41452 THE WORLD BANK MINDING THE GAPS MINDING THE GAPS Integrating Poverty Reduction Strategies and Budgets for Domestic Accountability Vera A. Wilhelm and Philipp Krause, Editors © 2008 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 1 2 3 4 11 10 09 08 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 volume do not neces- sarily 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 judge- ment 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 permis- sion 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; Internet: 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. ISBN: 978-0-8213-7205-0 eISBN: 978-0-8213-7206-7 DOI: 10.1596/978-0-8213-7205-0 Cover photo: Idiz/Shutterstock Cover design: Word Express Library of Congress Cataloging-in-Publication Data Minding the gaps: integrating poverty reduction strategies and budgets for domestic accountability / editors, Vera A. Wilhelm, Philipp Krause. p. cm. Includes bibliographical references and index. ISBN 978-0-8213-7205-0 -- ISBN 978-0-8213-7206-7 (electronic) 1. Poverty--Government policy--Developing countries. 2. Budget--Developing countries. 3. Eco- nomic assistance--Developing countries--Management. I. Wilhelm, Vera, 1960­ II. Krause, Philipp, 1978. III. World Bank. HC59.72.P6M49 2007 362.5'561091724--dc22 2007036780 Contents Acknowledgments ix Abbreviations xi 1. Executive Summary 1 Background, Objectives, and Methodology 1 Domestic Accountability in the PRS and Budget Processes 2 Linking PRSs and Budgets 3 Integrating PRS and Budget Reporting 4 The Agenda Ahead 6 Part One: Findings and Lessons from Country Case Studies 2. Introduction 11 Background 12 Objectives, Scope, and Methodology 13 Structure of the Review 15 3. Domestic Accountability in the PRS and Budget Processes: Expectations and Reality 17 Accountability: Ownership and Incentives 17 Ownership of the PRS and the Budget 19 Incentives in the PRS and the Budget 24 Summing Up: Accountability Gaps in the PRS and the Budget 26 4. Linking Poverty Reduction Strategies and Budgets 29 Mechanisms to Link PRSs and Budgets 29 Toward a More Effective Link 34 Summing Up 37 5. Integrating PRS and Budget Reporting 41 The Nature of Budget and PRS Reporting 41 Three Building Blocks 46 External and Internal Incentives to Report 47 Summing Up 51 v vi Contents 6. The Agenda Ahead 53 Key Findings 53 Practical Lessons and Entry Points for Reform 55 Conclusion 61 Part Two: Summaries of Country Case Studies 7. Albania: Creating an Integrated Planning and Reporting System 65 Country Context 65 Linking the PRS and the Budget 67 Integrating Reporting on the PRS and the Budget 68 Ownership and Accountability 69 The Integrated Planning System 70 Conclusions and Lessons 71 8. Burkina Faso: Aligning Donor Procedures with National Systems 75 Country Context 75 Linking the PRS and the Budget 76 Integrating Reporting on the PRS and the Budget 78 Ownership and Accountability 79 Conclusions and Lessons 81 9. Madagascar: Toward More Effective Government Leadership 85 Country Context 85 Linking the PRS and the Budget 87 Integrating Reporting on the PRS and the Budget 88 Ownership and Accountability 90 Conclusions and Lessons 91 10. Malawi: Building a Credible Budget for Better Accountability 95 Country Context 95 Linking the PRS and the Budget 96 Integrating Reporting on the PRS and the Budget 99 Ownership and Accountability 100 Conclusions and Lessons 102 11. Mali: Linking Policies and Medium-Term Budgeting 105 Country Context 105 Linking the PRS and the Budget 106 Integrating Reporting on the PRS and the Budget 108 Ownership and Accountability 108 Conclusions and Lessons 109 Contents vii 12. Mozambique: Promoting a Unified Process 113 Country Context 113 Linking the PRS and the Budget 115 Integrating Reporting on the PRS and the Budget 117 Conclusions and Lessons 121 13. Rwanda: Domestic Accountability as a Driver of Integration 123 Country Context 123 Linking the PRS and the Budget 124 Integrating Reporting on the PRS and the Budget 127 Ownership and Accountability 128 Conclusions and Lessons 129 14. Tanzania: Developing a Structured Link between the PRS and Budgets 133 Country Context 133 Linking the PRS and the Budget 134 Integrating Reporting on the PRS and the Budget 136 Ownership and Accountability 137 Conclusions and Lessons 139 15. Uganda: Building on Existing Systems for Monitoring Results 141 Country Context 141 Linking the PRS and the Budget 142 Linking PRS and Budget Reporting 145 Incentives and Accountability 147 Conclusions and Lessons 147 Annex 1: Lessons from Higher-Income Reformers 151 Country Summaries 151 Emerging Themes 156 Implications 158 Annex 2: Priority Sectors and Pro-Poor Spending in Selected Case Studies 161 Annex 3: Nature of the APR in the Case Study Countries 163 Notes 167 Bibliography 173 Index 177 viii Contents BOXES 3.1 Budget Challenge Functions in Case Study Countries 22 4.1 The Strategic Budget Allocation System in Tanzania 31 4.2 Role of Planning and Budgeting Instruments in Uganda 33 4.3 The Leadership and Management Program in Madagascar 38 4.4 The Integrated Planning System in Albania 39 5.1 Budget Reporting in Uganda 42 