Conflict Sensitivity in the PRS Process......................................................................40 Box 5: Conflict-Sensitive Policy Reforms ............................................................................................49 Box 6: Sierra Leone I-PRSP vis-ŕ-vis National Recovery Strategy......................................................55 Box 7: Donor Reviews on PRSPs .........................................................................................................57 Box 8: Alternative Models for Enhancing Country Ownership............................................................60 Box 9: Adapting Successful Experiments with Participation................................................................62 Box 10: Security System Reform............................................................................................................64 Box 11: Potential Conflict Impact of Policy Actions..............................................................................64 Box 12: Donor Harmonization ­ Donor Working Group, Sri Lanka......................................................66 6 TOWARD A CONFLICT-SENSITIVE POVERTY REDUCTION STRATEGY SUMMARY Building on a retrospective analysis of the poverty reduction strategy (PRS) experience in nine conflict- affected countries1 ­ Bosnia-Herzegovina (BIH), Burundi, Cambodia, Chad, Georgia, Nepal, Rwanda, Sierra Leone, and Sri Lanka,2 this report aims to determine how the causes and consequences of violent conflict can best be addressed within a country's poverty reduction program. The purpose is to provide lessons and guidance on how to increase the conflict sensitivity of the PRSP, which has become the primary tool in nearly 60 low-income countries for articulating a vision for growth and poverty reduction. The analysis identifies the main conflict challenges that have affected the nine countries in the sample; examines how the PRSPs took these conflict factors into account in developing their strategies for growth and poverty reduction; and attempts to draw lessons based on the findings. The analysis does not cover assessment of resource allocation, budget or implementation because of uneven data and implementation time-line between the nine cases. Inter-relationship between Conflict and Poverty The analysis is based on the recognition that conflict and poverty are closely interrelated. Empirical evidence shows that poorer countries are more likely to experience violent conflict, while conflict- affected countries tend to experience higher levels of poverty. Violent conflict results in the destruction of economic and human capital. A country emerging from conflict is faced with damaged physical infrastructure, scarce employment opportunities, reduced foreign investment, and increased capital flight. In addition, conflict increases military expenditures, which diverts resources from public and social spending, and erodes the government's ability to collect taxes and manage revenues, thus undermining post-conflict recovery. This situation is worsened by weak governing institutions, which are often unable to implement policy and uphold the rule of law. Conflict often contributes to poor-quality education, inadequate social service delivery, and high levels of brain-drain. The intersection of these factors increases both the depth of poverty and the risk of conflict being reignited. Viewed from the other direction, although poverty itself is neither a necessary nor a sufficient condition for conflict, poverty factors increase the likelihood of violent conflict in three main ways. First, a combination of poverty and unequal income levels tends to result in many people, particularly young men, who can be easily mobilized and recruited to armed groups. Second, weak and undemocratic governance structures, usually present in poor countries, are often incapable of preventing the onset of violence by peaceful means. Third, if a country with a large poor population is endowed with significant natural resources, rebel organizations are able to raise finances and galvanize public resentment against perceived or real injustices. The interrelationship between conflict and poverty is often affected by group-based inequalities, i.e. those that develop between and among distinct social groups on the basis of their ethnic, social, regional, or other characteristics. These horizontal inequalities may serve to escalate conflict into violence when ethnicity is politicized and social capital, defined as associations within and between groups in a society, is distorted via the strengthening of intra-group bonds at the cost of weakening ties between groups. 1 Conflict-affected countries are defined as those that have recently experienced, are experiencing, or are widely regarded as at risk of experiencing violent conflict. 2 The program, entitled "Ensuring Effective Poverty Reduction in Conflict-Affected Countries," is funded by the World Bank and DfID, and carried out in collaboration with concerned donors. 7 Findings and Lessons Learned The analysis identified several factors of conflict3 that were present to varying degrees in all nine countries. These factors constitute challenges related to governance; economic performance; in-country regional disparities; social divisions along ethnic, religious or clan lines; access to land and resources; militarized society; and external factors such as subregional politics, refugee flows, and the influence of the Diaspora.4 The following section discusses the extent to which different aspects of the PRS process ­ participation, poverty diagnostic, policy actions, institutional arrangements and donor behavior ­ were sensitive to conflict and seeks to understand what facilitated or hindered conflict sensitivity in each of these aspects. The report suggests ways to strengthen their sensitivity to conflict. Participation5 · In countries with traditions of limited public participation, and relative to their starting point, engagement with populations on poverty issues through the PRS process opened up space for greater inclusion and domestic accountability. · PRS formulation generally took place in environments of low state capacity and legitimacy with weak links among political power, bureaucracy, and conflict-affected populations. Prospects for the PRSP becoming a vehicle for stabilization increased where the government demonstrated that poverty reduction initiatives were for the benefit of all citizens. · Limited effort was made during the PRS process to diversify the means and geographic span of communication with conflict-affected groups on PRS goals. · Managing expectations was addressed most effectively through institutionalized dialogue between conflict-affected groups and the policy level. Efforts were less successful where the participatory process was misperceived to be a one-time consultation exercise with no impact at the policy level. Poverty Reduction Strategies are expected to be built upon a country-owned development model supported by principles of broad and deep stakeholder participation, domestic accountability,6 and social inclusion. These principles have special relevance for conflict-affected societies, where transparent policymaking and attention to inequality are likely to be limited during and immediately following periods of violence. Expectations of what can be achieved in such circumstances therefore need to be tempered with realism, while conflicting parties begin to build trust and a mutual desire for poverty reduction in the country as a whole. 3 The term factors of conflict (or conflict factors) is used for those features of a country's natural resource base, economy, social structures, or political environment that have, or could potentially contribute to escalation or de- escalation of conflict. The term may include causes of conflict, although causality may not necessarily be established, as well as consequences of violent conflict, especially if they are barriers to stabilization. 4The Conflict Analysis Framework (CAF) developed by the World Bank's Conflict Prevention and Reconstruction Unit was used as a guide to identify these conflict factors. The CAF was developed to enhance the conflict sensitivity and conflict prevention potential of World Bank assistance. For more details see: www.worldbank.org/conflict 5A recent update of World Bank policy with respect to adjustment lending is accompanied by a good practice note on supporting participation in development policy operations. Participation is defined in the note as "the process through which stakeholders (those affected by the outcome of the reform...) influence or share control over setting priorities, making policy, allocating resources, and ensuring access to public goods and services." 6 This can be loosely described as accountability of the executive branch of government, which has been assigned by the citizenry, legislature, and judicial branches with the governance and stewardship of institutions, resources, and service delivery. The term also implies that stakeholder groups and individual citizens should be accountable to one another and to society, abiding by the laws that govern them. 8 The case studies revealed conditions that posed obstacles to participation, but they also reflected the use of a variety of approaches which were seen as positive steps towards civil society engagement. Countries concentrated their participation efforts largely at the national level. Local government and community level participation was generally weaker, although the Rwanda case highlighted conflict mitigation mechanisms in participatory rural appraisals at the cellule-level7, thereby rooting participation in traditional processes. Weakened institutions and social divisions in conflict-affected countries often resulted in authorities relying more heavily on peace accords and donor support to jump-start development and social cohesion. Participation was generally defined as engagement with civil society organizations (CSOs), whose ability to engage depended on their level of economic literacy and ability to represent constituents. In post- conflict Bosnia and Herzegovina, a leading NGO stated: "For the first time, citizens have been asked to express their opinions directly through public debates. [The PRSP] has thus presented unique opportunities not only for improving the lives of the poor [...], but for empowering citizens to take part in the creation of future BiH priority actions and policies."8 The following capacity limitations were also discerned on the part of governments managing the PRS process: limited in-country experience with the country-driven model, a lack of know-how for engaging with unstable areas, and a belief that the PRSP was a technocratic exercise undertaken to fulfill donor requirements. On occasion, the PRSP lacked the ownership at the highest level of government needed to sustain dialogue with civil society. Conflict Sensitivity of the Participatory Process Given the diverse range of country settings, there was considerable variation in the recognition of conflict factors in participatory processes, and therefore in the extent to which these could be considered conflict sensitive. The potential for the PRSP to become a vehicle for stabilization depended on the government's ability and commitment to involve broad communities of stakeholders, and to demonstrate that poverty reduction is for all citizens, irrespective of ethnicity, religion, or region. In PRSP formulation, Bosnia-Herzegovina and Rwanda employed participatory processes that specifically addressed the re-integration of war-affected groups. The Rwanda PRS implementation process was reinforced by a well-structured approach to consultations; which were consciously used to include war-affected groups and prevent further outbreaks of violence. In other countries, the participatory process was partially sensitive to conflict factors, but was not implemented in conflict- affected areas of the country. A crucial component of participation was the extent to which communications were used strategically to engage conflict-affected populations. Authorities in Bosnia-Herzegovina, Rwanda, and Sierra Leone, for example, promoted participation of marginalized and war-affected groups through print and radio media. There was generally limited effort to provide documentation in languages other than English,9 including ethnic minority languages, or in formats accessible to illiterate audiences. Another aspect of communication related to the management of expectations, so that exchanges with the population could promote stability and mitigate a recurrence of violence. The Bosnia-Herzegovina, Georgia, and Rwanda strategies considered the PRSP to be a vehicle for social cohesion, and appear to have done this 7The cellule is the lowest level of administration in Rwanda. The country has 9,165 cellules, each comprising about 200 households. Above the cellule level are 1,545 sectors, 106 districts, and 12 provinces. 8See www.bospo.ba/eng/prsp.htm for more information. 9Some PRSPs were presented in French, Spanish, or Portuguese. 9 effectively, while in other cases participation was misperceived to be a one-time consultation exercise with little potential for impact at the policy level. PRS participatory processes could be more conflict sensitive if: · country ownership is manifested through the explicit inclusion of war-affected groups in socio- economic and policy dialogue; · each country finds its own formula for collaboration and reconciliation ­ it cannot be imported;10 · trust in the authorities is built through respect for laws on freedom of association and access to information, and when reconciliation commissions and ombudsmen are seen to administer justice in an equitable manner; · traditional ­ potentially conflict-mitigating ­ mechanisms are used to ensure communication and collaboration between and among groups;11 · parliaments are strengthened in order to carry out PRS oversight and constituency representation more effectively, especially for conflict-affected groups; · governments and partners capitalize on the energy and outreach of youth, especially where youth groups tend to be marginalized and at risk for recruitment; · media and strategic communications are used to disseminate information on the PRS to remote and conflict-affected communities, and more importantly, to channel feedback to policy levels, thereby enhancing the voice of those most affected by conflict; and · capacity building for domestic accountability and post-conflict participation processes is promoted and supported by partners. Poverty Diagnostic · Poverty diagnostics presented a multi-dimensional view of poverty that recognized not only the income dimension but also social, human, and structural dimensions of poverty. · Conflict issues were considered in the poverty diagnostic, but the discussion of the manifestations of conflict were not the result of systematic conflict analysis; there were only limited efforts to explore how factors of conflict and poverty drive each other. · Genuine political constraints prevented governments from explicitly addressing poverty-conflict linkages. · Lack of capacity and paucity of up-to-date socioeconomic information were the major practical constraints to the poverty diagnostic. The PRSP is expected to present a comprehensive diagnostic that sets out the determinants of poverty as a basis for developing a suitable program of actions. Conflict-affected countries would benefit by going beyond assessing traditional structural-based poverty to develop an in-depth understanding of conflict- induced poverty. This will contribute to the development of a conflict-sensitive poverty diagnostic which in turn will influence the formation of conflict-sensitive policy actions. Most governments in the sample put forward a multi-dimensional view of poverty that recognizes the social, human, and structural dimensions of poverty in addition to the income dimension. In direct consultations, poor people also stressed the multi-dimensional nature of poverty affecting their lives. In 10 There are many resources available on this issue, including Reconciliation ­ Theory and Practice for Development Cooperation, Swedish International Development Cooperation Agency (SIDA), September 2003. 11Use of traditional methods should be handled with sensitivity, as they may not always be helpful. In one country in the sample, drawing on the fora established by traditional chiefs in the PRS process proved to be unhelpful as the participation of youth in PRS discussions was not encouraged. 10 these countries, in addition to the structural poverty factors that afflict many developing countries, poverty manifests itself in new ways due to the devastation of physical, human, and social capital. The PRSPs in conflict-affected countries would benefit by the recognition of these specificities of conflict- induced poverty in their poverty diagnostic. Evidence from the poverty diagnostics across the nine cases demonstrate that the diagnostics considered aspects of conflict. Burundi and Sierra Leone (I-PRSP), and Rwanda, Nepal, Sri Lanka, and Chad (PRSP) considered conflict issues to some degree in their poverty diagnostic, and recognized the interaction between conflict issues and poverty. The ramifications of their being conflict-affected, however, were not informed by any systematic conflict analysis and there was limited effort to explore precisely how factors of conflict and poverty drive each other. While lack of a conflict analysis certainly weakens a comprehensive understanding of conflict and poverty, it seems that some governments faced genuine political and practical constraints that prevented them from explicitly addressing poverty-conflict linkages. On the political front, some countries suffered from the vestiges of violent conflict and tried to rebuild trust between groups, which made it politically untenable and insensitive to discuss conflict factors in the poverty diagnostic. In terms of practical constraints, weak capacity and the lack of recent and comprehensive socioeconomic data undermined the government's ability to undertake effective poverty analysis, particularly if it was recently emerging from war. Countries that were further removed in time from violent conflict were better able to conduct reliable poverty surveys due to internal security, the presence of international organizations and even international peacekeeping forces in some cases, internal capacity strengthened by external assistance, and ongoing reconciliation efforts. In some cases, the poverty diagnostic benefited from qualitative and quantitative data collected by humanitarian agencies and NGOs that were active in the conflict-affected regions of a country. However, sometimes data generated by humanitarian agencies were viewed with suspicion by the government because they tended to depict poverty in ways which were contrary to the profile the governments wished to project. As governments prepare a poverty diagnostic, donors can contribute by building capacity, providing technical support, and assisting in the development of methodologies like participatory poverty assessments (PPAs) suited to conflict environments. A well-developed poverty diagnostic that is sensitive to conflict factors could, in turn, influence the prioritization of policy actions. Poverty diagnostics could be more conflict sensitive if they: · focus on a discussion of conflict-induced poverty; · systematically integrate conflict analysis tools with poverty diagnostics, i.e. go beyond cognizance of conflict factors by undertaking systematic analysis of the interrelationship of poverty, poor governance, and marginalization; · collect data with the support of donors that can provide technical expertise and support capacity building; · combine quantitative and qualitative approaches (needs assessments, PPAs) to better cover the non-income dimensions of poverty; · use data collected by humanitarian agencies and NGOs that operate in the conflict-affected regions of a country; and · draw from expertise of humanitarian agencies, NGOs, or donors to develop proxy indicators that provide a reliable picture of conflict-related poverty. 11 Policy Actions · The PRSPs of all nine countries included policy actions or programs that sought to deal with the consequences of violent conflict. The sample displayed great variation in range and scope, with countries just out of war giving the most attention to such actions. · Security issues were considered important by most of the countries, but actions tended not to be part of an integrated security strategy. · In several of the countries, policy actions were clearly informed by knowledge about conflict; but overall, the conflict sensitivity of policy actions was constrained by a weak contextual analysis of conflict factors and their link to poverty. · The countries showed little systematic attempt to address sources of conflict through policy actions. They also showed little systematic attempt to consider the potential impact of the policy actions on the conflict situation. Analysis of Context None of the sample countries appear to have used systematic analysis of conflict as a method to guide the selection, prioritization, or content of policy actions. However, knowledge about conflict was applied to some degree in several cases. The Chad PRSP, for example, included a macroeconomic analysis that gave attention to conflict factors such as oil revenue management and diversification to non-oil sector productivity, and provided a basis for appropriate action. Weak contextual analysis made it difficult to determine how programs should be prioritized and sequenced. In one I-PRSP, the policy actions covered the entire gamut of war-torn recovery needs. Individually, the proposed activities made sense, but there were too many to be taken seriously, and they were considered too vague to have much impact. Getting the analysis right is crucial for the efficacy of an action program. There was a tendency in several of the PRSPs to frame complex problems, such as criminalization of the economy and resulting insecurity, as a purely technical issue rather than as one that is intimately linked to insecurity more widely, and to the quality of governance and the potential for escalating conflict. A problem as complex as criminalization of the economy cannot be addressed by simple technical solutions such as improving the tax administration or increasing the number of customs officers. Addressing Consequences of Conflict The PRSPs of all nine countries included policy actions designed to deal with the consequences of conflict. The Burundi I-PRSP aimed to address the challenges of transition from war to peace, and was framed as one of the tools to implement relevant parts of the Arusha peace agreement. The Sierra Leone government decided to tackle the effects of conflict first, and the top-priority in its I-PRSP was to improve the security situation by demobilizing ex-combatants and retraining the government security forces. Improved security was a key concern in several other countries. In Chad, the PRSP proposed an action program to deal with land mines, disarmament, demobilization and reintegration of ex-combatants, and justice sector reform ­ but not as an integrated or operational strategy. The Cambodia NPRS12 made explicit reference to the way land mines and unexploded ordinances (UXOs) contribute to poverty and 12The Cambodia PRSP was called National Poverty Reduction Strategy (NPRS), December 20, 2002. 12 inhibit poverty reduction efforts. It included plans to restructure security and reduce defense expenditures, and to disburse the funds on sectors such as health, education, agriculture, and rural development. Several of the PRSPs considered the question of internally displaced people (IDPs) and refugees. The Bosnia-Herzegovina policy action program highlighted the fact that despite considerable progress, an estimated half of the 1995 caseload of IDPs remained reluctant or unable to return to their homes. The situation of internally displaced people was also addressed in the policy action programs in Sierra Leone, Burundi, and Georgia. Policy Actions and Conflict Prevention To what extent did the policy actions attempt to address sources of conflict? There was a close thematic overlap between the conflict factors identified in the countries and the PRSP's policy actions because many of the factors affecting poverty and those affecting conflict are closely linked, but the linkages were not systematically explored. And while much of the planned action would contribute to preventing conflict if implemented, despite not having been primarily designed for that purpose, there were also missed opportunities given the limited exploration of the linkages with conflict and consideration of those linkages in the design. Characteristics of governance were identified among key conflict factors in all the country case studies. While most of the reviewed PRSPs did not explicitly recognize the governance-conflict link ­ some perhaps because it was politically impossible, and others because conflict did not figure in the analysis ­ a few did. The Rwanda PRSP promoted more inclusive political processes to improve relations between the government and citizens. The Cambodia NPRS promoted governance measures that would have contributed to reducing the chances for future conflict escalation; however, the actions necessary to support those measures, including development of the legal and regulatory framework, were never fully implemented. Actions on economic development offer both positive and negative lessons on sensitivity to conflict. The Nepal PRSP, for example, placed strong emphasis on stimulating rural growth based on the recognition that growth previously had excluded large parts of the population from development. Given the strong linkage between the conflict and the underdevelopment of certain rural areas in Nepal, effective rural- oriented action could potentially contribute to de-escalate conflict if designed sensitively. Policy actions could be more conflict sensitive if: · selection, prioritization and content of the policy actions are systematically assessed through a conflict lens; · strong contextual analysis includes identification of key conflict drivers and their interaction with poverty, and is an integral part of the poverty diagnostic; · security issues are integrated into a cohesive strategy for improved governance of the security sector and linked with policy actions for economic opportunities; · they assess and monitor the potential impacts of policy actions on the conflict dynamics; · conflict impact assessments consider the effects of individual action programs as well as the potential impact of policies and strategies, for example, the distributional impacts of a growth strategy; · they use innovative ways to consider politically sensitive issues such as ethnicity; · they take account of volatile situations by providing flexible implementation options. 13 Institutional Arrangements · Some governments placed a high premium on developing institutional arrangements that considered conflict issues (ethnic or religious divisions, regional imbalances) by designing structures that either consciously ignored conflict factors or purposefully took them into account. · Other governments made limited efforts to consider conflict issues, reinforcing beliefs that the establishment of pluralistic values was not a priority for the government. · Even though institutional arrangements were relatively broad-based in design across the sample cases, not surprisingly perhaps, they showed mixed results in their level of devolution in PRSP preparation. · In many cases, the PRS process has resulted in enhanced cooperation among sectors and ministries. · Parallel peacebuilding processes in-country have influenced and been influenced by the PRS framework. Institutional arrangements refer to formal structures and rules that determine the design and implementation of the PRSP. Given that the PRSP is the key policy document in some countries, the manner in which power is distributed through institutional arrangements, and the structure of relationships between government and non-government actors, can either reinforce the power imbalances that contributed to conflict or seek to redress them. By establishing inclusive and broad-based arrangements, it is likely that governments will be seen to recognize and perhaps even address conflict factors related to exclusion, concentration of power, and the control of public assets by a single group. The case studies found that some governments placed a high premium on developing these types of institutional arrangements for the PRS process, while others made limited efforts to consider conflict issues, including ethnic and religious divisions and regional imbalances, in their institutional arrangements. In turn, this reinforced beliefs among certain groups and regions that the re-establishment of pluralistic values was not a priority for the government. The analysis showed, however, that in a few cases, for example Bosnia and Sierra Leone, governments were cognizant of conflict issues and designed institutional arrangements that purposefully took them into account. Even if conflict fault lines were not reflected in the design of institutional arrangements, some governments, such as Rwanda consciously decided against incorporating conflict factors in institutional arrangements because they believed that doing so would risk cementing the divides that had led to war. The cases demonstrate that at least in design most institutional arrangements for the PRS process were relatively devolutionary, with different organs of government and NGOs being accorded specific responsibilities. The cases however showed mixed results on institutional arrangements being devolutionary during PRSP preparation, with the actual level of the influence and involvement of parliament, government bodies, NGOs etc., varying across the cases. BiH is a best practice example of actions being taken to ensure that the PRSP reflected a nationwide consensus. Nepal is another excellent example of effort being devoted to eliciting opinions of local government in the eastern, central, and western provinces during I-PRSP preparations. The PRSP also highlights the country's desire for an increased role for local government in development planning. Most governments made a good start by establishing comprehensive institutional arrangements for PRSP preparation. Since most cases are now in the early stages of implementation, it is neither appropriate nor possible at this point to make an informed judgment about how governments will follow through on their commitment to devolve their institutional arrangements, in terms of decisionmaking authority and resource allocations. Reluctance of governments to be inclusive in the implementation phase of the PRSPs could be counterproductive, since the success of the PRS process is predicated on institutional arrangements that devolve power and transfer the control of resources to lower levels of government. 14 PRSP principles envisage that collective responsibility will increase as cross-sector collaboration and coordination between the center and local governments improves. The case studies indicate that the PRS preparation process resulted in enhanced cooperation among sectors and ministries, and produced a strategic and mutually-beneficial outlook on the part of ministries. Going beyond the PRSP, parallel peace-related processes in a few cases have influenced and been influenced by the PRS framework and helped to strengthen inter-sectoral and inter-ministry relations. The PRSP has promoted the inclusion of poverty-related issues into peacebuilding processes, and conversely reconciliation issues have been integrated into the PRSP ­ as seen in Rwanda and Nepal ­ resulting in both processes learning from each other. Institutional arrangements could be more conflict sensitive if they: · result from conscious design, with the incorporation of conflict factors; · stress transparency and inclusion in design and implementation; · reflect broad-based and inclusive formations, which can serve as important vehicles for cohesion and reconciliation; · devolve power and transfer resources to support the implementation of PRSPs, particularly if the government is aware that if it does not follow through with their commitment of devolution, the institutional arrangements could collapse, with irreversible damage to both the PRS and the process of peace consolidation; and · draw from ongoing peacebuilding processes and in turn strengthen them. Donor Behavior · Country ownership of the PRSP tended to increase as the country moved further out of violent conflict. · Donors tended to have unrealistically high expectations of the quality of PRSPs in post-conflict countries, given capacity weaknesses and continuing divisions among population groups. · In some cases, donors consciously refused to align their strategy with the PRSP if it did not address conflict issues as this would imply that they endorse the exclusion of conflict. · Lack of prioritization in the PRSPs made it difficult for donors to align their programs with the country program. · Donor coordination was not optimal in many cases, although harmonization efforts are on the rise. For PRS preparation, donors need to examine their own behavior, consider how they can engage with conflict-affected countries more effectively and coordinate more systematically in order to support the PRS process, i.e. not drive it. The PRSP model envisages a partnership between donors and the country. Consequently the PRSP seeks to change donor behavior on three key fronts: Donor Assertiveness vs. Country Ownership Country ownership, a key pillar of the PRSP, suggests that the PRS be representative of the needs and priorities of the diverse stakeholders in a country. The case studies indicate that countries that had not experienced active conflict for a few years and were able to successfully recognize, even if not resolve, conflict challenges were more likely to lead the exercise with donors playing a supportive role and providing technical assistance. A common complaint across the countries however was that in countries weakened by conflict and with low capacity, donors had unrealistic expectations of the PRSP and set priorities that the governments had 15 no alternative but to follow. In some cases, donors made their support contingent upon the government preparing a PRSP that was congruent with donor vision. Donors respond that in conflict-affected countries, they encourage governments to make conflict issues a primary concern; and that they have the prerogative of not aligning their programming to the PRSP if it addresses conflict issues only superficially. They further argue that the country needs to create appropriate structures to tackle conflict challenges if it wants to receive donor support. Donor Coordination Donors acknowledge that differing priorities and limited coordination in the past resulted in overlapping programs or incompatible policy actions. In some cases, tension occurred between donors that provided budgetary support to the PRSP and those that continued to support projects that were not necessarily in the priority areas identified in the PRSP. In other cases, donors did not agree that the PRSP was the most useful strategy for the country and developed other strategies that they felt were more appropriate. Cases of inter-donor disagreements persist, and tensions over how countries should be encouraged to integrate conflict issues in strategy formation continue. Yet, by and large, the cases show that a uniform voice is gradually emerging that calls for donors to harmonize their efforts on conflict and development, with a focus on how to carry out development activities in a conflict context. DEPAC (the Development Partnership Committee) in Sierra Leone, the Donor Framework Group in Georgia, and the Donor Working Group in Sri Lanka are some examples of institutionalized coordination among donors. Their programs are concerned with both poverty reduction and activities supported outside of the PRSP framework, such as the multi-donor conflict assessments. Donor Alignment with PRSP Priorities Donors should ideally realign their country strategies to conform with the priorities presented in a country's PRSP. The case studies found, however, that donors only minimally considered the PRSP in determining country assistance because the PRSP encompassed a wide-ranging set of priorities. This is unfortunate and potentially a concern also in non-conflict countries, since one of the main aims of the PRSP is to encourage donors to rethink their strategy and help address the main concerns outlined by the country. In a few cases, when the PRSP ignored conflict challenges, donors consciously decided not to align their country strategy to it because alignment would suggest that they endorse the exclusion of conflict. In these cases, donors continued to support activities that contributed to poverty reduction and conflict mitigation independent of the priorities identified in the PRSP. Donor decisions to not align their strategies to the PRSP in these situations are understandable, given their belief that aid will not be effective if the PRSP does not account for conflict concerns. Donors could support conflict sensitivity more effectively if they: · make a concerted effort to prioritize country ownership over promotion of their own priorities, if the PRSP is to be an effective framework; · strengthen the country's capacity by providing technical assistance to ensure that conflict issues are carefully considered, and that structures have peacebuilding impacts; · create an environment that enables countries to deal with conflict-related sensitivities; · differentiate between legitimate reasons for omission of conflict issues and exclusionary policies that do not justify ignoring conflict; and 16 · improve harmonization and establish formal coordinating mechanisms to take advantage of comparative strengths, and to avoid duplication of support and contradictory policy advice and incentives. Key Issues for Conflict-Sensitive PRS Development A PRSP needs to be specific to the country context and flexible in responding to changing circumstances, while taking account of potential risks. · Country specific. The most fruitful PRS design is based on a thorough assessment of the country context, including specific conflict factors. This means there is a strong need for good contextual analysis and for avoiding the mechanical use of tools and lessons. The PRS should draw heavily on in-country processes such as peace agreements, joint needs assessments, and transitional results frameworks. In many conflict-affected countries, humanitarian, recovery and development needs will overlap and the PRS needs to take this into account. · Nimble and flexible. Conflict-affected countries are often characterized by great volatility and quickly changing situations, while facing serious capacity constraints. There is a real need for a PRS framework in such countries, but the process and strategy could be structured such that design and implementation allow the countries to (i) respond relatively quickly to changing situations; (ii) be flexible in their design and implementation; and (iii) produce alternative options when changes render current measures irrelevant. While flexibility is key for the PRS in conflict-affected countries, this should not be interpreted as allowing for a laissez-faire approach but as the ability to develop unique and innovative methods. · Risks. As any activity in a conflict-affected country, the development of a PRS carries certain risks. The more carefully the process takes conflict factors into account and the more realistically the PRS content reflects the context, the better the potential risks can be managed. Risks may include instability because of unmet expectations among specific groups; the government's avoidance of prioritization and hard choices in order to maintain support from divided constituencies; insensitive treatment of divisive issues; and attention to short-term needs in ways that undermine longer-term recovery. 17 INTRODUCTION This report presents a retrospective analysis of experiences with the Poverty Reduction Strategy (PRS) process13 in nine conflict-affected countries ­ Bosnia-Herzegovina (BIH), Burundi, Cambodia, Chad, Georgia, Nepal, Rwanda, Sierra Leone, and Sri Lanka. The analysis is part of a three-year program that aims to: (i) determine how the causes and consequences of violent conflict can best be addressed within a country's poverty reduction program; and (ii) generate and disseminate lessons, good practices, tools, and guidance to increase the conflict sensitivity of the PRSP. The PRSP has become the primary tool in nearly 60 low-income countries for articulating a vision for growth and poverty reduction ­ some 15-20 of these are affected by conflict, and with other conflict-affected countries still to begin the process. The ultimate goal of the three-year program is to contribute to more effective poverty reduction in countries affected by conflict. Development agencies are increasingly recognizing the importance of viewing their interventions in conflict-affected countries through a conflict lens, as a way to more fully understand the complexities of the country contexts in which they work. As a result, increased efforts are being made to assess conflicts and identify factors that drive their escalation and de-escalation in a country. The work is underpinned by a recognition that greater sensitivity to the sources and consequences of conflict throughout the PRS process will improve both poverty-reduction and conflict-mitigation outcomes. Based on a study of PRSP experiences in conflict-affected countries, the analysis aims to distill and disseminate lessons, good practice and potential measures through which PRS can become more conflict sensitive, in order to make poverty reduction more effective in such circumstances. The report assesses conflict sensitivity of each of the following PRS components: participation, poverty diagnostics, policy actions, institutional arrangements, and donor behavior. The report does not assess resource allocations and budgets, or implementation, of the nine PRSPs. A robust assessment of conflict-sensitivity based on resource allocations would go beyond the scope of this analysis given the uneven level of such information, especially related to the I-PRSPs, and the problem of comparing resource allocations to very different types of policy actions. Examination of actual PRS implementation was not included simply because of the time-line: most of the nine PRSPs were still too new to provide a clear picture at the time of the case studies. The report does, however, refer to implementation challenges in several instances. Organization of the Report This report is divided into four sections. Section I lays out the approach to the analysis and clarifies the conceptual definitions. Section II highlights the conflict factors prevalent in the sample countries at the time of PRSP development, and then uses them to discuss the main conflict findings that echo across the cases. Section III draws from the experiences (from both desk studies and fieldwork) of the nine cases to discuss how conflict issues are dealt with in the components of the PRSP. Section IV provides pointers on how the PRS can more effectively integrate conflict sensitivity in its development and implementation. 13While PRSP originally referred to the Poverty Reduction Strategy Paper, it has now come to imply a rolling process (sometimes called the PRS process), with the Paper being a tangible product but by no means the only outcome of the process. 18 SECTION I: CONCEPTUAL FRAMEWORK AND APPROACH Box 1: DefiningConflict-affected Countries and Conflict For purposes of this analysis, conflict-affected countries are defined as those that have recently experienced, are experiencing, or are widely regarded as at risk of experiencing violent conflict. Conflict is defined as a dynamic process involving two or more societal groups pursuing incompatible objectives, leading to positive and negative changes. The manifestations of conflict vary according to the means employed. A conflict is predominantly violent when the use of violence outweighs the use of political or other means to settle differences. Violent conflicts range in intensity and scope from small events that affect only a few members of the population, to full- scale wars that result in more than 1,000 combat deaths in a year. Interrelationship between Conflict and Poverty14 This analysis is based on the recognition that conflict and poverty are closely interrelated. Poorer countries are more likely to experience violent conflict evidenced by the fact that most civil wars since 1960 have erupted in low income countries. On the other hand, conflict-affected countries tend to suffer from aggravated poverty and falling national incomes. Research suggests, for example, that countries lose about 2 percent of annual economic growth during civil wars.15 After a seven-year civil war, therefore, incomes would be about 15 percent lower than otherwise, which translates into 30 percent increase in absolute poverty.16 Similar trends can be observed internally within conflict-affected countries, with poverty tending to be higher in regions that have experienced conflict than in other parts of the country. The destruction of economic and human capital and the distortion of social capital, defined as the associations within and between groups in a society, due to conflict have serious negative impacts on the levels of poverty. Severely damaged physical infrastructure and productive assets tend to reduce employment levels and foreign investment inflows. In addition, increases in security-related spending diverts scarce resources away from productive investments and public spending. Capital flight from conflict-affected countries also rises, along with dependence on imports and foreign aid. Furthermore, conflict undermines, if not destroys, governing institutions and reduces both their ability to plan and implement policy, and to effectively uphold the rule of law. Conflict also erodes the government's tax collection capability, and the tax base more generally. Lack of revenue, in turn, has important implications for post-conflict recovery, particularly rebuilding of infrastructure and social service delivery. The intersection of these factors increases both the depth of poverty and the risk of conflict being reignited.17 14In a post-conflict society, the determinants of poverty can be classified as two types: structural issues pre-dating the conflict, which may or may not be correlated with the conflict (traditional structural-based poverty); and new issues that may have arisen as a consequence of conflict (conflict-induced poverty). While there are similarities between these forms of poverty, it is important to differentiate between them for two main reasons: (i) failure to pay special attention to poverty resulting from conflict could undermine the peace, i.e. unequal access to social services, shelter, and economic opportunities could intensify existing social divisions, thus producing more conflict; and (ii) conflict-induced poverty requires the redress of consequences of conflict, i.e. destroyed roads and infrastructure. 15 For example, the Sri Lankan PRSP, Regaining Sri Lanka, December 5, 2002, notes that a Central Bank of Sri Lanka study found that the conflict reduced economic growth by 2-3 percentage points a year. If Sri Lanka had not experienced conflict, its income level would have been at least twice the current level ­ similar to that in Thailand and Singapore. 16 Paul Collier, et al, On the Economic Consequences of War, Oxford Economic Papers, No. 51, 1999. 17Paul Collier, et al, Breaking the Conflict Trap, World Bank Policy Research Report, 2003. 19 Conflict-affected countries also face significant human capital constraints. The high cost of engaging in violent conflict leads to a decrease in public expenditures on health and education in real terms, as well as in shares of GNI. Increasing illiteracy rates and poor education due to continuous disruptions, decline in health status of the population, and poor delivery of social services also weakens the ability of the population to recover economically and move out of poverty. The fall in the number of youth, i.e. productive section of the population, as a result of death, injury or drafting into rebel groups, reduces the human capital available for livelihood development. In addition, conflicts generally create large numbers of internally displaced persons (IDPs), orphans and widows, as well as handicapped and injured persons, who have lost their livelihoods and have often been brutally traumatized. These challenges in turn tend to be exacerbated by the emigration or death of educated segments of the population. Viewed from the other direction, although poverty itself is neither a necessary nor a sufficient condition for conflict, poverty factors increase the likelihood of violent conflict. Recent analysis indicates that a country's overall risk or proneness to violent conflict depends on baseline and other factors.18 The baseline risk or vulnerability is determined by a country's recent history of conflict19 and income level. These baseline factors interact with other factors such as economic decline or stagnation, presence of a dominant ethnic group, dependence on natural resources, unequal income distribution between different groups, and presence of a weak democracy. When these factors are present in low-income countries, those countries are more likely to experience violent conflict than are middle-income and upper middle- income countries. More specifically, poverty can act as a source of conflict in three main respects. First, a combination of poverty and unequal income levels tends to result in a significant number of poor and disaffected, particularly young men, who can be easily mobilized and recruited to armed groups. Second, poor countries tend to be characterized by weak and undemocratic governance structures, which are incapable of managing conflict and preventing the onset of violence by peaceful means. Third, if a country with a large poor population is endowed with significant natural resources, rebel organizations not only find it easier to raise finances, but also are able to galvanize public support against a perceived or real injustice in revenue distribution. This suggests that when poverty is seen as the outcome of a political process by which specific groups become marginal or are deprived of their resource base, the mobilizing capacity of the group increases manifold.20 In these ways, poverty increases the risk of conflict escalating into violence and extends the duration of war. The interrelationship between conflict and poverty is often affected by group-based inequalities, i.e., those that develop between distinct social groups on the basis of their ethnic, social, regional or other characteristics. These horizontal inequalities may escalate conflict into violence when differences such as ethnicity are politicized to mobilize for political causes, and when conflict has distorted social capital by cementing associations internal to the group while depleting ties to other groups. This in turn weakens the social glue that binds the larger society together. The distortion of social bonds in the form of trust and networks negatively affect the level of growth and development, claimed to be significantly inter- linked with the intensity of social bonds.21 18This section draws heavily on recent analytical work by Collier and O'Connell (2005), Collier and Hoeffler (2003), and Chalmers et al (2005 and forthcoming). 19For countries that have experienced civil war, the likelihood of violent conflict recurring within five years is 44 percent, and thus significantly higher than in countries previously unaffected by violent conflict. 20P. Douma, Poverty, Conflict and Development Interventions in Sub-Saharan Africa, Global Development Network, The World Bank, 1999. 21Robert D. Putnam with Robert Leonardi and Raffaella Y. Nanetti, Making democracy work: Civic traditions in Modern Italy, Princeton 1993. 20 Approach This analysis examines the PRSP experience in Bosnia-Herzegovina, Burundi, Cambodia, Chad, Georgia, Nepal, Rwanda, Sierra Leone, and Sri Lanka, with a focus on how conflict factors were taken into account in designing the country's poverty reduction strategy. The findings will contribute to developing an evidence base to guide future work. The countries in the sample were selected to represent: · a geographically broad range of experience (although the absence of a case from the Latin America and Caribbean region is recognized); · varying stages in the PRSP cycle (I-PRSP, full PRSP, implementation); · different conflict dynamics (civil war, regional conflict, resource-driven conflict, ethnic conflict); and · different stages of conflict (at risk, violent conflict, post-conflict). Table 1: Sample Cases Classified Region Country Status Board Date GNI/capita HDI Conflict Status24 2003 (US$)22 200223 Africa Burundi I-PRSP January 22, 2004 90 0.339 At risk Africa Chad PRSP November 13, 2003 240 0.379 At risk Africa Rwanda PRSP August 8, 2002 220 0.431 Post-conflict25 Africa Sierra Leone I-PRSP September 25, 2001 150 0.273 Post-conflict PRSP26 May 6, 2005 East Asia and Cambodia PRSP February 20, 2003 300 0.568 Post-conflict Pacific Europe and Bosnia and PRSP June 15, 2004 1,530 0.781 Post-conflict Central Asia Herzegovina Europe and Georgia PRSP November 6, 2003 770 0.739 In conflict Central Asia South Asia Nepal PRSP November 18, 2003 240 0.504 In conflict South Asia Sri Lanka PRSP April 1, 2003 930 0.740 At risk Institutional considerations also influenced the sample selection ­ most significantly, the interest of both World Bank and DFID country teams in the analysis, as well as a desire not to duplicate the work of other ongoing research studies, notably a study conducted by the World Bank's LICUS Unit.27 22Gross National Income calculated using Atlas method, in current US$. Source: The World Bank, World Development Indicators 2005 23UNDP, Human Development Report 2004, pp. 139-142. 24The stages of conflict are fluid with countries displaying characteristics of all the three main types (at risk, in- conflict, post-conflict). They are identified here according to the conflict stage that is most prominent at the time of writing this report. 25Post-conflict suggests that the country has moved out of violent conflict; it does not mean that the country is no longer at risk of violent conflict because the underlying conflict factors as well as factors that emerged because of the conflict still may be present. 26The retrospective study included an examination of the Sierra Leone I-PRSP and preliminary drafts of the PRSP. 21 Three main sets of issues bearing on the PRS process were considered for each country: · Conflict sensitivity of the PRS. The extent to which conflict factors28 were reflected and addressed in the PRS process and in recommended policy actions; and key contextual issues that either constrained or facilitated the conflict sensitivity of the process. · Challenges created by the conflict environment. For example, the extent to which diminished capacity, entrenched power interests, lack of full territorial control, political/social taboos, lack of vertical and horizontal trust, and donor behavior constrained the PRS process. · Lessons. How the countries addressed these challenges and integrated conflict considerations into the PRS process. How donor behavior affected the process. To examine these issues, the analysis was carried out in four steps: · Literature review and discussions with experts to identify the factors of conflict in each country; · Determination of what of a conflict-sensitive PRSP would entail for each country, given its particular factors of conflict; · A desk review of each PRSP, to consider how conflict factors were reflected in the document and its preparation; · Field work to assess how conflict factors informed the PRSP preparation and implementation, what constraints were encountered, and how challenges were addressed. The field work included semi- structured interviews with government officials, Bank staff, relevant donors, and other stakeholders involved in (or excluded from) the process. What does Conflict Sensitivity entail? For purposes of this analysis, a conflict-sensitive PRSP is defined as one that (i) recognizes the key factors of conflict at all stages of the PRS process; and (ii) seeks to address those factors, where appropriate. Box 2: Defining Conflict Factors (Factors of Conflict) The term factors of conflict (or conflict factors)a is used for those features of a country's natural resource base, economy, social structures or political environment that have, or could potentially contribute to escalation or de-escalation of conflict. It may include causes of conflict, although causality may not necessarily be established, as well as consequences of violent conflict, especially if they are barriers to stabilization. a'Conflict factors' are deliberately defined broadly in this context to capture the various features of society that affect or have been affected by conflict ­ both aspects need to be taken into account in a PRS. A more detailed conflict analysis conducted to guide country programming should identify the main drivers of escalation and de-escalation, and the dynamics between the different factors at play as this would help prioritizing as well as deciding the content of action. Identifying 'drivers' is a more pragmatic and practical method than attempting to determine 'causes' and 'causality' because of the methodological problems linked with the latter. The benefits for programming, however, would be similar. 