5.2 Fragmentation and Duplication in the Health Sector in Mozambique 49 5.3 The Annual Action Plan Process in Rwanda 50 5.4 The Role of Political Commitment and Leadership for Informed Decision Making in Uganda 51 FIGURES 3.1 Drivers of Accountability 18 3.2 Asymmetries of Ownership in the PRS and the Budget 23 3.3 Multiple Fractures in Planning and Budgeting 27 5.1 Use of Performance Information in Reporting Instruments in Uganda 45 5.2 A Typical Imbalance in Incentives to Report to External Stakeholders 48 12.1 Mozambique: Planning and Budgeting Instruments 116 15.1 Uganda: Planning and Budgeting Instruments 143 TABLES 3.1 Participation in the Budget Process 21 4.1 Breaking the MTEF into its Constituent Parts 32 6.1 Targeting Reporting to Specific Decision-Making Processes 59 6.2 Integrating Performance Information into Budget Documents 60 13.1 Rwanda: Annual Budget Calendar 125 Acknowledgments This report was jointly directed by the World Bank (Vera A. Wilhelm) and Gesellschaft für Technische Zusammenarbeit--GTZ (Philipp Krause). It draws on a work program developed by Vera A. Wilhelm and was under- taken by a team further consisting of Paolo de Renzio and Tim Williamson (both with the Overseas Development Institute--ODI). Zoran Stavreski (for- merly with the World Bank) worked on the report in its early stages and provided useful comments during the technical review process. The work was carried out under the guidance of Louise Cord (World Bank), Luca Barbone (World Bank), Danny Leipziger (World Bank), Jan Walliser (World Bank), Matthias Witt (GTZ), Albrecht Stockmayer (GTZ), and Jörg-Werner Haas (GTZ). Peer reviewers were Matthew Andrews (Kennedy School, Harvard University) and William Dorotinsky (World Bank). We are grateful for useful comments received from Gregor Binkert (World Bank), David Booth (ODI), Bianca Brétéché (GTZ), Siaka Coulibaly (World Bank), Wilfried Engelke (World Bank), Gord Evans (consultant), Kene Ezemenari (World Bank), Guenter Heidenhof (World Bank), Omer Karasapan (World Bank), Stefan Koeberle (World Bank), Kathie Krumm (World Bank), Florian Lang (GTZ), Greta Minxhozi (World Bank), Allister Moon (World Bank), Alia El Moubayed (World Bank), Paul Mpuga (World Bank), Antonio Nucifora (World Bank), Allen Schick (University of Maryland), Abdoulaye Seck (World Bank), Carlos Silva-Jauregui (World Bank), Agnes Soucat (World Bank), and Linda van Gelder (World Bank). We would also like to thank Rachel Weaving and Catherine Hull for their inspired editing, Jae Shin Yang for production assistance, and Stephen McGroarty from the World Bank's Office of the Publisher. The study builds on background reports prepared in nine low-income countries in 2006. The reports were prepared by Artan Hoxha (Institute of Contemporary Studies, Albania), Sawadogo Malick (Burkina Faso), Veronique Hubert (Madagascar), Philipp Krause (Malawi), Verena Fritz and Florian Lang (Mali), Paolo de Renzio and Jose Sulemane (Mozambique), Vincent de Boer and Robert Cook (Rwanda), and Tim Williamson (Tanzania and Uganda). Topical briefs on the experience in four higher-income countries ix x Acknowledgments were written by Malcolm Holmes (Australia), Azul del Villar (Chile), Albert van Zyl (South Africa), and Sang Dae Choi (Republic of Korea). In addition to funding from the World Bank and GTZ on behalf of the German Federal Ministry for Economic Cooperation and Development (BMZ), the work program has been financed through grants from the Trust Fund for Environmentally and Socially Sustainable Development sup- ported by Finland and Norway and the Belgian Partnership for Poverty Reduction, which are gratefully acknowledged. For any questions, comments, or suggestions on this volume, please con- tact Vera A. Wilhelm (Vwilhelm@worldbank.org) or Philipp Krause (Philipp. krause@gtz.de). For further information on poverty reduction strategies please visit http://www.worldbank.org/prsp. For further information on GTZ's work on public finance and poverty reduction please see http://www. gtz.de/en/themen/politische-reformen/oeffentliche-finanzen/886.htm. Abbreviations AAP annual action plan APR annual progress report BdPES annual report on the social and economic plan (Mozambique) BP budget­programmes (budget for programs, Mali) BPR budget performance report CCM Chama cha Mapinduzi CCS Commitment Control System CDMT cadre de dépense à moyenne terme (medium-term expenditure framework) CFMP medium-term fiscal framework (Mozambique) CPIA country institutional and policy assessment CPO Comissão do Plano e Orçamento CSFP Comité de Suivi des Finances Publiques (budget monitoring committee, Madagascar) CSLP Cadre Stratégique pour la Lutte contre la Pauvreté (Poverty Reduction Strategy, Mali) CSO civil society organization ERC Expenditure Review Committee EU European Union FEs forward estimates GBS general budget support GDP gross domestic product GNI gross national income HIPC heavily indebted poor countries HIV/AIDS human immunodeficiency virus/acquired immune deficiency syndrome IDA International Development Association IEG Internal Evaluation Group (World Bank) IMF International Monetary Fund