27The Low-Income Countries Under Stress (LICUS) unit has carried out a two-part review of PRSPs in LICUS countries. The first phase included a broad desk review of the PRSP approach in LICUS. The second phase involved further analysis of four country cases (Eritrea, Lao PDR, Niger, and Yemen), and desk reviews of Sudan, Liberia, and Timor Leste. A report on the findings will be available in mid 2005. 28The term, conflict factors, is used interchangeably with the term, factors of conflict. The definition is discussed in this section. 22 Recognition of conflict factors would entail that the PRSP takes them into account and integrates knowledge of them and the specific ways they link with poverty and growth in that country. A PRSP would seek to address conflict factors if it includes policy actions designed to resolve or mitigate any of the factors. It is recognized that conflict sensitivity at times may mean not attempting to address a specific conflict issue when doing so could potentially create or exacerbate tensions. Since conflict resolution or prevention is not the main purpose of a PRSP, the process should not be overburdened with unrealistic expectations. Nevertheless, to enhance the effectiveness of policy actions, the PRS should consider the conflict dynamic as much as feasible and take it into account in programming. For example, while it would be very sensible for a country to learn about factors of conflict in its poverty diagnostic, it may be unrealistic from either a political or a fiscal point of view to expect a country to address all of those factors through public actions. 23 SECTION II: CONFLICT FACTORS Country-Specific Findings Using the World Bank's Conflict Analysis Framework (CAF) as a guide, the country case-studies included an examination of the nature and manifestations of conflict factors in-country during the period of PRS development.29 The summaries of the case studies presented below highlight the background conflict factors in the countries at the time PRSP was being developed. Bosnia and Herzegovina Conflict in Bosnia and Herzegovina (BiH) followed the collapse of the state of Yugoslavia and the independence of Slovenia and Croatia in 1991. Ethnicity increasingly became the dominant discourse in politics, with leaders using ethnicity as a tool to mobilize support and gain power. In April 1992, the new state of BiH was recognized by the European Union and the United States. War broke out in 1992 and was ended in 1995 by the Dayton Peace Accords. Key conflict factors: · Stark divide between rural and urban populations in terms of income, welfare, and socioeconomic characteristics, with rural areas generally more ethnically homogenous. · Political mobilization along ethnic lines. · Lack of state capacity to protect rights of all ethnic groups. · Destruction of human/physical capital during the war. · Human rights abuses during the war, further entrenching grievances and divisions. · Refugee return and property restitution issues created by displacement during the war. · Economic tensions created by the transition to a market economy. · Corruption and a thriving black economy bred by the conflict. Burundi In the absence of institutions and processes that can resolve conflicts in a peaceful manner, political elites in Burundi have used ethnic polarization, mistrust, and fear to further their own political and economic interests. Widespread poverty, inequalities among social groups, competition over land in a small country with high population growth, soil degradation, and a low urbanization rate are key underlying sources of tension. The resulting conflicts have focused primarily on capturing or securing the state as the main avenue for power, security, and wealth. Key conflict factors: · Poor governance, clientelism, and pervasive corruption. · Widespread poverty and population pressure on scarce productive land. · Political exploitation of deep inter- and intra-ethnic divides among and within (i) clans ­ both horizontal divisions (rival lineages) and vertical divisions (castes); (ii) regions, and (iii) political and economic elites. · Weakened human, physical, and social capital from the ongoing conflict. 29 The World Bank's Conflict Prevention and Reconstruction Unit developed the Conflict Analysis Framework (CAF) to enhance the conflict sensitivity and conflict prevention potential of World Bank assistance. For more details see: www.worldbank.org/conflict 24 · Growing population of vulnerable persons and disenfranchised youth. · History of violence and impunity. · Spillover effects from conflicts in the Great Lakes region through refugee flows, arms trade, availability of resources, and rebel bases. Cambodia In the late 1960s, Cambodia plunged into a series of revolts against large landowners and the government and its economic policies, which began three decades of conflict. The Khmer Rouge regime of 1975- 1979, which committed widespread atrocities and persecuted urban populations and ethnic minorities, was replaced by a Vietnamese-backed government in 1979. Following the Vietnamese withdrawal and a peace agreement in 1991, the United Nations Transitional Authority in Cambodia helped establish a multi-party system. Since then, conflicts have been predominantly political in nature, with sporadic escalations into violence. Key conflict factors: · Social cleavages along class, ethnic, and urban/rural lines. · Unequal access to agricultural land, fishing areas, forests, and other productive resources. · Unequal distribution of benefits from economic growth, widening income disparities, profit-making and rent-seeking by privileged groups (particularly in logging). · Unemployment and limited opportunities for youth. · Militarized society and widespread violence. · Impunity for past human rights abuses complicating democratic consolidation. · Continuing challenges to the stability and equity of the political system. Chad Chad has a longstanding heritage of ethno-regional and clan-based conflict. Despite a period of progress in the 1990s, there are signs of increasing tensions and risks. Conflicts in Chad are more of a continuous presence than an event with a distinct beginning and end. They are best understood as the result of failure to build viable and responsive political institutions, due largely to a strong legacy of earlier conflicts. Armed incursions in the north, as well as in the eastern areas bordering Darfur, add to more permanent features of the situation, such as conflicts between farmers and stockbreeders, youth violence, human rights abuses, and impunity. The legacy of past conflicts also includes flawed demobilization efforts and continued militarization of society, with easy access to arms. Chad has been an oil producing country since 2003, and has put in place a complex oil revenue management program to mitigate risks of corruption. Key conflict factors: · Power concentrated in the executive, and a narrowing power base of certain clans and ethnic groups. · Threats to national and local stability from external developments, including the situation in Darfur and relations with Libya and the Central African Republic. · Clientelism and corruption, undermining rule of law and escalating local conflicts. · Persistent militarization of society, coupled with a lack of economic opportunities. · Regional imbalances and skewed socioeconomic opportunities between the resource-rich south and the northern-based political elite. 25 · Weak vertical social capital, especially between executive power and a mainly southern educated bureaucracy. · Socio-ethnic groups with competing livelihoods struggle for control of scarce natural resources, particularly fertile land and water. · Continuing concerns about respect for human rights and impunity for past abuses. Georgia The conflict in Georgia is generally understood to be a result of the wide range of demands and grievances which came to bear on the incipient state in the wake of the dissolution of the Soviet Union. The conflict is multifaceted and comprises a number of separate disputes. The two main areas of conflict are the regions of Southern Ossetia and Abkhazia, both of which have sought to secede from Georgia. In both cases, this has led to armed confrontations which ended in ceasefires. To date, however, no definitive solutions to these conflicts have been found. Key conflict factors: · Socioeconomic dislocations created by the post-Soviet collapse of the Georgian economy. · Failure of the state to put in place credible guarantees for minority rights. · Increasing politicization of ethnic identities, along with secessionist tendencies in a number of regions. · Georgia's strategic location; its transection by major commodity transportation routes, which provides incentives to fight for political control. · Foreign interference in support of different parties in the conflicts. · Widespread criminalization of the economy along with rising corruption, which further undermine effective central governance. Nepal The transition from an absolute monarchy to a multi-party democracy with a constitutional monarch combined with poor social service delivery to rural and remote areas were important contributors to the conflict in Nepal. Given the enormous poverty in the country, the Maoists were able to launch an armed struggle for a revolutionary transformation of society in 1996. In the early phases of the insurgency, the purported Maoist aim to redistribute land and promote equality received support. Their continuing reliance on violence and lack of continuous commitment to negotiations has however gradually led to increased disillusionment among the population. The conflict continues with no settlement in sight. Key conflict factors: · Wide socioeconomic disparities between the central and eastern regions and poor western territories, and between rural and urban populations. · Widespread poverty and income inequalities. · Unequal access to land, along with widespread landlessness and compulsory labor. · Social structures that impede socioeconomic mobility, restrict access to education, and limit opportunities for women, further compounding disaffection. · Political system unable to effectively address popular grievances, exacerbated by extended period of constitutional crisis and weak government. · Wide-ranging human rights abuses. 26 Rwanda In Rwanda, a cycle of violent riots, massive population movements, and increased militarization started a few years before independence in 1962. Power was reshuffled from Tutsi to Hutu elites during decolonization, resulting in massive outflows of Tutsi refugees. Raids by Tutsi militias on Rwanda triggered violent responses by the regular Hutu army against the remaining Tutsi population in Rwanda, and resulted in further refugee movements. This violence culminated in genocide in 1994 when up to one million Tutsi and moderate Hutu were massacred in 100 days, leaving about 3 million people (40 percent of the population) uprooted as IDPs or refugees. Later in the decade, border struggles intensified, particularly with the Democratic Republic of Congo. Key conflict factors: · Extreme poverty, horizontal inequalities, and population pressure on fertile land. · Political mobilization along clan and ethnic lines. · Weakness of state institutions and processes to manage conflicts in a peaceful manner. · Impunity for violence, which consolidated a culture of violence. · Conflicts in neighboring countries, which provided safe havens to various conflict groups, allowed easy arms trade, facilitated access to resources, and created refugee flows. · Dependence on natural resources, creating vulnerability to changes in commodity prices. Sierra Leone Sierra Leone's war ended in January 2002, following intervention by the United Nation's Mission to Sierra Leone (UNAMSIL) and British troops. Fighting had begun in 1991, when rural youth frustrated with socioeconomic and political stagnation were mobilized by the Revolutionary United Front (RUF). The RUF's aim was to topple the government, which had ruled for 24 years.30 Sierra Leone's war was characterized by several coup d'états, formation of civil militias and renegade soldier factions, employment of mercenaries, trade of diamonds, conscription of child soldiers, mutilation of civilians, and displacement of more than two million people. By the time the war was over, the country's infrastructure was destroyed and Sierra Leone found itself at the very bottom of the Human Development Index.31 Social cleavages between war-affected populations and ex-combatants, female heads of households and men (over property rights), young and old, and political parties had been deepened by the war. In the current post-war era, high expectations for improved governance and socioeconomic conditions as well as redress of human rights issues could be conflict-producing if not met. Key conflict factors: · Over-centralization of control over resources and political power caused disaffection in provincial areas. · Lack of accountability and capacity of governance structures. 30 In 1991 President Joseph Momoh of the All People's Congress (APC) was organizing multiparty elections, scheduled for 1992. Many people felt election plans were simply a ploy to legitimize the APC's 24-year grip on power. In 1991, before the elections could be held, the RUF invaded Sierra Leone with the assistance of Charles Taylor's National Patriotic Front (NPFL). 31The Human Development Index does not include all countries in the world, including some of the poorest; e.g., Somalia and Liberia. Source: Human Development Report 2002: Deepening Democracy in a Fragmented World, published by the United Nations Development Programme. 27 · Exclusion of youth from national and community politics. · Economic degeneration and extreme poverty. · Reverberations from political instability throughout the Mano River sub-region of West Africa. Sri Lanka While several conflicts have plagued Sri Lanka since its independence in 1948, none has impacted the country like the war between the government and the Liberation Tamil Tigers of Eelam (LTTE) on behalf of the Sri Lankan (or Jaffna) Tamils.32 The Sri Lankan Tamils are concentrated in the north and east of the country and had been demanding a separate state, to be known as Eelam. The literature usually points to 1983 as the critical juncture in Sri Lanka's ethnic conflict, although protests against the government's discriminatory policies began much earlier. From 1983 until the ceasefire in 2002, there was full-fledged war between the LTTE and the government (interrupted with periods of peace), resulting in 70,000 deaths, 800,000 IDPs, and one million refugees. The ceasefire continues to hold although with increasing tensions in recent years and has not led to a formal peace agreement. With the recent signing of the Memorandum of Understanding (MOU) for the establishment of a Post-Tsunami Operational Management Structure (P-TOMS) between the government and the LTTE, there is optimism that the peace process could resume. Key conflict factors: · Structure of government (majoritarian democracy) allowed for the concentration of political power. · Government policies on language, education, employment, and economics, intentionally or unintentionally resulted in exclusion of the Tamils. · Structural regional imbalances were worsened by inequitable distribution of resources. · Separate language of instruction and curriculum for each group fed beliefs of ethnic exclusivity. · Educated unemployed transformed into politically motivated Tamil Diaspora (based in the United States, Europe, Canada, Australia), which provided political support and funds to the movement. Support (sanctuary) also provided by kindred groups (Tamils in India) to sustain the movement. 32 Other sporadic insurrections include the Janatha Vimukthi Peramuna (radical rural-based youth movement) against the government; intra-Tamil warfare; and intermittent clashes between the Tamils and Muslims, particularly in the east, where Muslims speak Tamil but support the Sinhalese-dominated government. 28 Cross-Cutting Conflict Findings The country-specific findings demonstrate that seven broad conflict-related issues drawn from the CAF resonate across the nine cases. These include: (i) governance challenges; (ii) economic performance; (iii) regional disparities; (iv) social divisions; (v) access to productive and natural resources; (vi) militarized society, and (vii) external influences. These issues do not represent exclusive categories ­ but they interact, overlap, and vary in their degree of importance depending on the country. The following discussion attempts to highlight the main manifestations of these seven conflict-related issues as they played out in the sample country cases and seeks to examine how certain elements led to or worsened conflict conditions. Governance Challenges Most of the sample countries had state structures defined on the basis of religion, ethnicity, or clan power; in others, urban-rural divisions formed the basis of identity. As a result, in these cases, power was held by a specific group that made little effort to establish representative systems, protect minority rights, or allow for power sharing. Instead, structures consolidating security, political, and economic power were established to ensure that the armed forces, civil services, and bureaucracy were dominated by kin members. Quotas and preferential treatment were also offered in education and employment, such that they shut out opportunities for certain groups. This use of a specific identity to promote the interests of one group encouraged minorities to manipulate ethnicity or religion to construct a politicized identity as a way to gain power.33 The sample cases demonstrate that minority groups viewed the state as an instrument of domination, purposely excluding them from decision-making and access to economic and political resources. As a result, these excluded minorities challenged the authority of the state through violent means, either to overthrow it and establish yet another ethnically defined state, or to secede and establish an independent state. Some sample cases are in post-conflict stage, and at least two of them have implemented devolutionary policies to stress their commitment to restructuring government and promoting inclusion. Their performance, however, has been uneven and the center appears unwilling to devolve decision-making and share resources with lower levels of representative government, particularly if those levels are dominated by opposing groups.34 The reluctance to be transparent and inclusive is risky in these fragile situations since it raises fears that one ousted autocratic regime may have been replaced by another of similar character. Economic Performance In most cases under examination, weak governance interacted with economic factors to reduce growth and increase instability. For example, reductions in export commodity prices interacted with high dependency levels to reduce per capita incomes, often signaling the start of a negative downward spiral. Akin to the experience of many non-conflict developing countries, these governments usually did not 33While minorities are usually determined in numerical terms, minorities could be numerical majorities but are minorities in the sense they do not have the ability to make demands and suffer discriminatory treatment. 34In some situations, the central government may refuse to transfer power out of concern that devolution will lead to balkanization of the state. Thus, it further centralizes power to gain control over the country. While reasons of self- interest may provoke the center to concentrate power, it is seen by opposing groups as yet another excuse to avoid power sharing. 29 have any systematic policies in place to handle such volatile changes.35 State safety nets that could have mitigated the effects of economic contraction had usually been used to buy and retain support. Unlike non conflict-affected countries however, in a worsening economic situation, some states in the study provided protections to their ethnic or political support base to soften the impact of falling incomes. This increased the negative impact (and absolute size) of horizontal inequalities and pushed opposing ethnic or religious groups to become more politicized, thus strengthening the relationship between the promotion of economic interests and ethno-political symbolism. In other cases, stark inequalities in income and welfare between rural areas and urban areas worsened differences, thus increasing motivations for rural- based insurgencies. Most countries in this analysis have moved out of the phase of violent conflict and are coping with the economic repercussions of war. While the wars led to large increases in military spending, allocations to social expenditures (education, health) have become negligible. Youth unemployment has also increased. In many cases, lack of education and job opportunities made it attractive for youth to join the rebellion, and continuing youth unemployment can pose a serious challenge to peace, if youth once again consider violent transformation a desired alternative. Infrastructure destroyed in the war, low incomes and high inflation, enormous debts, deserted farms and plantations, increased costs of consumer goods, and poor social service delivery continue to impede the restoration of the economy. In-country Regional Disparities Certain regions of a country may have fertile soil favorable for producing export commodities, or may be blessed with valuable resources such as diamonds or oil. While these structural imbalances cannot be remedied, the cases demonstrate that the states used profits from richer areas to benefit its kindred ethnic, religious, or clan group. In cases where the more prosperous region coincided with the government's kin group, investment was concentrated in this region at the cost of other regions, thus exacerbating existing disparities. In cases where the productive land and precious natural resources were found in areas inhabited by other groups, the government tended to invest revenues from those regions to improve the social and economic infrastructures of the regions inhabited by its own group, where it enjoyed a solid support base. In some countries, regional disparities occurred along rural and urban lines, with rural areas displaying higher levels of poverty and unemployment, and urban areas being seen as benefiting disproportionately from growth and development interventions. Furthermore, in some cases, poor infrastructure and lack of social service delivery in rural areas raised the level of discontent among rural populations, who were convinced that the government was mainly interested in promoting development in urban and peri-urban areas. This discontent and frustration in turn formed the basis of a few rebellions. In some post-war situations, poor communications, transport, and infrastructure in rural areas continue to limit access to medical and education facilities, markets, and non-agriculture jobs, resulting in large-scale migration since rural areas are not seeing the benefits of the peace dividend. These type of grievances could reignite conflict. Social Divisions The case studies identified six main types of divisions: inter- and (occasionally) intra-ethnic or religious or clan fissures; socioeconomic divisions; inter-generational divides; returnees versus those who stayed behind; ex-combatants versus civilians, and gender-based divisions. No case was clearly characterized by only one type of social division, and commonly several types intersected to impact inter-group relations. 35Nepal is an interesting case that demonstrates increased growth despite conflict. According to the Nepal Living Standards Survey, growth and poverty reduction have progressed, particularly in the urban areas. These improved figures have occurred in an environment of conflict. 30 Still, in most cases, ethnic or clan or religious divisions remained the most serious division affecting the level of conflict. When such identity-based affiliations coincided with regional divides, they further increased the likelihood of conflict. Irrespective of the shape the divisions took in the sample cases, there were two disconcerting trends that underpinned them. First, social capital across divides (bridging social capital) depleted, while within- group social capital (bonding) strengthened. The politicization of identity by a group resulted in growing mistrust and hatred of the "other." Simultaneously, within the group, intensifying bonds resulting from polarized communities served as a powerful instrument of in-group mobilization. While bonding may have tightened intra-group relationships, it also had negative manifestations vis-ŕ-vis other groups that led to beliefs of superiority, hate media, and determination to dehumanize the other. In several cases, a polarized media contributed to worsening ethnocentrism through negative stereotypes and proliferation of historical untruths. Second, divides, especially those along ethnic, religious, and clan lines, led to differential access to employment and education opportunities, and unequal access to social services. In some cases, there were marked inequalities in education and employment opportunities, with those in power, without compunction, according their own clan or ethnic or religious group preferential treatment in political appointments and employment in the bureaucracy, armed forces, professional services, and private industry. Such ethnic exclusivity by the state was resisted by disenfranchised groups, which were strongly disinclined to accept the authority of the state. Access to Land and Other Resources Access to land and natural resources was a highly contentious issue that intensified tensions and strained relationships. Many of the sample cases were primarily agricultural societies where competition for control over land led to conflict. Increased stress on limited fertile areas resulted in a fall in available productive land, environmental degradation, and a decline in agricultural production. In a few cases, an expanding population was pressured to share both decreasing levels of food crops for personal consumption, and reduced revenues from primary commodities exports due to collapse in demand. As a result, poverty deepened and inequalities increased vis-ŕ-vis urban areas. Outdated land tenure systems also produced conflicts. When land reforms were poorly prepared and implemented (due, for instance, to the feudal and exclusionary nature of the rural economy), the reforms led to landlessness or indentured labor. Cases in our analysis point to evidence that the incidence of conflict correlated closely with areas where landlessness was highest. In some cases, government regulations prevented unregulated exploitation of high-value resources (e.g., timber, diamonds). While this was a sound decision in principle, the government did not apply the regulations equally, but instead awarded concessions and exemptions to certain private companies to exploit resources. This had a two-fold impact ­ it led to environmental degradation and, more significantly, destroyed the earning potential of groups that depended on these resources for their livelihoods. In other cases, both the state leadership and opposing rebel groups circumvented laws in order to capture and export valuable resources to finance the war and enrich themselves. Militarized Society The cases demonstrate that as the likelihood of violent confrontation increased, the state increased its defense expenditures and opposing groups amassed weapons. With weakened state authority, due to the collapse of the government or due to challenges from rebel groups, violent conflict ensued. The analysis revealed that once conflict erupted, there was widespread proliferation of weapons, growth of paramilitary groups, distortion of traditional power structures, human rights abuses, militarization of society, and a drop in social spending and service delivery. Under such conditions, human security was threatened, and the only safety nets available were based on ethnic or religious or clan loyalties. 