IPS Integrated Planning System (Albania) xi xii Abbreviations LAMP leadership action and management program LG local government M&E monitoring and evaluation MAP Madagascar Action Plan MDAs ministries, departments, and agencies MDGs Millennium Development Goals MEDEV Ministry of the Economy and Development (Burkina Faso) MEGS Malawi Economic Growth Strategy MEPD Ministry of Economic Planning and Development MFB Ministry of Finance and Budget MFPED Ministry of Finance, Planning, and Economic Development MGDS Malawi Growth and Development Strategy MIS management information system MoF Ministry of Finance MPRS Malawi Poverty Reduction Strategy MPRSP Malawi Poverty Reduction Strategy Paper MTBP Medium-term Budget Program MTEF medium-term expenditure framework NAO National Audit Office NATO North Atlantic Treaty Organization NGO nongovernmental organization NSDI National Strategy for Development and Integration (Albania) NSSED National Strategy for Social and Economic Development ODA official development assistance OE Orçamento do Estado PAF performance assessment framework PAP Priority Action Program (Burkina Faso) PARPA Plano de Acção para a Redução da Pobreza Absoluta (Poverty Reduction Strategy, Mozambique) PEAP Poverty Eradication Action Plan PED Poverty Eradication Division PEFA Public Expenditure and Financial Accountability PEIR public expenditure and investment review PES Plano Económico e Social PESS Plano Estratégico do Sector Saude PGE Politique Générale de l'Etat (Madagascar) PPEs protected pro-poor expenditures Abbreviations xiii PQG Programa Quinquenal do Governo (Five-year Plan, Mozambique) PRS Poverty Reduction Strategy PRSP Poverty Reduction Strategy Paper REO budget execution report (Mozambique) RPF Rwandan Patriotic Front SAA Stabilization and Accession Agreement (European Union) SBAS Strategic Budget Allocation System SISTAFE Sistema da Administraçåo Financeira do Estado (integrated financial management information system, Mozambique) SNLP Stratégie Nationale pour la Lutte contre la Pauvreté (National Strategy for Poverty Reduction, Mali) STA Secretariat Technique de l'Ajustement (Madagascar) SWAp sector-wide approach 1 Executive Summary Background, Objectives, and Methodology The Poverty Reduction Strategy (PRS) approach aims to enhance account- ability by promoting the participation of domestic stakeholders in the formulation of clear and realistic development goals. Institutions for moni- toring and evaluation, including annual progress reports, are designed to trigger learning and improved performance. Links between the PRS and budget, whether at the formulation, execu- tion, or reporting stage, have been considered integral to the successful implementation of the PRS for some time. These links are also increasingly recognized as vital for enhanced domestic accountability. This study offers practical insights for donors and national govern- ments on how to strengthen the links between PRSs and national budgets, with a view to improving domestic accountability. It aims to answer two principal questions: · What challenges have arisen in countries where efforts have been made to integrate the PRS with the budget? · What lessons have been generated by these experiences and what are the potential entry points for reforms to strengthen PRS-budget links? To answer these questions, the study reviews a series of case studies that document the status of budget and PRS integration in a sample of nine low-income countries--Albania, Burkina Faso, Madagascar, Malawi, Mali, Mozambique, Rwanda, Tanzania, and Uganda--and the links among policies, budgets, and service delivery in four higher-income countries that are internationally considered to be successful reformers in public financial management--Australia, Chile, the Republic of Korea, and South Africa. 1 2 Executive Summary Domestic Accountability in the PRS and Budget Processes The accountability relationship has three dimensions: ownership of the task; incentives to carry out the task effectively; and capacity to deliver results. While capacity issues in PRS implementation and public financial management reform have received a good deal of attention in previous reviews, and rightly so, this study focuses on issues related to ownership and incentives, because all three dimensions are required for successful development interventions. Both the PRS and budget processes can enhance domestic ownership by fostering transparency and dialogue between the executive and stakeholders in the sectors, parliament, and civil society. Typically, however, patterns of ownership are asymmetrical. For instance, while the executive and parlia- ment tend to be more involved in the formulation of the budget than in the PRS, civil society and the public often have less opportunity for meaningful engagement in the former than the latter. Thus, one of the principal chal- lenges in promoting links between planning and budgeting processes rests on aligning ownership across various groups of stakeholders. The PRS is intended to boost performance incentives by providing a results orientation to strategic planning and policy