31 Moreover, in the absence of viable economic alternatives, it became common for groups to resort to arms trading as a source of revenue, enabled by a system of impunity and clientelism. Most of the countries in the study are coping with the consequences of a militarized society. For example, weapons continue to proliferate and are readily available, and out of self-preservation, the states maintain high levels of defense expenditures, supposedly for security and stability. In addition, weapons as a means of conflict resolution remain an ongoing concern. Human rights abuses are continuing on both sides, and victims have limited recourse to justice. Most important, the conflict has engendered a culture of impunity characterized by a lack of accountability, and reliance on weapons to settle disputes. People continue to suffer from the psychological effects of war, including trauma and alienation. In cases where entire generations have grown up in the midst of war, lack of education and respect for human rights has made them contemptuous toward authority structures and non-violent mechanisms for addressing differences. External Impacts In many cases in our analysis, kindred groups, porous borders, inter-state political rivalries, refugees, Diaspora, and sub-regional politics conditioned the evolution of the conflict and exacerbated declining situations. These issues allowed rebel groups to use neighboring states as bases to launch attacks, regroup, and train forces. With sanctuary in and easy movement to neighboring states, groups exploited commodities such as diamonds and gold to engage in illegal trade to finance wars; increase recruitment, particularly among refugees; and purchase weapons and ammunition. Kinship ties or political rivalries between states further solidified support for these actions. In cases where rebel groups had ethnic ties, kinship provided moral support, funds, and safe havens. The Diaspora also provided political support and funds to the movement. In addition, inter-state rivalries increased political incentives to further destabilize neighbors in conflict and provide assistance to the rebel movement. In other cases, however, conflicts had spillover effects. Fear, ethnicity, flow of refugees, and arms trade, which endangered the stability of a larger region drove even relatively stable and neutral states into spiraling violence. Potential Responses to Conflict Factors Drawing from the key manifestations of the conflict-related issues discussed above, an initial judgment can be made on how these concerns could be systematically considered and integrated in the development of conflict-sensitive interventions and economic policies. Some of these concerns overlap with potential priorities in non conflict-affected countries but take on added significance in conflict countries because of existing divisions and grievances. Suggestions on how to address some of the conflict challenges are included here for purposes of illustration. They are generic, based on a cross-cutting analysis of conflict in the countries that were studied, and the actual issues in a specific country need to be based on an analysis of that specific country context. Note that several of the illustrative redress measures require substantive and long-term political effort: 32 Table 2: Potential Responses to Conflict Factors Governance Challenges Possible Redress Ethnic, religious, clan-based Implement power-sharing mechanisms that ensure state structures minority rights, balance disparities and challenge ethnocentric politics. Concentration of power and Build or recreate a functioning state with inclusive and limited representative systems representative institutions. Preferential treatment to a Implement power-sharing mechanisms that protect specific group, i.e. exclusion of minority rights and attempt to bridge ethnic, regional, or minorities religious divides. Limited state capacity Build institutional capacity to manage issues of (corruption, patron-client corruption, accountability and transparency. relationships) Displacement of the population Introduce measures that target specific vulnerable (consequence of war) groups; work toward the effective reintegration of IDPs and refugees in a manner that does not threaten the host community. Economic Performance Possible Redress Government protection of its Commit to and ensure level economic playing field. support base when economy weakens Youth unemployment (source & Focus on youth employment by developing strategies for consequence of war) youth education and employment. Poor social indicators Strengthen human capital and productive capacity of all (consequence of war) groups. Destroyed infrastructure Rebuild destroyed assets such as infrastructure and (consequence of war) housing; address challenges around the restitution of property. Weakened social service delivery Prioritize delivery of social services particularly to (consequence of war) groups that are excluded (remote, different identity) and suffered during the war. In-country Regional Possible Redress Disparities Structural disparities between Focus on systematic redistributive policies but in regions inclusionary ways, i.e. should not undermine relations between groups and between government and groups. Urban-rural disparities Provide equal social and economic opportunities. 33 Social Divisions Possible Redress Fissures along inter- or intra- Acknowledge divisions (religious, ethnic, clan) and group lines (religion, ethnicity, strengthen bridging social capital; ensure level playing clan) field on all sectors. Politicization of identity Foster trust and understanding across groups, and build social capital between and among groups. Ethnocentric approaches (hate Promote an independent and responsive civil society and media, polarized civil society) media, and encourage dialogue and consensus through inclusive participatory approaches, reconciliation and conflict-related trauma. Access to Land & Other Possible Redress Resources Limited fertile lands in Support non-land productive enterprises including rural agricultural societies leads to industries. increased rural competition and inequalities Outdated land tenure systems Support land reforms that take account of conflict sensitivities. Unregulated exploitation of Regulate and manage use of resources. resources, often as source of war financing Militarized Society Possible Redress Weakened ability by state to Rebuild or strengthen state security system without provide security (source & reducing social and productive budgets. consequence of war) Arms Market Strengthen control of borders and illegal arms trading. Child soldiers, unemployed Focus on youth education and employment. youth constitute base of rebel movement Regional Instability Possible Redress Inter-state political rivalries Promote regional cooperation. (source & consequence of war) Easy movement of rebels and Work towards regional cooperation and disarmament; weapons (source and strengthen border control. consequence of war) Politicized Diaspora Encourage Diaspora to serve as peace-builders. Some conflict-related challenges are closely related to poverty reduction and development, and it is natural to address them within the framework of a PRS. However, it may be beyond the purview of the PRS to directly deal with all aspects of a militarized society or external pressures. Moreover, for PRSs prepared in ongoing conflict situations, success in tackling conflict-affected poverty problems are predicated on finding a binding political resolution to the conflict. That said, while PRSPs do not normally include the resolution of human rights concerns, flow of weapons, or sub-regional instability, ideally a PRSP should account for these factors in the design and implementation of policy actions, and encourage prioritization of programs that strengthen resilience to the negative impacts of these issues. It is equally important that the PRS process be inclusive and aware of the importance of balancing conflicting priorities across social and ethnic divisions. Issues of marginalization and exclusion tend to be magnified in conflict environments, such that even a PRSP based on relatively broad participation risks 34 failing if groups feel their views have not been incorporated. Thus, ideally the process resulting in the PRSP should be characterized by transparency, consultative and representative discussions, and carefully designed feedback mechanisms with validation exercises. This is the case in any society, but the chances that the country may face escalating conflict makes this an issue with very high stakes. 35 SECTION III: KEY LESSONS What is Realistic? The degree to which the PRSPs consider conflict factors needs to be understood against the background of the situation in which they were developed. · The goal of the PRSP is to achieve poverty reduction and growth, not conflict prevention or mitigation. Conflict issues would thus be dealt with as part of the poverty reduction and growth action programs, not as a separate objective. · The PRSP is a government-led, country-owned strategy. Political willingness and feasibility influence the determination of priorities. The degree to which conflict has been taken into account is thus driven by each country's specific situation and demands. · The PRS process is political as well as technical, and PRS action plans are developed within the parameters of domestic and, in some cases, foreign policy. This would account for political limitations as well as political opportunities. · The situation in a conflict-affected country may require conflict issues to be discussed and presented in specific ways to avoid exacerbating tensions. This may mean that analysis and actions related to conflict are, in some cases, dealt with implicitly rather than in explicit terms, and may account for the lack of specific reference to ethnicity or other divisions in some countries. · State capacity suffers in most conflict-affected countries, and the country's ability to plan and implement complex programs may be poor. This would often manifest itself as weak prioritization, as disconnect between overall goals and specific action, and as slow implementation. In some cases, however, such problems may be a result of the need to satisfy divided constituencies. This section considers the main components of the PRSP in the nine countries, in terms of both process and content. The objective is to assess the extent to which different aspects of the PRS process ­ participation, poverty diagnostic, policy actions, institutional arrangements and donor behavior ­ were sensitive to conflict, and to understand what facilitated and hindered efforts to identify and address conflict factors. Participation36 Summary Findings · In countries with traditions of limited public participation, and relative to their starting point, engagement with populations on poverty issues through the PRS process opened up space for greater inclusion and domestic accountability. · PRS formulation generally took place in environments of low state capacity and legitimacy with weak links among political power, bureaucracy, and conflict-affected populations. Prospects for the PRSP becoming a vehicle for stabilization increased with the government's ability to demonstrate that poverty reduction efforts are for the benefit of all citizens. 36A recent update of World Bank policy with respect to adjustment lending is accompanied by a good practice note on supporting participation in development policy operations. Participation is defined in the note as "the process through which stakeholders (those affected by the outcome of the reform...) influence or share control over setting priorities, making policy, allocating resources, and ensuring access to public goods and services." 36 · Limited effort was made during the PRS process to diversify the means and geographic span of communication with conflict-affected groups on PRS methods and goals. · Managing expectations was addressed most effectively through the institutionalization of dialogue, and by ensuring that the concerns of conflict-affected groups were taken up by policy planners. The effort was less successful where the participatory process was misperceived to be a one-time consultation exercise with no expected impact at the policy level. Poverty Reduction Strategies are expected to be built upon a country-owned development model that is underpinned by broad and deep stakeholder participation. In order to achieve and sustain such participatory processes for the long term, governments are encouraged to incorporate the views and priorities of stakeholders including civil society, parliament, social partners, vulnerable populations, and the media into the design, monitoring, and implementation of the strategy. It is expected that poor people and their legitimate representatives will be increasingly included in debates on policy choices, and that public actions prioritized in the PRSP will be developed in light of their concerns. Citizens and external partners anticipate that the PRS process will continue to open up space for participation throughout the policy cycle, leading progressively to a greater impact of stakeholder inputs into strategy implementation, monitoring, and adjustment. The principles of sustained participation, domestic accountability,37 and social inclusion have special relevance for conflict-affected societies, where transparent policymaking and attention to inequality are likely to be limited during and immediately following periods of violence, particularly in cases where inequalities were a key source of conflict. In addition, given that failure to address the concerns of certain sections of the population through legitimate processes is often a key precursor of conflict, efforts to achieve consensus-based policymaking is a critical component of any peacebuilding program. However, opening up policymaking entails real risks, most obviously that the government will be unable to manage multiple demands effectively, and that expectations will be unmet, causing disillusionment, withdrawal from the political process, and damage to the credibility of representative institutions. Expectations of what can be achieved through participatory processes therefore need to be tempered with realism and seen in light of modest increments of change over time, as conflicting parties begin to build trust and a common desire to reduce poverty for the country as a whole. Stakeholder participation and society's appreciation of how it is conducted presents challenges in any country. Ensuring that participation contributes to stability and growth in conflict-affected countries presents one of the more difficult steps in the PRS process. Experience from the sample countries revealed a range of complex conditions that posed obstacles to participation. Some of the key challenges with respect to participation are summarized below. 37 This can be loosely described as accountability of the executive branch of government, which has been assigned by the citizenry, legislature, and judicial branches with the governance and stewardship of institutions, resources, and service delivery. The term also implies that stakeholder groups and individual citizens should be accountable to one another and to society, abiding by the laws that govern them. 37 Challenges to Participation in Conflict-Affected Countries · Absence of the State. Certain parts of the country are beyond the territorial control of the government; there is little tradition of or resources for service provision to those areas. · Limited in-country experience with the country-driven model, and a belief that the PRSP is a technocratic exercise undertaken largely to fulfill donor requirements. · Limited capacity and tools at the disposal of authorities to structure or facilitate stakeholder engagement in unstable or hostile areas. · Timing. The sequencing of the PRS process with HIPC decision points and other international commitments, makes it difficult for some governments to ensure meaningful participation, while the country is at war (Sierra Leone I-PRSP) or in a state of emergency. · Poor outreach/communication strategy. There is an assumption that social divisions resulting from conflict preclude the use of mechanisms to engage certain populations located outside the authorities' sphere of influence. · Low incentive to engage with communities or groups considered to have a conflictual relationship with the authorities. · Limited voice among certain sections of the population, reinforced by legacies of obedience, subordination, and respect for hierarchy. In all of the sample cases, participatory processes undertaken in the PRS context testified to a new departure for the authorities and the population. Countries employed a variety of participatory approaches as part of PRSP formulation. These were widely seen as positive developments in their own right, as observers felt that the decision to engage the population opened up valuable space for information dissemination and, in some instances, debate on vital socioeconomic issues. Country experiences bore out the importance of PRS participation as a starting point for engaging populations previously unable to voice their basic concerns and poverty reduction priorities. They also reflect countries which were setting out from traditions of limited participation in public debate. Sample countries concentrated their participation efforts largely at national, and to a modest extent, regional levels, typically through a series of workshops covering various sector issues and involving large numbers of participants. In Cambodia, a total of 650 participants engaged in 3 national workshops, in addition to a number of regional events. In Bosnia and Herzegovina, it is estimated that more than 300 roundtables were convened and several thousand people consulted over the course of PRSP preparation. Active participation at local government and community levels was generally weaker, but some emerging good practice was found. In Rwanda, conflict mitigation mechanisms were incorporated into cellule- level38 consultations, which linked participatory rural appraisal methodologies (PRA) with the traditional concept of ubudehe,39 in order to root participation in traditional processes. During the preparation of the Burundi I-PRSP, the authorities made efforts to overcome problems of representation by sending members of parliament and senators to districts to inform populations about the PRSP and elicit local input. 38The cellule is the lowest level of administration in Rwanda. The country has 9,165 cellules, each comprising about 200 households. Above the cellule level are 1,545 sectors, 106 districts, and 12 provinces. 39Ubudehe is a participatory method developed from a Rwandan traditional concept of working collectively in agriculture. Ubudehe took place when all social and ethnic groups prepared the fields together before the rains came and the planting season arrived. It now refers to a participatory process of budgeting and planning at the village level, whereby citizens themselves allocate decentralized funds according to village priorities. 38 Compared to generally stable PRS country environments, the weakened institutions and social divisions in conflict-affected countries often resulted in authorities relying more heavily on peace accords and donor support to jump-start development and social cohesion building. Although there was modest involvement of other stakeholders such as civil society, parliament, and social partners, sample countries primarily defined participation as engagement with civil society organizations (CSOs), with participation consisting largely of consultation workshops and focus groups. CSOs frequently struggled with capacity limitations, which hampered their ability to analyze and respond to complex policy matters or voice their constituents' views effectively. In some instances, the predominant role of international NGOs tended to undermine the involvement of national organizations. One of the more encouraging participatory processes from the sample ­ the case of Bosnia and Herzegovina ­ is summarized in Box 2 below. Box 3: Bosnia and Herzegovina PRSP Participatory Process Civil society tradition was limited in pre-war BiH, but the massive influx of aid to the region during and after the war, and a recognition of the need to build civic institutions, resulted in a proliferation of new organizations in the mid-1990s, many of which delivered services previously provided by the state. Few of these organizations were skilled in macro policy, or mastered the advocacy role of civil society in the policy cycle. By all accounts, the participatory process in BiH was impressive. More than 300 roundtables were convened and several thousand people consulted over the course of the exercise. Stakeholders, including international and local NGOs, Youth Coordinating Committee, and social partners, commented on all drafts of the PRSP. Regions and ethnicities were included in the process, and the strategy was reviewed and approved by the parliament. Serious efforts were made to incorporate conflict-affected voices early in the process, including refugees and IDPs, women, and single-headed households. Thematic consultations considered the rights of war veterans, the families of fallen soldiers, missing persons and military personnel disabled in the war, refugee return, corruption, human rights, creation of a single economic space, unemployment, and rural poverty. NGOs rated the process as highly participatory, and one which they committed to remain engaged with over time through monitoring. One leading NGO stated, "The creation of BiH's PRSP has allowed citizens to directly participate in its drafting. For the first time, citizens have been asked to express their opinions directly through public debates. It has thus presented unique opportunities not only for improving the lives of the poor residing in [our] country, but for empowering citizens to take part in the creation of future BiH priority actions and policies."a In describing the BiH experience, it is important to acknowledge legitimacy questions surrounding NGO participation in the process, as many of them were closely allied to international donors, and thus perceived to be proxies for donor opinions. Notwithstanding this fact, the participatory process contributed to the strengthening of governance processes in BiH. The PRSP Coordinator's office stated, "The government was able to learn more about the problems faced by the most vulnerable categories of the population, while on the other hand, civil society representatives were able to mobilize and establish dialogue with government representatives."b aSee www.bospo.ba/eng/prsp.htm for more information bHow the Development Strategy of BiH ­ PRSP ­ Came to Be, available at www.bih.prsp.info/knijga/ZA-WEB/english%2003/prsp%20eng/4.pdf Capacity limitations were also discerned on the part of governments managing the PRS process. These included limited in-country experience with the country-driven model, and a belief across the less- integrated parts of governments that the PRSP was a technocratic exercise undertaken to fulfill donor 39 requirements. While these challenges are common in developing countries, participatory provisions were particularly uneven in sample countries that had areas beyond the territorial control of the government, and consequently, where there were limited resources and few mechanisms for public service provision. Authorities often had limited tools at their disposal to structure stakeholder engagement in unstable areas, and tended to assume that the social divisions resulting from conflict precluded engagement with populations located outside their sphere of influence. This is not to overlook the fact that even in areas where the government had control, groups were often selected on the basis of political affiliations. Furthermore, in some of the sample countries, the PRSP lacked the ownership needed at the highest level of government to ensure institutionalization of and sustained participation in the process. In this environment, certain sections of the population had very limited voice in the process, a shortcoming reinforced by the limited practice of participation prior to the launch of the PRSP. Conflict Sensitivity of the Participatory Process Given the diverse range of country settings, there was considerable variation across the sample in the extent to which participatory processes recognized key factors of conflict, and could therefore be said to be conflict sensitive. The potential for the PRSP to become a vehicle for stabilization depended on the government's ability and commitment to involve a broader group of stakeholders and demonstrate that poverty reduction efforts are for all citizens, irrespective of ethnicity, religion, or region. The sample suggests a spectrum in the degree of conflict sensitivity in PRSP formulation. Bosnia- Herzegovina and Rwanda employed participatory processes that specifically addressed re-integration of war-affected groups, war veterans, families of fallen soldiers, missing persons, and disabled citizens. In other countries, the participatory process was partially sensitive to conflict factors, but the process was not implemented in conflict-affected areas of the country. In Georgia, the government addressed IDP/refugee issues in areas under its geographic control through a general, non-conflict-specific poverty lens. In the Chad PRSP, and the Burundi and Sierra Leone I-PRSP preparations, war-affected populations were identified and consulted, but there is little evidence of their voice in the final document, and little indication of how their involvement can be sustained over time. In another group of countries, the PRS document refers to key conflict factors in the country, some of which may have been based on pre-PRSP analysis, although they are not identified as such. However, the document stops short of defining the role of participatory processes in PRS implementation, or the ways that participation might contribute to addressing conflict factors or building better governance. In some country cases, there was evidence of authorities beginning to use the PRSP as an opportunity to promote dialogue with war-affected groups. For example, the Bosnia-Herzegovina, Sierra Leone, and Rwanda PRSPs, and the Burundi I-PRSP, presented a strong rationale during the PRS design phase for addressing the effects of conflict through broad-based participation. In each case, the decision appears to have provided an entry point for addressing inclusion issues. The Rwanda PRSP (Box 3) is an example of a conflict-sensitive PRS process that was reinforced by a well-structured approach and comprehensive institutional arrangements. Box 4: Rwanda: Conflict Sensitivity in the PRS Process The PRSPa described the strategic importance of participation as follows: "Broad consultation is particularly necessary in a country emerging from conflict, and cooperation on local questions can help direct attention away from the divisive national policies of the past." Rwanda's PRSP is an encouraging example of an effort to ensure social inclusion, dialogue, and consensus building in a PRSP context. If one adopts a static view of consultations, the constraints that the country's unique conflict placed on open and free debates were overwhelming. A static view might therefore suggest that PRSP 40 consultations did not contribute much to dialogue and consensus building. It might be more appropriate, however, to take a dynamic view of PRSP-engendered participation in relation to its starting point. In Rwanda, citizens had rarely been involved in decision- making, and social exclusion was one factor underlying the conflict. PRSP consultations therefore were a key step toward sensitivity to conflict. The authorities demonstrated a firm political commitment to consultative processes; consultations were consciously used to prevent further outbreaks of violence; and participatory processes included war-affected groups (although especially vulnerable groups may not have had adequate voice). Further, engagement with citizens was institutionalized as a tool for enriching political processes. There were, at the same time, elements of tradition that constrained free and open debate. Mistrust of authorities and lack of capacity, particularly on the CSO side, were additional factors. The dynamic view acknowledges that because of the longstanding social divisions in Rwanda, it will take time to reach the requisite level of trust and incentive for open and free debate. Field work conducted for this case revealed the special importance of country ownership in providing an enabling environment for transparency and meaningful conflict sensitivity. aRwanda PRSP, June 30, 2002. In other cases, fewer efforts were made to encourage opposing groups to be part of the participatory process in the country as a whole, resulting in limited ownership of the PRSP, at least for the first round. Translating the identification of war-affected and vulnerable populations into a socially inclusive process proved to be challenging for most governments. Authorities may have had low incentives to engage with communities or groups with which they were in conflict. For the longer term, however, it is important to highlight the rolling nature of the PRS process, which will evolve in response to the building of trust between the governors and the governed, and in tandem with the political and foreign policy developments which lie beyond the scope of the PRSP. A crucial element of the participatory processes in the sample countries was the extent to which strategic communications were used to engage conflict-affected populations. Authorities in Bosnia-Herzegovina, Rwanda, and Sierra Leone promoted participation of marginalized and war-affected groups through print and radio media. A key challenge in this respect, however, was to go beyond one-way dissemination of PRS-related information in order to sustain constructive discussion on policy issues. Some cases revealed that participants were reportedly unaware that they were expected to provide input to the PRSP. In this connection, the use of languages widely spoken by target populations was of critical importance. Limited effort was made to provide documentation in languages other than English,40 including ethnic minority languages, and pictorial or local language brochures on PRSP goals and processes were rarely available. A second key aspect of communication related to the management of expectations, so that exchanges with the population could encourage stability and help prevent the recurrence of violence. This was addressed mostly effectively through the institutionalization of dialogue and by ensuring that the concerns of conflict-affected groups were taken up by policy planners at the national level. The Bosnia-Herzegovina, Georgia, and Rwanda strategies considered the PRSP to be a vehicle for social cohesion and institution building, and appear to have managed expectations effectively. In other cases, the participatory process was frequently misperceived to be a one-time consultation exercise that would have no impact at the policy level. 40Some PRSPs were presented in French, Spanish or Portuguese. 41 Poverty Diagnostic Summary Findings · Poverty diagnostics presented a multi-dimensional view of poverty that recognized not only the income dimension but also social, human, and structural dimensions of poverty. · Conflict issues were considered in the poverty diagnostic, but the discussion of the manifestations of conflict were not the result of systematic conflict analysis; there were only limited efforts to explore how factors of conflict and poverty drive each other. · Genuine political constraints prevented governments from explicitly addressing poverty-conflict linkages. · Lack of capacity and paucity of up-to-date socioeconomic information were major practical constraints to the poverty diagnostic. The PRSP is expected to present a comprehensive diagnostic that sets out the determinants of poverty as a basis for developing a suitable program of actions. In a country affected by conflict, it may be useful to go beyond traditional structural-based poverty to develop an in-depth understanding of conflict-induced poverty. This will contribute to the development of a conflict-sensitive poverty diagnostic that will enable the specificities of conflict-induced poverty to be reflected in policy actions, a subject discussed in the following section. Most governments in the sample went beyond the basic needs approach and put forward a multi- dimensional view of poverty that recognizes not only the income dimension, but also social, human, and structural dimensions of poverty. In these countries, in addition to the structural poverty factors that afflict many developing countries, poverty manifests itself in new ways due to the devastation of physical, human, and social capital. Key manifestations of conflict-induced poverty that contribute to worsening poverty rates and insecurity include: (i) weakness of the state and its inability to meet the needs of the poor; (ii) destroyed infrastructure and decreased production, affecting livelihoods; (iii) climate of insecurity deterring investors; (iv) exacerbated regional disparities and rural-urban divisions; (v) deteriorated education and health facilities; (vi) increased overlap of ethnic and economic divisions; and (vii) increased numbers of vulnerable people, including orphans, widows, handicapped, IDPs, and refugees. Given that the PRSPs were developed in conflict-affected environments, the poverty diagnostic should identify these types of conflict factors. It appears that although most PRSPs considered conflict issues to some degree in the poverty diagnostic, the ramifications of being conflict-affected were not informed by any systematic conflict analysis. In the I-PRSPs of Burundi and Sierra Leone, there is consensus that conflict aggravated poverty, and the poverty diagnosis highlights the consequences of conflict. In Rwanda, not only did the poverty diagnosis stress the impact of conflict on poverty; it also promoted the potential for peace-building through links with important institutions and processes. The Nepal PRSP views the conflict as the manifestation of not just political but economic and social grievances, and admits that the nexus of poverty, poor governance, and marginalization needs to be urgently addressed. The Sri Lanka PRSP discusses poverty in the North East, noting that it is qualitatively different than in the South. It assesses poverty in the North East quite effectively, and systematically outlines the phenomenon of conflict-affected poverty. In Chad, the consultations during the PRS informed the poverty diagnostic on local conflicts, landmines, and governance. In another significant case, the poverty diagnostic recognizes that conflict has led to poverty and that the high risk of violence continues to inhibit development. That diagnostic, however, seems to focus on the economic impact of conflict, e.g., economic collapse increasing poverty, while only superficially touching 42 on issues that are likely to increase the country's vulnerability to conflict, such as social dislocation, ethnic divisions, displacement, and criminalization of the economy. Still, even though the poverty diagnostic does not show an understanding of the initial determinants of conflict or its social repercussions, it has managed to recognize the poverty impacts of conflict that continue to undermine peace and poverty reduction efforts. Challenges to Poverty Diagnostics in Conflict-Affected Countries · Weakened capacity. Severely weakened capacity makes it difficult for the country to undertake a systematic conflict analysis and discuss the manifestations of conflict-induced poverty for a country. While this is a significant constraint, governments have to develop ways of better using the capacity at hand to achieve a deeper understanding of the dimensions of conflict-related poverty. · Lack of data. Incomplete and outdated data as well as lack of access to certain areas reduces the government's ability to prepare a comprehensive poverty diagnostic. In these circumstances, it may be beneficial for governments to collaborate with donors, NGOs, etc. to develop innovative qualitative and quantitative methods to collect data and understand the special needs born out of conflict-affectedness. · Political sensitivities. It may be politically difficult to address some conflict factors (for e.g., regional disparities, unequal access) particularly when a country has reconciliation as its objective. The question therefore is how these issues can be addressed effectively and sensitively. Evidence from the poverty diagnostics across the nine cases demonstrate that the diagnostics considered aspects of conflict, thus introducing some level of conflict cognizance into the diagnostics, albeit not systematically. There was, however, limited effort to analyze how the conflict dynamic informs the poverty diagnostic, or to explore precisely how factors of conflict and poverty drive each other. While this weakens a comprehensive analysis of conflict and poverty, it seems that some governments faced genuine political constraints that prevented them from explicitly addressing poverty-conflict linkages. Some countries were suffering from the vestiges of violent conflict, which made it politically untenable to discuss conflict factors in the poverty diagnostic. Since regional disparities, unequal access, and ethnic relations were common factors that led to conflict, highlighting them in a poverty diagnostic could have been counterproductive. Many governments were trying to rebuild relationships and trust between and among groups. In such circumstances, it would have been politically insensitive to discuss conflict factors in a way that showed certain groups in poor light. If any reference to conflict had been considered biased by a certain section, it could have threatened the progress toward peace. There are also practical constraints that hinder the systematic analysis of conflict and its relationship to poverty in conflict-affected areas of a country. Foremost are capacity constraints and the lack of recent and comprehensive socioeconomic data, which undermine the government's ability to undertake effective poverty analysis. The countries in the study suffered from serious data problems, some more than others. In a few cases, the PRS process began while the conflict was still active. In one of those cases, the government prepared the I-PRSP while rebels controlled more than two-thirds of the territory. New data could not be collected due to the security situation, and the I-PRSP therefore used pre-war data to create a poverty profile of the country. However, the depth and characteristics of poverty could not be assessed with any accuracy because the most recent nation-wide statistics were from before the war, and poverty had changed considerably in extent (exacerbation of poverty indicators) and manifestations (impact of conflict) during the war years. In contrast, countries that were relatively further removed in time from violent conflict were better able to conduct reliable poverty surveys due to internal security, the presence of international organizations and 43 even international peacekeeping forces in some cases, internal capacity strengthened by external assistance, and ongoing reconciliation efforts. In BiH, for example, the poverty diagnostic provided disaggregated data by geographic unit; when cross-referenced with demographic data, some interesting findings on the relative welfare of different ethnic groups emerged. The geographic analysis was complemented by assessments of rural versus urban and peri-urban poverty; and of specific war-affected populations, including IDPs. Such a comprehensive analysis of poverty provided a compelling picture of poverty in-country. In some cases, the poverty diagnostic benefited from qualitative and quantitative data collected by humanitarian agencies and NGOs that were active in the conflict-affected regions of a country. In Cambodia, for example, the poverty diagnostic was informed by data from a Participatory Poverty Assessment, Demographic Health Surveys, National Survey on Public Attitudes towards Corruption, and the National Population Census. However, sometimes data generated by humanitarian agencies were viewed with suspicion by the government because they tended to depict poverty in ways which were contrary to the profile which governments wished to project. Donors can contribute to the development of a conflict-sensitive poverty diagnostic in a conflict-affected environment by building capacity, providing technical support, and assisting in the development of methodologies like Participatory Poverty Assessments suited to conflict environments. They can impress upon the government that the nature of their conflict-affectedness needs to be captured in the poverty diagnostic, otherwise important opportunities for poverty reduction may be missed. A well-developed poverty diagnostic that is sensitive to conflict factors could, in turn, influence the prioritization of policy actions. Policy Actions Summary Findings · The PRSPs of all nine countries included policy actions or programs that sought to deal with the consequences of violent conflict. The sample displayed great variation in range and scope, with countries just out of war giving the most attention to such actions. · Security issues were considered important by most of the countries, but actions tended not to be part of an integrated security strategy. · In several of the countries, policy actions were clearly informed by knowledge about conflict; but overall, the conflict sensitivity of policy actions was constrained by a weak contextual analysis of conflict factors and their link to poverty. · The countries showed little systematic attempt to address sources of conflict through policy actions. They also showed little systematic attempt to consider the potential impact of the policy actions on the conflict situation. Priority policy actions constitute the heart of a PRSP: they are designed to increase sustainable growth and reduce poverty. It is common to delineate four key areas of content41: (i) macroeconomic and structural policies to support sustainable growth in which the poor participate; (ii) improvements in governance, including public sector financial management; (iii) appropriate sector policies and programs; and (iv) realistic costing and appropriate levels of funding for the major programs. The following section discusses the extent to which the policy actions outlined in the PRSPs have been informed by recognition and assessment of conflict factors, and how the identified conflict factors are 41Jeni Klugman (ed.), A Sourcebook for Poverty Reduction Strategies, Volume 1: Core Techniques and Cross- Cutting Issues, World Bank, Washington, D.C., 2002. 44 dealt with in the countries' policy actions. The discussion also considers the potential impact that policy actions outlined in the sample PRSPs may have on the conflict environment. Analysis of Context The policy action program needs to be guided by a thorough understanding of the operating environment, and in the case of conflict-affected countries to include an analysis of conflict and how it interacts with poverty. Ideally, such an analysis would determine the key drivers of conflict, and assess the ways they constrain or provide opportunities for policy actions that promote growth and reduce conflict. To what extent have the sample policy action programs been guided by an assessment of conflict? As noted in the section on the Poverty Diagnostic, none of the sample countries appears to have systematically analyzed conflict as part of the poverty analysis. It is therefore not surprising that assessment of conflict was not used as a method either to guide the selection and prioritization of policy actions, or their content. This does not mean that knowledge about conflict was not applied. The Chad PRSP, for example, included a thorough macroeconomic analysis that gave attention to conflict factors such as oil revenue management, and diversification to non-oil sector productivity. This shows that the government was aware of the potential adverse effects of oil revenues, as well as the temporality of these resources (expected to decline drastically by 2015). On this basis, the PRSP emphasized the need to strengthen the non-oil economy and avoid dependency on petroleum resources. One apparent weakness in several of the policy action programs was a lack of prioritization. This problem would stem, at least in part, from weak analysis of the context ­ the purpose of analysis would not only be to identify opportunities and content of action but also determine how programs should be prioritized and sequenced. In one I-PRSP, the policy actions covered the entire gamut of war-torn recovery needs. Individually, the proposed activities made perfect sense in a conflict-affected environment, but they were too many to be taken seriously, and they were considered too vague to have impact. Getting the analysis right is crucial for the efficacy of an action program. One of the I-PRSPs in the sample was developed with the belief that key underlying factors of the conflict would be taken care of if the effects of the war were addressed. Tackling the immediate effects of the war made perfect sense in the recovery phase, but did not address the underlying sources of conflict, which are still very much alive. Related to this, there was a tendency in several of the PRSPs to frame complex problems, such as criminalization of the economy and resulting insecurity, as a purely technical issue rather than as one that is intimately linked to insecurity more widely, and to the quality of governance and the potential for escalating conflict. A problem as complex as criminalization of the economy cannot be addressed by simple technical solutions such as improving the tax administration or increasing the number of customs officers. Challenges to Policy Actions in Conflict-Affected Countries · Contextual analysis. Thorough assessment of the operating environment, including of conflict factors, is essential for effective policy actions ­ the capacity for such analysis in most conflict-affected countries. · Consequences of violent conflict. Addressing such consequences may need to be prioritized before other actions designed to raise growth and reduce poverty. Priority actions may thus include rebuilding destroyed or damaged infrastructure, demobilizing combatants, reintegrating IDPs, or removing land mines from agricultural land. · How best to prioritize? This is especially tough if the needs are great, the capacity and resources are limited, and the government is under pressure to satisfy multiple and divided 45 constituencies. · Preventing escalation of conflict. There may be good reasons for prioritizing issues that, if addressed, may help mitigate escalation of conflict and thus ensure a longer-term poverty reduction impact. For example, improved public accountability and transparency in the use of oil or forest revenues would help ensure that these revenues are used for the good of the general public; and by so doing, eliminate a potential source of violent conflict.42 · Unintended impact. Policy actions can theoretically have a zero or neutral effect on potential escalation or de-escalation of conflict, but in most cases they will impact conflict positively or negatively, and often strongly · Security and political stability. Policy actions can only succeed in an environment that is conducive to reform. This would include a reasonable level of security and political stability, as well as some access by the government to all parts of the country, including those dominated by rebel groups. Addressing Consequences of Conflict The PRSPs of all nine countries included policy actions designed to deal with the consequences of conflict. Sierra Leone and Burundi, for example, both developed their I-PRSPs while still in the early stages of a transition from war to more peaceful conditions. The Burundi I-PRSP explicitly aimed to address the challenges of transition from war to peace, and was framed as one of the tools to implement relevant parts of the Arusha peace agreement. Its proposed policy actions had the ambitious goal of restoring poverty and social indicators to their pre-crisis level by 2010 and 2015.43 The Sierra Leone government made a conscious decision to tackle the effects of conflict before anything else, which was perfectly understandable in the war-recovery phase. The top priority in the I-PRSP's action program was to improve the security situation by demobilizing ex-combatants and retraining the government security forces. The second goal was to kick-start the economy by increasing public expenditure in social sectors and support IDPs and other vulnerable groups. The third goal was to improve access by the poor to basic education and health care. These goals were reconfirmed three years later in the full PRSP, which stressed the importance of completing the post-conflict recovery process, and identified the need to address the conflict-induced aspects of poverty as a basis for further development and poverty reduction.44 In addition to Sierra Leone, improved security was a key concern in several other countries. In Chad, the PRSP proposed an action program to deal with land mines, disarmament, demobilization and reintegration of ex-combatants, and justice sector reform ­ however, not as an integrated or operational strategy. The Cambodia NPRS45 made explicit reference to the way land mines and unexploded ordinances (UXOs) contribute to poverty and inhibit poverty reduction efforts. It argued that land mine clearance would not only reduce the number of casualties, it would improve access to services and infrastructure such as water, schools, roads, and hospitals, as well as to additional land for settlement. The strategy also planned to reduce expenditures for security and defense, and to restructure by removing ghost workers and soldiers from the payroll, and disburse the funds on priority sectors, including health, education, agriculture, and rural development. 42For a discussion on this, see Ian Bannon and Paul Collier, eds., Natural Resources and Violent Conflict, The World Bank, 2003. 43Both the I-PRSP and the Arusha accord recognize that although some of the Millennium Development Goals (MDGs) are realistic to accomplish by 2015, decades of violent conflict have rendered others unrealistic. 44 During fieldwork, Sierra Leoneans across the country emphasized the importance of tackling development constraints caused by conflict, before other reforms are attempted. 45The Cambodia PRSP was called National Poverty Reduction Strategy (NPRS), December 20, 2002. 46 Several of the PRSPs considered the question of internally displaced people and refugees. The Bosnia- Herzegovina policy action program highlighted the fact that despite considerable progress, an estimated half of the 1995 caseload remained reluctant or unable to return to their homes. Property disputes were among the chief problems inhibiting return, and the PRSP proposed a plan to build new homes and to expedite the resolution of such disputes. The situation of internally displaced people was also addressed in the policy action programs in Sierra Leone, Burundi, and Georgia. Policy Actions and Conflict Prevention How sensitive were the PRSPs to the likely impact of policy actions on the conflict, and to what extent did these actions attempt to address sources of conflict? First, while there was a close thematic overlap between the conflict factors identified in the countries and the PRSP's program of policy actions because of the close link between many of the factors affecting poverty and those affecting conflict, the linkages were not systematically explored. Second, while much of the planned action would contribute to preventing conflict if implemented, despite not having been primarily designed for that purpose, there were also missed opportunities given the limited exploration of the linkages with conflict and consideration of those in the design. Improved governance is a central theme in the PRSP process overall, and characteristics of governance had been identified in all the country case studies as key conflict factors. While most of the reviewed PRSPs did not explicitly recognize the governance-conflict link ­ some perhaps because it would be politically impossible, and others because conflict really did not figure in the analysis ­ a few did. The Rwanda PRSP, for example, stated that "good governance, conflict and economic prosperity are deeply interlinked,"46 and proposed civil service reform, democratization, decentralization, and reforms for improved accountability and transparency to increase equity and strengthen political institutions. Proposed reforms in the Rwandan PRSP also aimed at transforming the centralized system into more inclusive political processes, in order to improve relations between the government and ordinary citizens. Other governance programs, including national reconciliation, human rights, gacaca,47 ubudehe,48 and democratization, would potentially impact positively on social and ethnic relations. In the Cambodia NPRS, the measures promoted under the heading of improved governance would, if successfully implemented, have included local governance that would empower communities and provide effective public services. The measures would also have curbed corruption and enhanced social justice. Such achievements would have gone a long way toward reducing the chances for future conflict escalation; however, the actions necessary to support these measures, including development of the legal and regulatory framework, were never fully implemented. The economic development actions offer both positive and negative lessons on sensitivity to conflict. The Nepal PRSP, for example, placed strong emphasis on stimulating rural growth, based on the recognition 46Rwanda PRSP, June 30, 2002, page 32. 47Gacaca are traditional community-level courts, originally used to resolve conflicts between two families. The system has been adopted for trials of genocide perpetrators with members of the courts drawn from the local community. 48Through the ubudehe process, communities came together to identify problems, solutions, and priorities. During 2003, the ubudehe pilot project in Butare was evaluated and found to be useful for nationwide efforts to decentralize. Ubudehe was launched nationwide, beginning with 6 provinces. Fiscal decentralization was introduced by allocating part of the government's budget to districts. In 2002, provinces became responsible for their own budgets. A Common Development Fund (CDF) for development purposes was introduced, and regulations established so that it would get an increasing proportion of inter-governmental transfers. Districts also got the legal right to collect their own income from property tax, rental tax, and licensing fees. 47 that growth previously had left large parts of the population excluded from development. Given the strong linkage between the conflict and the underdevelopment of certain rural areas in Nepal, effective rural-oriented action could potentially contribute to de-escalate conflict. This linkage was recognized in the PRSP, which as a rationale for the rural emphasis cited the country's poverty problem as well as "social and political context".49 The PRSP presented a series of actions, including improved inputs, diversification, access to credit, and marketing as well as better-targeted irrigation, electrification, and other infrastructure programs. The challenge for the government, however, is how it will be able to implement these activities, given the limited access to many rural areas on account of insecurity. Another PRSP saw growth as a key driver of poverty reduction, and proposed a number of reforms designed to stimulate aggregate growth. The strategy highlighted the importance of private sector development and trade for attracting foreign direct investment and expanding export markets. From a conflict perspective, however, the discussion of growth failed to take account of several key issues. For example, the PRSP highlighted the country's strategic location as a transit route for the region, but failed to note that many key parts of this route pass through conflict regions and are currently not in operation. In the same PRSP, a discussion on energy supplies failed to note the potential contribution to the national grid from a hydroelectric power plant based on the border with a neighboring country, with which strong tensions exist. Finally, the section on stimulating tourism ignored the constraints posed by continuing conflict and insecurity, and failed to discuss how different groups in the conflict regions may perceive the attempts to boost tourism. Land was at the heart of conflict in several of the countries ­ how was it dealt with in the PRSPs? The Cambodia NPRS put land policy and administration, including titling, at the center of the strategy. It also planned to link land registration to dispute resolution, which indicated that the government was aware of the potential land conflicts that could result from these actions. Although it promoted a legal and policy framework, the NPRS did not offer a mechanism for redressing land acquired illegally in the past. In Nepal, land issues did not figure much in the poverty diagnostic, but the action program included a short section on land reform and land management, which touched on some of the critical issues underlying the conflict, and set the goal of increasing poor people's access to land. In Rwanda, the PRSP highlighted land as a source of conflict and as an increasingly pressing issue due to population pressure and soil degradation.50 While the PRSPs varied in their consideration, and especially systematic consideration, of how policy actions might affect conflict, there are positive examples. This includes Bosnia-Herzegovina's program to stimulate private sector-driven growth through three key sets of actions: increasing privatization, facilitating entrepreneurship, and attracting increased foreign investment. The program recognized a number of risks that could increase social tensions, most notably a rise in unemployment caused by privatization and other structural measures. Consequently, the program emphasized the need to put in place an adequate social safety net, dismantle barriers to improved inter-entity cooperation in commerce, and address discrimination in the labor market, with the aim of creating a single economic space. This program could positively impact the conflict environment if it succeeds in creating bridging social capital and demonstrating the benefits of cooperation among different ethnic groups. 49Nepal PRSP, May 30, 2003, para. 99. 50The Land Policy Act and Land Law have been drafted and are waiting to be approved by the parliament. 