making. The low-income case study countries with second-generation PRSs (such as Burkina Faso, Rwanda, Tanzania, and Uganda) have placed significant emphasis on developing results-oriented strategies, influenced by the strong inter- national focus on development results and achieving the Millennium Devel- opment Goals (MDGs). This has generated more attention to the use of performance indicators and targets, and to the improvement of information systems. Alignment of donor monitoring and reporting requirements around national PRS monitoring systems remains a challenge, however. Institutional arrangements in the budget can facilitate performance incentives by enabling the scrutiny of resource allocation decisions, budget execution, and outcomes. According to heavily indebted poor countries (HIPC) and country institutional and policy assessment (CPIA) indicators, as well as a recent Global Monitoring Report developed under the Public Expenditure and Financial Accountability program, public financial man- agement in low-income countries has improved in recent years. National governments require deeper support and capacity development if they are to accelerate these recent improvements, because poor budget formulation and the weak enforcement of rules continue to undermine incentives to execute the budget effectively in a number of low-income countries. In sum, while both the PRS and budget offer scope for enhanced domestic accountability, fractures in planning and budgeting systems pose Executive Summary 3 obstacles for donors and national governments. The following challenges emerge from our case study countries: the prioritization of plans and coor- dination between planning and budgeting units; the creation of incentives to formulate realistic budgets and execute them as planned; the expansion of ownership of the PRS at the sector level; the development of a multi- annual perspective in strategic resource allocation; and the integration of reporting mechanisms. Linking PRSs and Budgets Since the instigation of the PRS approach, donors have encouraged the introduction of a number of technical instruments aimed at improving the integration of planning and budget processes. These include (i) the enforce- ment of pro-poor spending priorities; (ii) the introduction of medium-term expenditure frameworks (MTEFs); and (iii) results-oriented program bud- geting. These reforms are at different stages in different countries, reflecting the level of development of public financial management systems. Early efforts in countries such as Madagascar, Malawi, and Mali have tended to focus on narrow technical solutions to the challenge of linking PRSs and budgets, while the objectives of forging ownership beyond a narrow group of technocrats and creating adequate institutional incentives for integration have often been neglected. Across countries, reforms that have taken account of the interests of a broad set of stakeholders have been able to generate the strong incentives for the behavioral changes that are needed for successful implementation. Despite the difficulties that the majority of countries encountered while trying to effectively implement highly technical reforms, some of the low-income case study countries have achieved important successes. In Tanzania, a Strategic Budget Allocation System (SBAS) has enabled a strong link between ministry priorities and PRS goals. In Uganda, the MTEF and budget process have been effectively aligned at the strategic level. And in Mali and Uganda, governments have developed systems for program budgeting around preexisting institutions. If the goal is to move beyond narrow technical solutions to broader reforms, three principal lessons (derived from the experiences of the higher- income countries and the better performing case study countries) may be instructive. First, domestic ownership of the reform agenda is often crucial and may need to extend beyond the ministries of finance to the cabinet, and, in some cases, parliament and civil society. Second, an integrated approach to reform rests on strong institutional connections between planning and 4 Executive Summary budgeting. And third, the most effective reforms begin with getting the basics of public financial management right, before moving on to more sophisticated, performance-oriented reforms. Each of these lessons is also reflected in the most successful reform expe- riences in the low-income case study countries. An innovative and--by early accounts--successful approach to building capacity, ownership, and incentives at the core of government has been applied in Madagascar in the form of a Leadership and Management Program. The Albania case study points to the importance of increased PRS ownership developed under that country's integrated planning system for budget formulation and implementation. Tanzania's initiative to harmonize planning and bud- geting instruments demonstrates sensitivity to the goal of integrated reforms. Mozambique's efforts to avoid duplicating reporting instruments for donors and domestic stakeholders have helped to strengthen domestic processes and to