48 Box 5: Conflict-Sensitive Policy Reforms Conflict-sensitivity is not only a matter of the problems addressed and how they are framed; it is also a matter of how action is designed. One of the PRSPs focused on labor market reform with the aim of removing obstacles to job creation, investment, and business expansion. While the reforms were well designed from an economic perspective, they did not take conflict into consideration. In particular, they did not consider how these reforms would affect the absorption of demobilized soldiers and former combatants into the workforce. In the immediate term, demobilized soldiers and ex-combatants might be provided with rehabilitation benefits, but in the longer term, a favorable employment environment would have to be created. In such a situation, how might ex-combatants affect the labor market reforms, and conversely, what might be the potential impact of labor market reforms on former combatants. Institutional Arrangements Summary Findings · Some governments placed a high premium on developing institutional arrangements that considered conflict issues (ethnic or religious divisions, regional imbalances) by designing structures that either consciously ignored conflict factors or purposefully took them into account. · Other governments made limited efforts to consider conflict issues, reinforcing beliefs that the establishment of pluralistic values was not a priority for the government. · Even though institutional arrangements were relatively broad-based in design across the sample cases, not surprisingly perhaps, they showed mixed results in their level of devolution in PRSP preparation. · In many cases, the PRS process has resulted in enhanced cooperation among sectors and ministries. · Parallel peacebuilding processes in-country have influenced and been influenced by the PRS framework. Institutional arrangements refer to formal structures and rules that determine the design and implementation of the PRSP. Given that the PRSP is the key policy document in some countries, the shape that institutional arrangements assume during the PRS process are particularly important, as they will influence both the content of the PRSP and the implementation process. The manner in which power is distributed through institutional arrangements, and the structure of relationships between government and non-government actors, can either reinforce the power imbalances that contributed to conflict or seek to redress them. It follows that the PRS process can benefit significantly from establishing institutional arrangements that are sensitive to factors of conflict, particularly those associated with concentration of power. Moreover, by instituting inclusive and broad-based arrangements, it is likely that governments will be seen to recognize and perhaps even address conflict factors related to exclusion, concentration of power, and the control of public assets by a single group. The case studies found that some governments placed a high premium on developing these types of institutional arrangements for the PRS process, while others made limited efforts to consider conflict issues, including ethnic and religious divisions and regional imbalances, in their institutional arrangements. In turn, this reinforced beliefs among certain groups and regions that the re-establishment of pluralistic values was not a priority for the government. The analysis showed that in a few cases, the government established arrangements that did not take existing conflict fault lines into account, but favored and overrepresented one group in these arrangements. In other cases, governments were cognizant of conflict issues and designed institutional arrangements that purposefully took them into account. Even if conflict fault lines were not reflected in the design of institutional arrangements, some governments consciously decided not to have conflict factors be incorporated in institutional arrangements, rather than simply failing to account for them. In post-war Rwanda, for example, the 49 government deliberately set about to establish ethnic-blind institutions, and ignored ethnicity in setting up institutional arrangements because it believed that this would risk cementing the divides that had led to war. In BiH, however, the institutional system under the PRS process was deliberately designed to balance the interests of different ethnic groups, although this was a natural arrangement that mirrored the existing government structures. The PRSP in BiH did not fall hostage to one or more ethnic interest groups, and made a positive contribution to challenging exclusive ethnic politics. In Sierra Leone, the government realized that over-centralization and exclusion of rural areas in development planning was one of the main sources of conflict. Hence, it made a conscious decision to institutionalize arrangements for the PRS process that were broad-based and allowed for the incorporation of inputs from different government and non-government actors, particularly in the rural areas. The cases demonstrate that at least in design most institutional arrangements for the PRS process were relatively devolutionary with different organs of government and NGOs being accorded specific responsibilities. The cases however showed mixed results on institutional arrangements being devolutionary during actual PRSP preparation with the level of the influence and involvement of parliament, government bodies, NGOs etc. varying across cases. In terms of design, the central government had overall responsibility for the PRSP, and typically appointed a steering committee to closely oversee the institutional arrangements. Under this arrangement, in some cases, members of parliament, lower levels of government (provincial, district), line ministries, and civil servants were given the opportunity to engage in the process and provide inputs to the PRSP. They were also assigned a principle role in implementation, supervision, and monitoring. Relevant actors from NGOs, civil society, CBOs, grassroots communities, and the private sector were also involved In terms of preparation process of the PRSP, BiH is a best practice example. The Office of the PRSP Coordinator realized early on that "it is only if every sector of society in BiH is engaged in drawing up the Strategy that we can expect it to be implemented."51 Actions were taken to ensure that the PRSP would reflect a nationwide consensus on both the diagnostic and the program of action. The parliament then endorsed the final version of the PRSP, and its interest in the roll-out of the program is evidenced by the fact that it has requested a status report every six months on its implementation.52 In Nepal, considerable effort was devoted to eliciting opinions of local government in eastern, central, and western provinces during I-PRSP preparations. Subsequent disruptions in the security situation have constrained further efforts on this front, but the PRSP highlights the country's desire for an increased role for local government in development planning. On the other hand, in at least one country in this analysis, the Secretariat assigned to lead the PRSP appears to have developed it in isolation, rarely convening the sub- commissions and essentially drafting the preliminary versions. Line ministry involvement was minimal, with many parts of government unaware of the existence of the exercise. As a result, the first draft of the PRSP was rejected by the donor community and reframed instead into a set of discussion materials. With significant donor support, the revised PRSP was significantly improved but with a change in government, there is uncertainty as to whether the PRSP will be updated or a new PRS process will be initiated. Most governments made a good start by establishing comprehensive institutional arrangements for PRSP preparation. Since most cases are now in early stages of implementation, it is neither appropriate nor possible to make an informed judgment about how governments will follow through on their commitment 51 How the Development Strategy of BIH ­ PRSP ­ Came to Be, See: www.bih.prsp.info/knijga/ZA- WEB/english%2003/prsp%20eng/4.pdf 52 On the involvement of parliaments in PRS preparations, BIH is an exception. In most sample cases, parliamentary involvement dissipated in the course of the PRSP formulation. 50 to devolve operations, in terms of decisionmaking authority and resource allocations. In one case, the institutional arrangements under the PRSP reflected a clear devolution of power and contributed to broad consultations and strengthening of institutions, but key decisionmaking power still rests with the President and a narrow group of advisors. In at least one other case, due to capacity constraints, the institutional arrangements were dominated by an ethnic group different from the ruling political group, leading to lack of political will to implement the PRSP. In contrast, the Nepal PRSP places strong emphasis on responsible agencies' monitoring activities and impacts, and reporting on these regularly to the government and other stakeholders. The National Planning Commission has a new supervisory responsibility to verify the accuracy of line ministry reports. Under the PRSP, further devolution to enhance community management of facilities is expected. While progress is expected to be slow because of capacity constraints, the PRSP certainly encodes a vision of a more open and responsive government. The evidence from the case studies does not indicate the extent to which governments will adhere to their decision of inclusion and broad-based consultations in implementation of PRSPs. The reluctance of some governments to be inclusive, however, could be counterproductive, since the success of the PRS process is predicated on institutional arrangements that devolve power and transfer the control of resources to lower levels of government. Challenges to Institutional Arrangements in Conflict-Affected Countries · Devolutionary arrangements. Governments may be willing to design devolutionary institutional structures for the PRSP. Given their lack of experience with devolution and inclusiveness however, they may be less willing and able to transfer power and resources during PRS preparation and implementation as this could undermine their influence and dissatisfy constituencies. · Centralization of power. The government has to overcome its natural tendency to centralize power by involving a range of state and non-state stakeholders in the PRS process. · Battles for control of PRSP. To be effective, the PRSP has to be a collaborative effort. Economic ministries however tend to struggle for control of its preparation and implementation. Thus, ministries have to come to an understanding so that they do not undermine each other, and by extension the entire PRS process. PRSP principles envisage that collective responsibility will increase as cross-sector collaboration and coordination between the center and local governments improves. The case studies indicate that the PRS preparation process resulted in enhanced cooperation among sectors and ministries. Governments are trying to clearly articulate the specific roles that each actor needs to play in the PRSP institutional arrangements in order to prevent territorial battles, establish cross-sector thematic groups that build linkages among various parts of government, and provide training that encourages inter-sectoral cooperation. In some cases, the PRS process created a more level playing field in which no single ministry had disproportionate influence; rather, the dialogue established during the PRS process strengthened the voice and influence of other ministries. In other cases, the government failed to define the roles of different actors in PRS preparation, leading to inter-ministerial struggles for control, and complaints by excluded government representatives. For example, during fieldwork in one of the case studies, many government officials and civil servants claimed that they were deliberately excluded from PRS participatory processes. If their grievances are not addressed, PRS implementation may be affected, since their cooperation and leadership are needed for successful implementation. Cross-sector thematic groups, an initiative undertaken by a few governments, have been an effective forum for enabling engagement across line ministries and producing a strategic outlook on the part of 51 ministries. In BiH, cross-sectoral thematic groups founded as part of the PRSP "strengthened cooperation among different government members at both a technical and senior level, which is important for a post- conflict society."53 Similarly in Rwanda, in terms of improving collective responsibility, the PRSP strengthened a number of institutions and created new ones to assume responsibility for key areas of the PRSP. As a consequence, cooperation throughout government has improved due to the consultative processes. In a few cases, parallel peace-related processes in-country have influenced and been influenced by the PRS framework and helped to strengthen inter-sectoral and inter-ministry relations. The PRSP has promoted the inclusion of poverty-related issues into peace-building processes, and conversely reconciliations issues have been integrated into the PRSP. Independent budgetary frameworks have also interacted with the PRSP, enhancing cooperation among different government organs. In Rwanda, continuous and interlinked consultations with the National Unity and Reconciliation Commission (NURC) pushed reconciliation and human rights programs to address poverty issues; and the decentralization policy was continuously adapted to promote poverty reduction in line with the PRSP. Simultaneously, several concurrent programs have fed into the PRSP. The NURC's action plan, the work of the National Commission for Human Rights (NCHR), the gacaca courts built on traditional justice at the grassroots level, and the decentralization policy have been incorporated into the PRSP in a variety of ways. In Nepal, the National Planning Commission and the Ministry of Finance took the lead in preparing the PRSP. Simultaneously, the Ministry of Finance prepared the Medium-Term Expenditure Framework (MTEF), which served as the resourcing framework for the PRSP. The explicit link between the PRSP and the MTEF has led to improved relations among ministries and better integration of the planning and resourcing functions of the government. Donor Behavior Summary Findings · Country ownership of the PRSP tended to increase as the country further moved out of violent conflict. · Donors tended to have unrealistically high expectations of the quality of PRSPs in post-conflict countries, given capacity weaknesses and continuing divisions among population groups. · In some cases, donors consciously refused to align their strategy with the PRSP if it did not address conflict issues as this would imply that they endorse the exclusion of conflict. · Lack of prioritization in the PRSPs made it difficult for donors to align their programs with the country program. · Donor coordination was not optimal in many cases, although harmonization efforts are on the rise. Just as governments need to seriously consider conflict issues in the development of PRSPs, it is equally important for donors to examine their own behavior, consider how they can engage with governments more effectively and coordinate more systematically to support PRSP development, i.e. not drive it. The PRSP model envisages a partnership between donors and the country, with the aim of clarifying the aid relationship. To this end, the PRSP seeks to change donor behavior on three key fronts. First, donors are expected to respect the principle of country ownership by tempering their advice and encouraging the government to listen first and foremost to the voices of its own citizens. Second, development partners are encouraged to coordinate more effectively, ideally under the leadership of the government, and provide more consolidated inputs into the policy dialogue, in order to avoid fragmenting the policy process and absorbing limited government capacity. Third, donors should ideally align their financing 53Zlatko Hurtic, Realities of the PRSP Process, 2002. See www.bih.prps.info/english/odnosi/aktivnosti/ presentation06.02.ppt 52 with PRSP priorities; provide funds through government systems (ideally the budget) rather than through parallel processes; and simplify and harmonize their administrative procedures to achieve economies of scale. Post-conflict environments present a series of challenges in moving toward these ideals. Conflict-affected countries are more likely to be dependent on international assistance, arguably giving donors more influence than in non-conflict affected countries; governments tend to be weaker and consequently the power balance is more likely to favor the donors; and international agencies are likely to operate with a humanitarian as opposed to development mindset, focusing on short-term needs and often by passing government and implementing projects through non-governmental intermediaries. Shifting from an interventionist role, which often makes sense in a conflict setting, to one that is more respectful of country-driven efforts, demands new incentives and sensitivity on the part of donors. The case studies found that donor behavior in conflict countries was uneven, and that there is considerable room for improvement. In a few cases, donor behavior undermined PRS development, particularly in the countries either coming out of conflict or trying to deal with unresolved conflict challenges. In general, however, donor interactions and mechanisms for coordination have improved significantly. Challenges for Donors in Conflict-Affected Countries · Country ownership. Given heavy engagement and greater influence in conflict-affected countries, donors have to consider how they can simultaneously be less interventionist and more supportive so that the PRSP is a country-driven, home-grown product. · Harmonization. It is understandable that donor policy agendas differ in emphases, however they should aim to coordinate activities so as to reduce additional burdens on governments or enable governments to take advantage of the lack of donor harmonization. · Align support. To strengthen the effectiveness of the PRS process, donors should align country programming to the PRSP. Some donors may be reluctant to do so if conflict factors are overlooked in PRS development. This is a serious dilemma that donors need to address if the PRSP is to be the accepted framework for development assistance. Donor Assertiveness vs. Country Ownership Country ownership, a key pillar of the PRSP, suggests that the PRS be a product of genuine national efforts and be representative of the needs and priorities of the diverse stakeholders in a country. The case studies indicate that countries that had not experienced active conflict for a few years and were able to successfully recognize, even if not resolve, conflict challenges, were more likely to ensure that their poverty reduction strategy was country owned. In these cases, it seems that the country determined the prioritization and pace of development, and managed the exercise without excessive intervention. Donors played a supportive role, providing technical assistance and advice, and in some cases pushing for the consideration of conflict-related issues. In some of these cases, the donor community was included as a key stakeholder for consultations, given their involvement during the conflict period and understanding of the conflict conditions. BiH made robust efforts from the onset to ensure that the PRSP was country owned rather than donor driven. The government used domestic resources to fund the preparation process, and drew on local expertise rather than using international consultants. Donors were included as members of thematic groups and UNDP played a constructive role in organizing donor inputs by establishing working groups and synthesizing donor comments at various stages of the strategy's development, but the ultimate product was country owned. In the case of Sierra Leone, the country was expected to complete its PRSP in 2002 when war trauma and emergency needs were top priorities. The government delayed preparation of the PRSP until it was ready to develop a long-term vision for growth and poverty reduction. Now, two years later, the government is willing to confront these issues and has 53 finalized the PRSP, for which it received considerable technical assistance from donors. While the PRSP was certainly accelerated by donor pressure, the numerous delays in hindsight were not necessarily disadvantageous for early indications are that the PRSP is dealing with conflict challenges and is the product of a country-driven process. A few cases illustrate that donors used financial assistance as leverage to pressure governments to modify priorities and revise frameworks to align them with donor preferences. If the PRSP is intended to be a country-driven effort, donors should certainly advise and assist the PRS process but could undermine it if they try to drive the process. For example in one of the sample cases, a key donor did not demonstrate a willingness to support the PRS process and consequently refrained from providing inputs to the government. The donor believed that the macro strategy outlined in the PRSP was incompatible with its macro framework that was more suited the environment. It pressured the government to redraft the PRSP to make it more congruent with its vision. This resulted in tensions between the government and donors and between government and civil society. In some cases, there was an apparent lack of commitment to the PRSP, and its preparation seemed to be a superficial attempt undertaken to appease donor requirements; i.e., mainly to access concessionary lending. In these cases, the PRSP could be considered government owned (not country owned), since it did not have the buy-in of other institutions or constituencies, was not based on systematic consultations, and often excluded the more conflict-ridden parts of the country or groups without any justification for their exclusion. In some cases, donor pressures to complete the PRSP led to the development of a strategy with little regard for conflicting sensitivities and diverse constituencies. Such PRSPs did not reflect the needs of the country, particularly the special concerns of conflict-affected areas, leading to its rejection by excluded groups and often being rendered irrelevant with a change in government. A common complaint across the countries was that the PRSP is the latest in a series of requirements imposed by International Financial Institutions (IFIs) on countries in urgent need of resources. Given that the country had been weakened by conflict and most likely had low capacity, donors had unrealistic expectations of the PRSP and set priorities that the governments had no alternative but to follow. Donors respond that they encourage conflict-affected countries to make conflict issues a primary concern; and that they have the prerogative of not aligning their programming to the PRSP if it addresses conflict issues only superficially. Donors, particularly bilaterals, add that country ownership in instances where it is inadequately executed cannot be the only criterion of strategy formation. The countries need to create appropriate structures to tackle conflict challenges if they want to receive donor support. Donor Coordination Donors acknowledge that differing priorities and limited coordination in the past resulted in overlapping programs or incompatible policy actions; and there were many instances where their lack of effective coordination affected the PRSP and created inter-donor disagreements. In some cases, differing policy agendas led to different priorities or, under the guide of coordination, attempts to push a specific line of thinking. In one case, a donor coordination unit was temporarily suspended (and continues to be so) because it was perceived as inefficient with meetings dominated by UN appeals rather than donor coordination. In other cases, tension occurred between donors that provided budgetary support to the PRSP and those that continued to support projects that were not necessarily in the priority areas identified in the PRSP. While there may have been sound reasons for providing support outside of the PRSP, particularly when the PRSP ignored the needs of a conflict region, this behavior was generally an obstacle to effective cooperation and policy dialogue. Some donors did not agree on the PRS structure and prioritization for conflict-affected countries, and developed other strategies that they felt were more appropriate. This point is illustrated by the use of the 54 UN agency-developed National Recovery Strategy in Sierra Leone because the UNDP leadership did not consider the I-PRSP as the appropriate implementation strategy in the post-war environment. Box 6: Sierra Leone I-PRSP vis-ŕ-vis National Recovery Strategy In Sierra Leone, the government prepared the I-PRSP in order to meet HIPC eligibility requirements, at a time when the country was still at war and key regions were under rebel control. The strategies and the benchmarks outlined in the I-PRSP were so broad that they encompassed most ongoing donor initiatives. This broadness introduced flexibility but prompted few programmatic changes. Soon after being endorsed by the Bank's board, the I-PRSP was replaced by the National Recovery Strategy (NRS) drafted by the UNDP and other UN agencies.a The NRS is now considered the key framework that guided Sierra Leone's recovery from war. The NRS was a manifestation of the UN's increasing interest in conflict-sensitive development, which sprang from an "emerging understanding of the context and the Security Council's focus on peace-building as an integrated, comprehensive process."a It encouraged the government to identify specific recovery needs that NGOs, United Nations Mission for Sierra Leone, government, and donors could jointly address; and pledged total support for the government's recovery efforts. UNDP's ability to help the government see the NRS as effective and provide funds quickly was due in part to the bridging role UNDP played between humanitarian relief and long-term development. The strategizing behind the NRS differed from that of the I-PRSP. The NRS document emerged organically from a need to strategically assist the recovery of the country, while the I-PRSP was a document required by donors for the government to access funds. Sierra Leone has phased out of emergency recovery period and now is attempting to strengthen peace and embark on development. The PRSP is recognized as the strategy that will enable them to achieve these goals, and donors have coordinated on technical assistance and thrown their support behind the PRSP. aNational Recovery Strategy Assessment, Sierra Leone. Regardless of the reason for the lack of effective donor coordination and unwillingness to use the PRSP as a coordinating mechanism, donors need to guard against creating a "donor circus" characterized by contradictory programs, duplication of support, and lack of communication, so as not to confuse or overburden governments or enable them to take advantage of donor differences. The cases show that a uniform voice is gradually emerging that calls for donors to harmonize their efforts on conflict and development, with a focus on how to carry out development activities in a conflict context. Donor groups have been established to facilitate collaboration on programmatic priorities and the strengthening of country structures needed to address poverty, conflict, and development more generally. DEPAC (the Development Partnership Committee) in Sierra Leone, the Donor Framework Group in Georgia, and the Donor Working Group in Sri Lanka are some examples of institutionalized coordination among donors. Their programs are concerned with both poverty reduction and to activities supported outside of the PRSP framework, such as the multi-donor conflict assessments conducted in Nigeria, Somalia and Sri Lanka. From the case studies, it seems that donors increasingly see the PRSP as a tool that can enhance their communication and coordination of aid and programs, and enable them to transform aid from humanitarian to development assistance. In several cases, however, donors seem to have had competing views on the degree to which countries should be compelled to incorporate conflict issues. Despite such tensions, donors are generally trying to generate common positions and coordinate support. Governments, in turn, seem to appreciate not having to manage conflicting agendas and appease divergent demands. 55 Donor Alignment with PRSP Priorities Donors should ideally realign their country strategies to conform with the priorities presented in the country's PRSP. The case studies found, however, that donors only minimally considered the PRSP in determining country assistance. Given sketchy and vaguely designed policy actions in PRSPs that encompassed a wide-ranging set of priorities, donors were able to say that their existing programs were consistent with the PRSP and did not require realignment. Thus, the PRSPs in our sample did not seem to precipitate major revisions in donor programming. This is disappointing and potentially a concern also in non-conflict countries, since one of the major aims of the PRSP is to encourage donors to rethink their strategy, in order to address the main concerns outlined by the country. In a few cases, when PRSPs ignored conflict challenges or brushed over these issues, donors consciously decided not to align their country strategy to it because alignment would be tantamount to endorsing the exclusion of conflict. In these cases, donors continued to support activities that contributed to poverty reduction and conflict mitigation independent of the priorities identified in the PRSP. Donors argue that if they are expected to align their country strategy to the PRSP, the PRSP should have focused on conflict in a more direct and meaningful manner. In this vein, donors often highlighted the lack of a conflict perspective in preliminary PRSP drafts and encouraged governments to address conflict-related issues. In most cases, however, governments did not respond to these suggestions, so donors felt justified in not considering the PRSP in their country strategies. Donor decisions to not align their strategies to the PRSP are understandable in these circumstances, given their belief that aid will not be effective if the PRSP does not account for conflict concerns. At the same time, given that the PRSP is the accepted framework for development assistance, donors need to balance the government's desire for assistance to meet urgent needs with the recognition that the government may be unwilling or unable to address conflict problems at that time. 56 Box 7: Donor Reviews on PRSPs Several donor studies have raised important issues about the use of the PRSP in conflict situations: A DfID studya questioned whether the PRSP is the most relevant instrument for donor engagement in conflict-affected countries, since the PRSP is premised on the existence of a modern state capable of making a strong commitment to poverty reduction. The study noted that conflict-affected countries generally have "major shortcomings in basic institutional capacity and governance" (p. 4); and that "the degree to which the PRSP can assist in the process of state (re)building will be a crucial test of its relevance" (p. 5). The DfID study also questioned whether there is always a synergy between poverty reduction and conflict prevention. "There may be some kinds of conflict prevention that actually freeze existing structures of exploitation and the often-hidden violence of peacetime, suggesting that there is a problem in placing conflict and peace at two ends of a spectrum" (p. 6). There may also be a danger that "reformers who talk the language of development assistance are acting as cover for those who are prepared to use violence to maintain criminal or violent economies behind the scenes" (p. 10). Another study, by the Netherlands Institute of International Relations,b found, to the contrary, that "processes aimed at the joint formulation of a PRPS may play an important role in the sharing of power and the allocation of government assets" (p. 12). Quoting the Ministry of Foreign Affairs, the study suggested that the criteria for assessing the PRPS in conflict-affected countries should be limited to three main questions: "Is the process sufficiently inclusive? Is the budget transparent? Are enough pro-poor choices being made?" This study agreed with DfID that "a PRSP requires a functional state" (p. 15), but noted that "it is through the PRSP...that the government publicly acknowledges [the causes of conflict] and its responsibility in addressing them. Whether this constitutes a real change or just ... mastering donor language remains to be seen" (p. 16). A third study, by UNHCR,c criticized the PRPS for failing to take adequate account of refugee issues in conflict-affected countries. "Generally, displacement issues are missing in existing PRSP and I-PRSP strategies ... [and if refugees are] mentioned, [they] are perceived as a constraint to economic growth and development, as a security threat or a health risk" (p. 7), rather than as people who can contribute to development. Since in many instances, "the number of displaced people makes up a substantial part of the total population and an even bigger part of the poor population," PRSPs that do not factor in the needs and potentials of displaced people risk not being very effective in achieving poverty reduction" (p. 8). The UNHCR report was also critical of the PRSP's general failure to use community development approaches ­ to target displaced people as well as the local population -- as a way to avoid creating tensions and conflicts in host communities and when refugees return to their homes (p. 8). a National poverty reduction strategies (PRSPs) in conflict-affected countries. Department for International Development (DfID), 2002. b Using PRSPs in conflict-affected countries. Bart Klem, Netherlands Institute of International Relations `Clingendael,' Conflict Research Unit, July 2004. c Poverty reduction strategy papers ­ a displacement perspective. United Nations High Commissioner for Refugees (UNHCR), October 2004. Donors can support fragile governments in conflict-affected countries by providing technical assistance on subjects that are difficult for the government to address. For example, donors can assist governments in understanding the linkages between conflict and poverty, and help them realize that conflict impedes 57 development and poverty reduction efforts. In one instance, a donor-commissioned conflict analysis resulted in the government gaining a deeper understanding of conflict issues. None of the donors, however, helped the PRSP committee to understand the relationship between conflict and poverty as the document was being drafted. Instead, comments on the document were purely technical in nature. In a minority of cases, donors were able to persuade the government to conduct PRS consultations in a transparent manner and incorporate conflict issues in the PRSP, but it is too early to determine whether this increased recognition of conflict will continue in the implementation phase. In one country, the government accepted support from donors on both the procedural and technical aspects of addressing conflict issues in the PRSP. While those donors were not completely satisfied with the process, they generally believe that the PRSP reflects the needs of the entire country, and have committed themselves to aligning their country programs with the PRSP. 58 SECTION IV: TOWARD A CONFLICT-SENSITIVE PRSP Based on findings and lessons of the nine cases, this section lays out suggestions to strengthen the conflict sensitivity of PRSPs. It borrows from experiences in the nine cases and other field and analytical work to discuss innovative ways that the PRSPs can address conflict-related issues, and thereby become more effective in their design and impact. Conflict-Sensitive PRSP ­ The Way Forward To be sensitive to conflict, a PRSP must be specific to the country context and flexible in responding to changing circumstances, while taking into account potential risks. · Country specific. The most fruitful PRS design is based on a thorough assessment of the specific country context, including specific conflict factors. This means there is a strong need for good contextual analysis that systematically discusses the underlying conflict factors as well as conflict factors that emerged due to the war. Countries should also avoid the mechanical use of tools and lessons from other country situations. The PRS should draw heavily on in-country processes such as peace agreements, joint needs assessments, and transitional results frameworks. In countries still facing violent conflict or are in transition to more stable conditions, humanitarian, recovery and development needs will overlap ­ and the PRS needs to take this overlap into account. · Nimble and flexible. Conflict-affected countries are often characterized by great volatility and quickly changing situations, especially in terms of political arrangements, security, and urgent needs. At the same time, many of the countries in such circumstances face serious capacity constraints. Given these challenges, a question brought up during this analysis was how a PRSP can be more responsive to quickly changing situations without weighing down a strained government with repeated demands for resources, capacity, and organization to develop or update the strategy document. The experience is that there is a real need for a PRS framework in such countries, but that the process and strategy could be structured such that design and implementation allow them (i) to respond relatively quickly to changing situations; (ii) be flexible in their design and implementation; and (iii) develop alternative approaches when changes render current measures irrelevant. While flexibility is key for the PRS in conflict-affected countries, it should not be interpreted as allowing for a laissez-faire approach on part of the country leadership or donor partners. Instead flexibility suggests the development of unique and innovative methods to address PRS analysis, participation, and implementation. · Risks. Like any activity in a conflict-affected country, the development of a PRS carries certain risks. The more carefully the process takes conflict factors into account and the more realistically the PRS content reflects the context, the better the potential risks can be managed. Risks may include: (a) Unrealistic perception among people of what the PRS would deliver may contribute to renewed instability if tangible changes do not take place. Such a perception would increase if the process is not well explained e.g., through media, or if politicians use the process for political mobilization. (b) Participation in the PRS process may lead to beliefs among interest groups that their concerns will be taken care of, and may result in a PRSP that avoids hard choices or prioritization in an effort by government to maintain support from diverse constituencies. (c) Potentially divisive issues, e.g., ethnic differences, may be treated insensitively in the PRS process or content, and may increase tensions. (d) Immediate post-war needs may be addressed in a way that undermines the longer-term interests of the country, e.g., by creating (or prolonging) dependencies of war-affected communities, or providing different level of support to different groups such as demobilized combatants and returning IDPs. 59 Participation Country ownership. In some of the sample countries, the PRSP lacked the necessary ownership at the highest level of government to ensure institutionalization and sustain participation. Formulation of the PRSP took place in environments of low state capacity and legitimacy, as well as weak linkages among the political powers, the bureaucracy, and the population. The potential for the PRSP to be a vehicle for stabilization would increase with the government's ability to involve a broader group of stakeholders and demonstrate that poverty reduction efforts are for the benefit of all citizens. As the discussion of conflict factors showed (section II), the exclusion of population groups from economic and political processes constitutes an important source of conflict in many countries. A particular challenge for the government in such cases would be to demonstrate the political will to break this pattern and find a way to constructively engage excluded groups in the PRS process. Box 8: Alternative Models for Enhancing Country Ownership Countries that experienced conflict prior to the advent of the PRSP, including Mozambique and Senegal, have introduced a Poverty Observatory to encourage joint official and independent oversight of PRS implementation. Stakeholders in these countries ­ including government officials, civil society groups, religious bodies, business leaders, trade unions, NGOs and external partners ­ are charged with the task of conducting a comprehensive review of PRS progress, which the Government applies to an assessment of poverty outcomes. Conflict-affected countries could consider adapting such an approach to their own circumstances. Collaboration and reconciliation. There is evidence that the participatory processes in a conflict- affected country can be a vehicle for increased collaboration between and among population groups that have experienced tension or conflict, and can help promote reconciliation. The impact in terms of reconciliation would be enhanced if the participation process is designed with this outcome in mind. This would have an effect on the timing and facilitation of participatory processes, the selection of participants, and their opportunity to engage. Every country would have to find its own formula for such a process ­ it cannot be imported.54 Building trust and managing expectations. The sample testifies to PRS participation being a tool for initiating engagement with previously overlooked populations. For participation to be meaningful, it needs to be tempered by a realistic view of what actions government has the capacity and resources to undertake at any given time. The expectations of the populations also need to be managed by institutionalizing the dialogue, especially with vulnerable, war-affected communities and excluded groups. The validity of participation processes then depends on the extent to which ownership of the PRS is maintained from one administration to the next, as trust in the authorities will grow when laws on freedom of association and access to information are respected, and when independent bodies such as reconciliation commissions and ombudsmen are seen to administer justice in an equitable manner. Injecting transparency into the input-output cycle. It would be beneficial to establish structures that capture the input-output cycle. Given limited resources and the need for prioritization, there are always groups and organizations whose views are not going to be incorporated. These groups are likely to criticize the PRSP as being non-representative. To avoid this problem, monitoring mechanisms for participation should be developed that (i) identify the different groups/organizations who provide input and record their input; (ii) highlight the inputs that were relevant for the process; and (iii) outline how the 54 There are many resources available on this issue, including Reconciliation ­ Theory and Practice for Development Cooperation, Swedish International Development Cooperation Agency (SIDA), September 2003. 60 input was fed into and influenced policy development. By stressing transparency and openness, the PRS would be grounded in a more participatory process. In Bosnia and Herzegovina, for example, the authorities deliberately prepared a comprehensive first draft PRSP that included as many as possible contributions from civil society. Government wished to indicate clearly that stakeholder views would be given due consideration in the strategy and would be reflected in the prioritized PRS approved by the State Parliament. By taking the inputs from civil society seriously, the government introduced transparency and legitimacy into the process, resulting in support from all civil society groups for the final PRSP. Use of traditional mechanisms. A number of countries in the sample reflected considerable experience using local practices and mechanisms to ensure communication and collaboration between and among groups, which could help to mitigate conflict. Dissemination of these and similar practices both within and among countries could be used as a basis to enhance prospects for conflict prevention and social cohesion elsewhere. It would be important, however, to use traditional practices with sensitivity, as some of them may be counterproductive and even exacerbate conflict.55 Role of parliament. An inclusive system requires meaningful participation and representation that integrates society at large into decisionmaking processes. Improving the representative capacity of parliament strengthens its ability to reach out to all sectors of society and potentially play a role in mitigating conflict. Opportunities can be sought to strengthen parliamentary oversight of PRS implementation, in order to build country ownership of the strategy, generate broad country support for reforms, build parliamentary capacity to improve PRS linkages with the national budget, and raise the level of parliamentary debate on conflict mitigation. Elected bodies and participatory mechanisms also need to be strengthened at the local and community levels in order to take account of diverging opinions and preferences, thereby helping the PRS to become more conflict sensitive. Capitalizing on the energy of youth. In conflict-affected countries and new democracies, civil society tends to be weak and to focus on short-term goals. In order to sustain the participation of civil society, and thus support its ability to influence policy over the longer term, networking with youth groups and organizations could be supported, so as to capitalize on the energy and outreach of youth as building blocks for their participation over time. Mobilization of young people for poverty reduction is especially relevant in conflict-affected countries where youth are often marginalized from social and economic opportunities and are easily recruited into rebel groups. Media and strategic communications. The media's capacity to analyze the PRS process, and disseminate information on the PRSP and other development initiatives to remote and conflict-affected communities, should be strengthened. Media involvement would also help to channel feedback from vulnerable groups to policy levels, thereby enhancing the voice of those most affected by conflict. As evidenced in the case of Bosnia and Herzegovina and Sierra Leone, television and radio debates were used effectively to make the PRS real in the eyes of the average citizen. Use of the media also increased the obligation of the authorities to implement a process that was familiar to and owned by the society as a whole. South-South and North-South capacity building. Capacity shortcomings on the part of both civil society and government agencies constitute a major obstacle to the successful implementation of participatory approaches. Knowledge sharing on promising approaches to domestic accountability and post-conflict 55Use of traditional methods should be handled with sensitivity as they may not always be helpful. In one PRS process in the sample, drawing on the fora established by traditional chiefs proved to be counterproductive because they precluded the possibility for youth to express their priorities or concerns.. 61 participation processes, not only from developed countries but also from other developing countries that may or may not be conflict-affected, could be promoted and supported by partners. Box 9: Adapting Successful Experiments with Participation Countries can learn from the Ghana's experience with developing a successful consultative structure.a For the SAPRI (Structural Adjustment Participation Review Initiative), extensive consultations and town hall meetings were held at 10 regional forums, led by a civil society council that was very broad in its reach, including public interest groups, unions and NGO consortia. These regional forums saw participation from teachers, nurses, farmers' cooperatives, local businesses, NGOs, CSOs, and other groups. From each regional forum, 25 representatives were selected to participate in a national forum. The 250 representatives at the national discussions set the agenda and highlighted priorities, which were discussed with the donors. Decisions were relayed back to the regional forums. Most groups agree that this experience with participation was fairly representative and meaningful, and it created a civil society coordination structure that lasted well beyond the SAPRI program. Given that Ghana is a relatively stable country, it may not be possible to duplicate this experience completely. It is possible, however, for conflict-affected countries to design consultative structures using the principles of Ghana's successful experiment. aThis box benefited from the comments of Peter Harrold, Country Director, Sri Lanka, based on his experience with the SAPRI initiative when he was World Bank Country Director, Ghana/Liberia/Sierra Leone. Poverty Diagnostic Focusing on conflict-induced poverty. Political and practical constraints may make it difficult for conflict-affected countries to focus on conflict in their poverty diagnostic, but it is nevertheless important that there be a discussion of conflict-induced poverty. Such a discussion could highlight differences from traditional structural-based poverty, which would promote understanding of the special needs of the conflict-affected poor. Recognizing conflict factors, however, is not sufficient; governments should also consider undertaking some form of systematic analysis that would inform the relationship between conflict and poverty. They could also potentially learn from humanitarian agencies that have been active in the conflict zones. Exploring conflict-poverty linkages. Country programs would benefit from an increased understanding of the interrelationship among poverty, poor governance, and marginalization. This could be achieved by systematically integrating conflict analysis tools with poverty diagnostics, especially through an in-depth analysis of conflict factors that perpetuate poverty, such as regional disparities and unequal access to services. Data collection with donor support. Data collection should be considered a key component of the post- war recovery process. In an environment of fragility, governments can ask donors to provide technical expertise and support capacity building. Poverty diagnostics could be built by combining quantitative and qualitative approaches to better cover the different aspects of poverty in conflict-affected countries. Needs assessments and Participatory Poverty Assessments in conflict-affected countries can help provide data on issues such as non-income dimensions of poverty and participation. These in turn could feed into the diagnostic, thus making the poverty profile more dynamic and relevant. Data collection, drawing from humanitarian agency and NGO expertise. In conflict-affected environments, the security situation, political and capacity constraints, and limited trust (between government and citizens, between and among different population groups) make data collection difficult. 62 In these instances, it may be prudent for governments to build on the experience of humanitarian agencies and NGOs that operate in the conflict-affected regions of a country. Data collection in inaccessible areas. Following from the above two points, in regions where normal tools do not work well and the situation is fluid, the government may request assistance from humanitarian agencies, NGOs, or donors in developing proxy indicators that provide a reliable picture of conflict-related poverty. The PRSP should clearly articulate that certain areas or groups are represented in the diagnostic by proxy indicators. This would increase understanding of how different areas and groups are captured by the poverty diagnostic. Policy Actions Systematic integration of conflict factors. The policy action programs in many of the PRSPs recognized conflict to some degree but did not integrate this perspective in a systematic manner. Some of the programs would, if implemented, contribute to preventing future escalation of conflict; and a more systematic inclusion of conflict sensitivity into the planning process would strengthen the conflict prevention outcome manifold. In particular, the policy actions would be more conflict sensitive if their selection, prioritization, and content are systematically assessed through a conflict lens. There are three key steps to including a conflict perspective in the PRSP: (i) a recognition among all stakeholders that inclusion of a conflict perspective will make policy actions more effective; (ii) selection, prioritization, and design of actions based on contextual analysis that has determined conflict drivers and linkages with poverty; and (iii) assessment of the planned policy actions for their potential effects on the conflict environment. Stronger contextual analysis. In a conflict-affected or at-risk country, a contextual analysis would determine key drivers of conflict escalation and de-escalation, and how they interact with factors affecting growth and poverty. If possible, the analysis should be a component of the poverty diagnostic, not conducted as a separate exercise. Conflict studied as an integral part of the wider poverty/growth diagnostic would increase the chances that conflict sensitivity will be organically integrated into the planned policy actions, not just tagged on. This would provide for more realistic goals, planned actions that have a reasonable chance of being implemented, and design that can achieve the desired outcomes. Countries that have been affected by violent conflict need to continue focusing on repairing the damage done, but strong analysis would help the government to prioritize according to agreed criteria. Security issues. While security issues may not normally be considered as part of poverty reduction, the experiences of conflict-affected countries show that such issues are of critical importance for longer-term poverty reduction and sustainable growth. A successful PRS depends on a reasonable level of security, and on sound management of security-related expenditures. Security and rule of law are essential public goods that help create the conditions for government accountability, robust private sector development, and the participation of populations in conflict-affected societies in social and development processes. While most of the PRSPs included some security actions, such as demobilization or de-mining, such actions would be more effective if they are informed by strong contextual analysis and better integrated into a cohesive strategy for improved governance of the security sector and enhanced economic opportunities. 