promote integration. Last, relatively informal approaches to results-oriented budgeting in Mali and Uganda have arguably enjoyed more success than more formal approaches, because they have been in step with, rather than ahead of, the level of sophistication of the budget. Integrating PRS and Budget Reporting A well-functioning, results-oriented reporting system needs to integrate financial information on actual expenditure inputs with performance infor- mation on the outputs and outcomes that result from public spending. While the PRS goes beyond the budget, because it includes strategies and priority actions that do not require the allocation of public funds, the systems for mon- itoring government performance in implementing budgets could provide a solid basis for reporting on a critical part of PRS implementation. From the case study countries, three essential building blocks for the integration of PRS and budget reporting emerge: · Regular reporting from various government implementing institutions, such as central ministries/agencies and local governments, is needed. · More advanced sectoral monitoring and reporting processes, particu- larly in health and education, would be well advised. · Improvements in surveys and statistics on poverty outcomes on a more regular basis would assist in assessing government performance. Although the three elements for successful integration may be lagging due to inadequate capacity, they depend more fundamentally on bureau- cratic and political incentives to integrate financial and performance reporting. Executive Summary 5 They thus depend on two forms of incentives: first, incentives to produce information and reports, using both financial and performance information; and second, incentives to link such information and analysis with policy- making processes. In most case study countries, incentives to generate information come pri- marily from donors. This imbalance is particularly evident in annual progress reports (APRs), which have been largely a response to external demands for information on PRS implementation, first linked to the HIPC initiative and subsequently in the context of direct support to government budgets. As a result, APRs tend to be de-linked from domestic reporting mechanisms. Among the few countries that have used existing annual reports as the basis for PRS reporting, Mozambique has successfully generated a domestically owned product, primed to be integrated into policy processes. Excessive external demands for reporting compound the risks of frag- mentation and duplication. For instance, at the national level in Rwanda, the report on the annual action plan, joint sector reviews, the APR, and budget support reviews could benefit from better streamlining and inte- gration. External requirements can also undermine the legitimacy of gov- ernments' own reporting systems. For instance, sector-wide approach (SWAp) mechanisms in Mozambique have set up reporting systems that run parallel to mainstream budgetary systems and distract attention from national processes. While donor-driven demands have created significant incentives to supply information, they have provided fewer incentives to use this information in an integrated reporting process. Despite the fact that much of the information produced for compliance is potentially powerful for informing decisions, a large amount of reporting is done simply to comply with donor-related or statutory requirements. As a consequence, little effort is put into analyzing the information collected or applying analysis to policy making. Experience with successful reporting mechanisms in the higher-income countries emphasizes the importance of core policy-making processes. Most of these processes relied on cabinet-level structures--such as committees and review bodies--that discussed fiscal frameworks and budget priorities and monitored the execution and performance of the overall budget. These structures ensured the direct engagement and involvement of politicians (mostly ministers, but in some cases also parliamentarians) at key decision points, thus ensuring the necessary buy-in and political peer pressure for "sticking to the path." This engagement allowed the use of information in policy- and decision-making processes, with positive effects on the design of programs and sectoral policies. 6 Executive Summary The Agenda Ahead Integrating PRSs, budgets, and their corresponding reporting processes is challenging but has the potential to strengthen government accountability to citizens and the implementation of more effective policies. Handling the multiple difficulties inherent in integration demands a more effective approach to two dimensions of domestic accountability: ownership and incen- tives. An improved approach would also require a renewed commitment from donors to build on what exists and to focus on reforms with potential political traction. Four lessons, in particular, can be gleaned from the experiences of our country case studies. Lesson 1: Focus on strengthening and harmonizing existing processes and adopt a step-by-step approach to reform. The enhancement