63 Box 10: Security System Reform (OECD/DAC) The SSR policy agenda covers three inter-related challenges facing all states:a (i) developing a clear institutional framework for providing security that integrates security and development policy and includes all relevant actors and focuses on the vulnerable, such as women, children, and minority groups; ii) strengthening the governance and oversight of security institutions; and iii) building capable and professional security forces that are accountable to civil authorities and open to dialogue with civil society organizations." For DAC donors , this policy agenda, therefore, focuses primarily on governance- related and democratic oversight dimensions. aOECD Policy Brief: Security System Reform and Governance: Policy and Good Practice, May 2004. Impact assessments. As important to consider as the direct effects of individual action programs or projects are the potential impacts of overall policies and strategies; for example, the distributional impacts of a country's growth strategy. This goes beyond assessing the risk for specific vulnerable groups and developing corresponding safety nets: it deals with the longer-term and often indirect effects, which are often mediated by complex social structures. Many organizations have developed tools to assess and monitor immediate and longer-term policy impacts on the conflict environment; these can be adapted for use by governments to systematically consider the impact of PRS policy actions. These tools should help increase the understanding of: (i) how policy actions could impact conflict-affected communities; (ii) how policy actions could affect (already tenuous) relations between groups; (iii) how policy actions may impact conflict factors (e.g., regional imbalances); and (iv) how factors of conflict could threaten the planned policy actions. The tools should not be used mechanically, but tailored to the specific country context. Box 11: Potential Conflict Impact of Policy Actions There may be a trade-off between the intended and unintended consequences of a specific policy or action ­ it may succeed in achieving its intended objective at the cost of exacerbating conflict. For example,a an education program may succeed in increasing the number of students passing the state-wide examinations, but if the bulk of those students are members of one particular social group, the program may exacerbate inter-group tensions by underscoring the perception that one group is being privileged at the expense of another. The converse also holds true: an education program may fail to produce students able to pass state-wide exams, but may succeed in reducing tensions between particular social groups by creating and institutionalizing an environment that increases constructive contact and decreases tensions by dispelling stereotypes and misconceptions. aThis example is taken from Conflict in Somalia: Drivers and Dynamics, World Bank, Washington DC, January 2005. Ethnic issues. The issue of ethnicity was enormously problematic in most of the country cases, largely because of its role in past political mobilization. If poverty diagnostics lack analysis of how ethnicity correlates with poverty, it is not surprising that the policy action programs do not include activities aimed at ameliorating ethnic tensions. One PRSP, for example, asserted the need for special efforts to ensure health and education services for marginalized groups, but stopped short of identifying actions that could have addressed ethnic divisions in service delivery. Rather than ignoring this issue and thus risking that 64 skewed priorities and allocation of resources might further exacerbate tensions, such actions could be framed in other terms, such as linking them to regional interventions. Furthermore, the issue is not whether the problem is explicitly discussed in the document ­ that is often not possible ­ but whether the problem is analyzed and addressed as much as feasible, through the PRS process. Flexible implementation of actions. Conflict-affected countries are volatile, and changes can be difficult to predict. The possibility always exists that the operating environment will not improve as predicted or will deteriorate; thus the PRSP should plan appropriate alternative actions. The Nepal PRSP offers a good example of such foresight. Based on an explicit recognition that continuing insecurity would be a risk to the program, it outlined a number of actions to be taken if the situation worsened. For example, it proposed that NGOs and CBOs could deliver essential services in areas where the government cannot operate; that mobile teams may be able to provide services and perform maintenance functions in areas where a permanent state presence may not be viable; and that programs with quick results such as income or employment generation would be needed "to prevent further alienation and win public support."56 Institutional Arrangements Conscious design of arrangements. The case studies demonstrate that there is no single formula for shaping institutional arrangements, but that such arrangements should be the product of conscious design and the consideration of conflict factors. The government may decide to either deliberately ensure that conflict issues and ignored groups play a key role in institutional structures, or exclude them for legitimate reasons. The government should not, however, use these as pretexts for shutting groups out or disregarding key conflict issues. It is also possible that the design of institutional structures might give political signals, perhaps unintended, of power distributions, i.e. who has power and who does not. To avoid criticism, it would serve governments well to be transparent in making decisions on design and implementation arrangements. Broad-based and inclusive institutional arrangements. The PRSP can be an important vehicle for cohesion and reconciliation if it manages to bring different stakeholders together to develop a comprehensive national framework. To achieve this objective, structural formations that represent diverse interests should be advanced. Institutional arrangements for the PRSP could increase commitment among a diverse range of stakeholders by supporting the gradual devolution of power and responsibilities across state (different levels of government) and non-state (NGOs, civil society, CBOs, private sector) actors. Devolve power and resources. Governments should ensure the progressive institutionalization of inclusive, transparent, and broad-based arrangements for PRS implementation and monitoring. They should follow through on their decision to devolve power and transfer resources to support the implementation of PRSPs. Governments should be aware that if they do not follow through with their commitment to devolve power and resources, the institutional arrangements could collapse, with significant damage to their credibility. This could cause irreversible damage to both the PRS and the process of peace consolidation. Linkages with ongoing processes. As the case studies demonstrate, the PRSP can strengthen ongoing processes, and in turn be strengthened by them. Human rights programs, peace-building exercises, and reconciliation processes can learn from the PRS process how to be more attentive to poverty-related issues. The PRSP can also draw from the lessons of peacebuilding activities. Such reciprocity could strengthen both poverty reduction strategies and peacebuilding initiatives. 56Nepal PRSP, May 30, 2003, para. 205. 65 Donor Behavior Prioritizing country ownership over donor assertiveness. Donors have traditionally had significant influence in conflict-affected countries, but need to consider how to reconcile country ownership with promoting their own priorities. Donors may be eager to have a country address conflict issues in the PRSP, but how to do this effectively without pressuring the government is a key challenge that donors have to address upfront if the PRSP is to be an effective umbrella framework. Donors could strengthen the country's capacity to prepare a comprehensive, conflict-sensitive PRSP by providing technical assistance (e.g., implementation and monitoring mechanisms), commissioning studies, and organizing consultation workshops to ensure that conflict areas and issues are carefully considered, and that structures have peace-building impacts. Donors should not determine PRSP priorities or lead its preparation; rather they should create an enabling environment that makes it more possible for governments to deal with conflict-related issues and sensitivities. Facilitating the inclusion of conflict issues. Donor decisions to provide limited support to a PRSP that does not take conflict concerns into account may be valid because of the belief that aid will not be effective, and could in fact contribute to the exacerbation of conflict if the PRSP remains in its current form. Given that the PRSP is the accepted umbrella framework for development assistance, donors should ideally foster an environment in which governments are able to prepare PRSPs that address conflict concerns. Donors should be able to differentiate between legitimate reasons for omission of conflict issues and exclusionary policies that do not justify ignoring conflict. Donors need to be sensitive to the motivations that drive the cognizance of conflict or its lack thereof in the PRS, and should accordingly make decisions on how to engage with the country's PRS process. If a PRSP does not integrate conflict-related issues into the poverty diagnostic and policy actions, the IMF-World Bank Joint Staff Assessment should point out that the limited conflict sensitivity of the PRSP could affect the basis for financing. Improving donor harmonization is critical. Evidence from the case studies clearly shows that donor coordination has significantly improved, and that donors are cooperating on the nexus of conflict and development issues. Donors should continue moving forward by communicating and collaborating on country programming. More specifically, donors could establish formal coordinating mechanisms to identify their activities, implementing partners, monitoring and evaluation activities. They could also increase their cooperation on conflict-related activities, such as the multi-donor conflict assessment collaborations. Coordination would enable them to take advantage of their comparative strengths, avoid duplication of support and contradictory policy advice and incentives. Box 12: Donor Harmonization ­ Donor Working Group, Sri Lanka Since the Tokyo Conference on Reconstruction and Development in Sri Lanka in June 2003, international donors have been involved in numerous efforts to strengthen harmonization in order to contribute to Sri Lanka's development and a durable peace. Donors have set-up a trust fund, managed by the World Bank, that aims to enhance their coordination and harmonization. Two initiatives supported through this trust fund are the Donor Working Group and the multi-donor conflict assessment.a Heads of mission tasked a donor working group in early 2004 with developing a strategy for monitoring progress on peace. The Donor Working Group on the Peace Process (DWG) was formed with widespread representation from bilateral and multilateral agencies. The DWG encourages shared analysis and provides robust information for collective or individual donor assessments, while still leaving the decisions on aid allocation and conditionality up to the discretion of individual donors. This approach has 66 won the support of a wide range of donors. Key activities of the DWG included a scenario planning exercise, which has identified significant factors affecting progress on peace in Sri Lanka. A local policy and research organization has been hired to provide quarterly reports analyzing trends against these critical factors, with the aim of helping donors make more informed programming and financing decisions. A multi-donor (World Bank, UK, Sweden, Netherlands, Asia Foundation) conflict assessment was commissioned in late 2004 to contribute to conflict-sensitive donor programming. That report is due to be issued in Summer 2005. a This box was prepared with input from Anthea Mulakala, DFID Sri Lanka. Next Steps This report is the product of the first stage of a wider three-year program. Following on from the findings and lessons presented here, the program envisages the following main steps, to take place through 2006: · Further dissemination and discussions of the lessons generated with practitioners and policy-makers in PRSP countries, civil society organizations, and international donors and organizations. The content of the discussions will develop as the program evolves. · Further refinement of lessons and development of appropriate measures to help conflict-sensitivity in the PRS. This will incorporate outcomes of the workshops as well as support of selected country cases. · Support of conflict-sensitivity measures in a selected few countries through the Bank's teams in the countries. This would include development of tailored activities as well as piloting of measures of wider application. · Support of capacity building and cross-fertilization across countries and organizations. · Monitoring of the continuing PRS experience in selected countries as they develop or implement their PRS. · Final dissemination of program outcomes. 67 BIBLIOGRAPHY Addison, T. 2003. From Conflict to Recovery in Africa. London and New York: Oxford University Press. Asian Development Bank. 2000. Cambodia: Enabling a socioeconomic renaissance. Manila (July). Asian Development Bank. 2002. Cambodia: indigenous peoples/ethnic minorities and poverty reduction. Manila (June). Asian Development Bank. 2002. Poverty Reduction in Nepal: Issues, Findings and Approaches. Amnesty International. 2004. Human rights briefing on Burundi. London. Aron, J. 2002. Building institutions in post-conflict African economies. Helsinki: World Institute for Development Economics Research (WIDER). Bannon, Ian and Paul Collier (eds.). 2003. Natural Resources and Violent Conflict. Washington DC: World Bank. Beleli, Özsel. 2001. Aid as a peacemaker: a view from Burundi. Washington DC: Georgetown University (masters thesis). Bigsten, A. and S. Lundström. 2004. Aid and growth in Rwanda. Göteborg, Sweden: Göteborg University (January). Bose, Sumantra. 1994. States, Nations, Sovereignty: Sri Lanka, India and Tamil Eelam Movement. New Dehli: Sage Publications. Bougarel, Xavier, in David Dyker and Ivan Vejvoda (eds.). 1996. Yugoslavia and After. New York: Longman. Bradbury, Mark. 2002. A review of the DfID commissioned study of PRSPs in conflict-affected countries. London: Department for International Development (DfID, August 14). Brouneus, K. 2003. Reconciliation ­ theory and practice for development cooperation. Stockholm: Swedish International Development Cooperation Agency (SIDA). Bugingo, E. (no date). Missing the mark? Participation in the PRSP process in Rwanda. London and Dublin: Christian Aid. Cambodia Development Review. 1999. Vol 3, no. 4 (December). Cambodia Development Resource Institute (CDRI). 2001. Social assessment of land in Cambodia. Phnom Penh. Center for Strategic and International Studies. 2004. South Asia Monitor Bulletin No. 72 (July). Country Indicators for Foreign Policy (CIFP). 2002. Conflict Risk Assessment Report: Cambodia Indonesia, Philippines. Ottawa: Carleton University (January). Colletta, Nat J. and Michelle L. Cullen. 2000. Violent Conflict and the Transformation of Social Capital. Washington DC: World Bank. Collier Paul, et al. 1999. On the Economic Consequences of Civil War. Oxford Economic Papers, No. 51. Collier Paul et al. 2003. Breaking the Conflict Trap. Policy Research Report. Washington DC: World Bank. 68 Collier, Paul and Anke Hoeffler. 2002. Aid, Policy, and Growth in Post-Conflict Societies. Policy Research Working Paper WPS 2902. Washington DC: World Bank. Collier, Paul and Anke Hoeffler. 2002. Greed and grievance in civil wars. Centre for the Study of African Economies, Oxford University. Oxford. Collier, Paul and Anke Hoeffler. 2004. The challenge of reducing the global incidence of civil war. Copenhagen Consensus Project, Environmental Assessment Institute. Copenhagen. Department for International Development (DfID). 2002. National Poverty Reduction Strategies (PRSPs) in conflict-affected countries. London. Department for International Development (DfID). 2003. PRSP synthesis report: experience of PRSs in Asia. London (July). Department for International Development (DfID). 2001. Strategic conflict assessment ­ Sri Lanka. London (July). Department for International Development (DfID). 2002. Economic aspects of insurgency in Nepal. ECON Centre for Economic Analysis. London. Douma, P. 1999. Poverty, conflict, and development interventions in Sub-Saharan Africa. Global Development Network, World Bank. Ebata, Michi. 2001. Mainstreaming conflict prevention in analysis and programming: a review of CCA/UNDAF processes (October). (Common Country Assessment/UN Development Assistance Framework.) Economist Intelligence Unit. 2004. Country Report, Cambodia. London. Economist Intelligence Unit. 2003. Country Profile, Chad. London. Economist Intelligence Unit. 2003. Country Report, Nepal. London. Economist Intelligence Unit. 2003. Country Profile, Georgia. London. Fafo Institute for Applied International Studies. 2002. The logs of war: the timber trade and armed conflict. Oslo (March). Goodhand, J. 2001. A synthesis report: Kyrgyzstan, Moldova, Nepal and Sri Lanka. Conflict, Security and Development Group (CSDG), Centre for Defence Studies. London: Kings College (July). Government of the Federation of Bosnia-Herzegovina and others. 2004. Mid-Term Development Strategy 2004- 2007 (PRSP, March). Government of the Republic of Burundi. 2003. Interim Poverty Reduction Strategy Paper (November). Government of the Republic of Chad. 2003. Poverty Reduction Strategy Paper and Joint IDA-IMF Staff Assessment (October). Government of the Republic of Chad. 2002. Stratégie nationale de bonne gouvernance cellula. Ministry of Development Plan and Cooperation Economic, "Senarec" (August). Government of the Republic of Chad. 2003. Poverty Reduction Strategy Paper (October). 69 Government of the Republic of Chad. 1995. Enquęte sur la consommation et le secteur informel au Tchad (Household Survey). Government of the Republic of Rwanda. 2002. Poverty Reduction Strategy Paper (June). Government of the Republic of Rwanda. 2001. Core Welfare Indicators Questionnaire. Government of the Republic of Rwanda. 2001. Household Living Conditions Survey. Government of the Republic of Rwanda. 2001. Multiple Indicator Cluster Survey. Government of the Republic of Rwanda. 2001. Report on poverty profile survey in Rwanda (October). Government of the Republic of Rwanda. 2000. Demographic and Health Survey Government of the Republic of Rwanda. 2000. Food Security Survey. Government of the Republic of Rwanda. 2000. Interim Poverty Reduction Strategy Paper. Ministry of Finance and Economic Planning (November). Government of the Republic of Rwanda. 2000. Public Expenditure Tracking Survey (PETS). Government of the Republic of Sierra Leone. 2004. Economic Bulletin. Economic Policy Research Unit. Government of the Republic of Sierra Leone. 2002. National Recovery Strategy, Final Report. Government of the Republic of Sierra Leone. 2001. Interim Poverty Reduction Strategy Paper June. Government of the Republic of Sierra Leone. 2000. Multiple Indicator Cluster Survey 2 (MICS2). Government of Sri Lanka. 2002. Regaining Sri Lanka: Vision and Strategy for Accelerated Development. Government of Sri Lanka. 2002. National framework for relief, rehabilitation and reconciliation, (June). Hamilton, Kate. 2004. Politics and the PRSP approach: Georgia case study. Overseas Development Institute (ODI, May). Haughton, J. 1998. The reconstruction of war-torn economies. Harvard Institute for International Development (June). His Majesty's Government of Nepal. 2004. Harmonization of donor assistance in Nepal. Ministry of Finance (May). His Majesty's Government of Nepal. 2003. The Tenth Plan 2002-2007 (Poverty Reduction Strategy Paper, May). How the development strategy of BiH ­ PRSP ­ came to be. www.bih.prsp.info/knijga/ZA- WEB/english%2003/prsp%20eng/4.pdf Hughes, Caroline. 2003. The Political Economy of Cambodia's Transition 1991-2001. London: RoutledgeCurzon. Hughes, Caroline. 2001. An investigation of conflict management in Cambodian villages: a review of the literature with suggestions for future research. Centre for Peace and Development, Cambodia Development Resource Institute. Phnom Penh. Hughes, Caroline and T. and Conway. 2004. Understanding pro-poor political change: the policy process in Cambodia. Overseas Development Institute (ODI, January). 70 Hughes Caroline with Real Sopheap. 2000. Nature and causes of conflict escalation in the 1998 national election. Cambodian Centre for Conflict Resolution. Phnom Penh. Human Rights Watch. 2002. Prospects for sustainable peace in the Mano River Union. http://www.hrw.org/reports/2002/liberia/liberia0402-03.htm. Hurtic, Zlatko. 2002. Realities of the PRSP process. www.bih.prps.info/english/odnosi/aktivnosti/ presentation06.02.ppt. International Crisis Group. 2003. A framework for responsible aid to Burundi. Brussels. International Crisis Group. 2001. Burundi ­ breaking the deadlock: the urgent need for a new negotiating framework. Brussels. International Development Association. 2004. Chad Implementation Completion Report: Fourth and Fifth Structural Adjustment Credits. Washington DC (June). International Development Association and International Monetary Fund. Bosnia and Herzegovina Poverty Reduction Strategy Paper ­ Joint Staff Assessment (May). www.worldbank.org/poverty/***: International Development Association and International Monetary Fund. 2003. Cambodia Poverty Reduction Strategy Paper ­ Joint Staff Assessment (January). International Development Association and International Monetary Fund. 2003. Chad Poverty Reduction Strategy Paper ­ Joint Staff Assessment. Washington DC (October). International Development Association and International Monetary Fund. 2003. Sri Lanka Poverty Reduction Strategy Paper ­ Joint Staff Assessment. Washington DC (March). International Development Association and International Monetary Fund. 2002. Rwanda Poverty Reduction Strategy Paper ­ Joint Staff Assessment (July). International Development Association and International Monetary Fund. 2001. Sierra Leone Joint Staff Assessment (July). International Monetary Fund. 1997. Burundi ­ recent economic developments. Washington DC. International Monetary Fund and World Bank. 2004. Policies and actions for achieving the Millennium Development Goals and related outcomes. Global Monitoring Report. Washington DC. Japanese Bank For International Cooperation. 2003. Sri Lanka ­ Conflict and development: the role of JBIC. Tokyo (August). Keen, David. 2002. A review of the DfID-commissioned study of PRSPs in conflict-affected countries. London: DfID (May). Kiernan, B. and Ch. Boua (eds.). 1982. Peasants and Politics in Kampuchea, 1942-1981. London: Zed Press. Kievelitz, U. and T. Polzer. 2001. Nepal Country Study on Conflict, Transformation and Peace Building. Eschborn: GTZ. Klem, Bart. 2004. Using PRSPs in conflict-affected countries. Conflict Research Unit Netherlands Institution of International Relations "Clingendael" (July). 71 Klugman, Jeni (ed.). 2002. A Sourcebook for Poverty Reduction Strategies, Volume 1: Core Techniques and Cross- Cutting Issues. Washington DC: World Bank. Lama-Tamang, M., S. Gurung, D. Swarnakar, and S. R. Magar. 2003. Social Change in Conflict Affected Areas: Assessment Report. Prepared for DfID. London. Management Systems International. 2002. Rwanda conflict vulnerability assessment. Prepared for USAID (October). Management Systems International. 2002. Rwanda democracy and governance assessment. Prepared for USAID (November). Mateeva, Anna. 2002. Minorities in Georgia: future challenges for development. Meek, Sara and Thokozani Thusi. 2003. Sierra Leone ­ Building the road to recovery. Institute for Security Studies, Pretoria. Moore, Karen, Chris Squire, and Foday MacBailey. 2003. Sierra Leone National Recovery Strategy ­ Assessment (December). Morazan, Pedro and Irene Knoke. 2005. PRSP is about people: experiences and reflections with partner organizations in five African countries. Church Development Service, Bread for the World, and World Vision. Murshed S. M. and S. Gates. 2003. Special horizontal inequality and the Maoist insurgency in Nepal. Mutebi, F., S. Stone, and N. Thin. 2003. Rwanda. Development Policy Review Vol. 21, No. 2. Mutebi, F., S. Stone, and N. Thin. 2001. Institutionalising the PRSP approach in Rwanda. Oxford Policy Management. Submitted to the Strategic Partnership with Africa (September). Mysliwiec, E. 1988. Punishing the poor: the international isolation of Kampuchea 1970-1987. London: Oxfam. Nepal Human Rights Yearbook. 2004. Informal Sector Service Center (INSEC). Kathmandu. Netherlands Institute of International Relations "Clingendael." 2003. Rwanda 2003 and beyond: prospects in stabilizing governance. Workshop Report. The Hague (February). Nkurunziza, Janvier. 2001. Economic determinants of violent conflict in Burundi. Nkurunziza, Janvier and Floribert Ngaruko. 2002. Explaining growth in Burundi: 1960-2000. Norwegian Refugee Council. 2003. Nepal: a hidden IDP crisis. Oslo (October). Nzabonimana, V. 2004. PRSP and civil society participation report: experience of Rwanda. International Training Programme, Christian Aid. London (May-June). Obidegwu, C. 2003. Rwanda: The Search for Post-Conflict Socio-Economic Change, 1995-2001. Africa Region Working Paper Series, No. 59. World Bank (October). Ojha, D. 2004. Inventory of poverty and impact monitoring (PIM) approaches to Nepal (draft, March). OECD. 2004. OECD Policy Brief: Security Sector Reform and Governance: Policy and Good Practice. Overseas Development Institute. 2003. National Poverty Reduction Strategies (PRSPs) in Conflict-Affected Countries in Africa. PRSP Briefing Note 6 (March). 72 Oxfam. (no date). Cambodia: Is the donors' medicine worse than the ailment? oxfamgb.org/eastasia/documents/CambPRSPfinal2.doc Oxfam. 2001. Cambodia's socio-economic development plan: issues for the NGO. London (June). Painter, G. 2002. Quality participation in poverty reduction strategies: experiences from Malawi, Bolivia and Rwanda. Christian Aid. London. Peou, S. and S. Ha. 2004. International Assistance for Institution Building in Post-Conflict Cambodia. Working Paper No. 26. Conflict Research Unit, Netherlands Institute of International Relations "Clingendael" (May). Protais, Musoni. 2003. Innovations in governance and public administration for poverty reduction in post-conflict countries in a globilised world: the experience of Rwanda. Paper presented at UN Ad Hoc Expert Group meeting on Innovations in Governance and Public Administration for Poverty Reduction (January). Putnam, Robert D., with Robert Leonardi and Raffaella Y. Nanetti. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Renard, R. and N. Molenaers. 2005. Civil society participation in Rwanda's poverty reduction strategy. University of Antwerp. Richards, Paul, Khadija Bah, and James Vincent. 2004. Social Capital and Survival: Prospects for Community- Driven Development in Post Conflict Sierra Leone. World Bank Social Development Papers No.12 (April). Royal Government of Cambodia. 2002. National Poverty Reduction Strategy (December). Scott, Alison. 2000. Poverty Reduction Strategies in Conflict Countries: How Are They Different? IMF Note. worldbank.org/poverty/strategies/review/semseries/scott.pdf Seddon, D. and K. Hussein. 2002. The consequences of conflict: livelihoods and development in Nepal. Overseas Development Institute, London. Shukla, K. 2004. Forgotten people: the internally displaced people of Nepal. Refugees International, Washington DC. Smillie, Ian. 1996. Service delivery or civil society? NGOs in Bosnia-Herzegovina. CARE Canada (December). Swedish International Development Cooperation Agency (SIDA). 2003. Reconciliation ­ Theory and Practice for Development Cooperation. Stockholm. Swedish International Development Cooperation Agency. 2004. A strategic conflict analysis for the Great Lakes Region. Stockholm (March). Thornton, Nigel and Marcus Cox. 2005. Developing Poverty Reduction Strategies in Low-Income Countries Under Stress (LICUS). Agulhas Development Consultants Ltd., London. Tommasoli, Massimo. 2003. Inequality, vulnerability to violent conflict, and aid incentives for peace. Organisation for Economic Co-operation and Development. Paris. Torres, Magui Moreno and Michael Anderson. 2004. Fragile States: Defining Difficult Environments for Poverty Reduction. DfID, Poverty Reduction in Difficult Environments Team. PRDE Working Paper 1 (August). United Nations (no date). Chad country profile. http://esa.un.org/socdev/unyin/country3a.asp?countrycode=t United Nations Development Programme. 2004. Human Development Report 2004: Cultural Liberty in Today's Diverse World. New York. 73 United Nations Development Programme. 2004. The macroeconomics of poverty reduction in Cambodia (March). United Nations Development Programme. 2002. Crisis in Nepal: a development response. Bureau for Crisis Prevention and Recovery. New York (May). http://www.undp.org/bcpr/smallarms/index.htm United Nations Development Programme. 2002. Human Development Report 2002: Deepening Democracy in a Fragmented World. New York. United Nations High Commissioner for Refugees. 2004. Poverty Reduction Strategy Papers ­ a displacement perspective. Geneva (October). United Nations Security Council. 2001. Report of the panel of experts on the illegal exploitation of natural resources and other forms of wealth of the Democratic Republic of the Congo (April). United States Agency for International Development. 2000. Chad 1999/2000 current vulnerability assessment. Famine Early Warning Project. Washington DC (March). United States Agency for International Development. 2002. Rwanda conflict vulnerability assessment (October). United States Department of State. 2004. Chad Country Report on Human Rights Practices. http://www.state.gov/g/drl/rls/shrd/2003/31014.htm#chad Van Loocke J. H. and L. Philipson. 2002. Report of the EC conflict prevention assessment mission, Nepal. European Commission, Brussels. VENRO. 2003. Poverty reduction and crisis prevention: how can poverty reduction be shaped conflict sensitively? Berlin (December). WIDER. 2004. Commune-Level Estimation of Poverty Measures and its Application in Cambodia. Tokyo: United Nations University (July). World Bank. 2002. Georgia poverty update. Washington DC (January). World Bank. 2002. Bosnia and Herzegovina local level institutions and social capital study. Economically and Socially Sustainable Development Department. Washington DC. World Bank. 2003. Conflict Analysis Framework (CAF). Conflict Prevention and Reconstruction Unit. www.worldbank.org/conflict. World Bank. 2005. Conflict in Somalia: Drivers and Dynamics. Washington DC (January). World Bank. 2004. OED review of the poverty reduction strategy (PRS) process: Cambodia case study. Operations Evaluation Department. www.worldbank.org.oed/prsp (July). World Bank. 2003. Bosnia and Herzegovina Poverty Assessment, Vol. 1 (November). World Bank. 2003. Chad Country Assistance Strategy with Chairman's Concluding Remarks, Washington DC (December). World Bank. forthcoming. Developing Poverty Reduction Strategies in Low-Income Countries Under Stress. LICUS Unit (Low Income Countries Under Stress). Washington DC. World Bank. 2004. The Poverty Reduction Strategy initiative: an independent evaluation of the World Bank's support through 2003. Operations and Evaluations Department. Washington DC (July). 74 Zaman, M. 2002. Are we getting lost in exclusive anti-poor, adjustment lending policy cycles? A rapid review of preliminary ActionAid engagement with Poverty Reduction Strategies in Kenya, Haiti, Uganda, Vietnam, Nepal, Rwanda and Malawi. ActionAid. Johannesburg. Zambelis, Chris. 2002. The South Caucasus: nationalism, conflict and minorities. 75