of the policy-making function of the budget in line with PRS priorities often requires a flexible approach to PRS design, building on existing processes and tailoring interventions to local context and realities. An increased focus on results requires not only technical capacity but also a change in culture and thinking, in the sense of organizational and institu- tional capacity development. Experience in the case study countries shows that successful budget reform processes tend to start simply, evolve gradually, and be framed within an integrated approach linking improvements in public financial management with broader civil service reform. In essence, short- cuts tend not to work. Lesson 2: Build support from within, through high-level ownership of poli- cies, a challenge to the executive, and clear roles for sectoral ministries. Intra-governmental accountabilities, starting at the political (cabinet) level, can be a powerful tool in promoting reform and have so far received insuffi- cient attention. Initiatives that may help to catalyze the improvement of core policy processes and internal accountability include · Introducing or reforming cabinet committee structures, to strengthen high-level political ownership and a challenge function within the executive. · Building policy links "from the budget up," rather than "from the PRS down." Successful reform efforts have focused on core policy processes, and on turning budgets into more policy-oriented tools, rather than starting from long-term plans and translating them into annual budgets. The Poverty Reduction Strategy Paper (PRSP) Executive Summary 7 should therefore guide resource allocations, but not be the basis for the structure of the budget. · Introducing a strategic phase in the budget process to ensure that there is a stage, before the preparation of detailed operational budgets, at which sector managers and the cabinet review high-level policy prio- rities and their impact on inter-sectoral resource allocation. · Developing sector-level policy processes. While the PRS should be seen as a mechanism for developing high-level policy priorities, sec- tor ministries and institutions should be charged with establishing processes to elaborate sector policies and budgets and review policy implementation, and their roles should be clearly defined with respect to larger national objectives. Lesson 3: Foster incentives for integration: target reporting to decision- making processes. Reporting instruments best serve their purpose when they are instigated by, and remain linked to, particular decision-making processes involving spe- cific actors who are likely to demand and use the information. Strengthening reporting on budget execution, rather than focusing on externally driven APRs, should be given priority, introducing outcome information alongside financial reporting whenever possible. Providing easy access to quality information helps decision makers to monitor policy implementation and take corrective measures. Better target- ing of reports, both financial and performance related, to users' needs could facilitate and improve decision making. This study provides a relatively simple typology of how various reports could be targeted to the different audiences that demand them, including donors, finance ministries, parlia- ments, and presidents. Lesson 4: Keep it simple. A strengthened results orientation within the budget need not require sophisticated technical solutions. As the case studies show, comparatively simple budget reforms can significantly improve the budget's responsive- ness to policies. Such reforms start with making budget execution credible and transparent. Structuring a PRSP in a more budget-friendly manner, for example by sector, would facilitate the interface with the budget by involv- ing sector agencies in elaborating policy priorities and establishing the resource implications. It could expand ownership and boost incentives for integration of a greater number of key stakeholders, thereby strengthening domestic accountability. Part One Findings and Lessons from Country Case Studies The following synthesis of the case studies is based on a draft prepared by Paolo de Renzio, Vera A. Wilhelm, and Tim Williamson. 2 Introduction The World Bank and the IMF introduced Poverty Reduction Strategies (PRSs) in 1999, as a new approach to engagement with low-income countries. They provide a framework for these countries to articulate their development priorities and to specify the policies, programs, and resources needed to meet their goals. PRSs were expected to be poverty focused, country driven, results oriented, and comprehensive. They were also expected to serve as a framework for better coordination of development assistance among devel- opment partners and encourage donors to provide predictable, harmonized assistance, aligned with country priorities. The PRS approach has become widespread. Between late 1999 and 2006, 51 countries prepared national PRSs; several countries are in the process of revising their original strategies; and Burkina Faso, Ghana, Mauritania, Mozambique, Nicaragua, Senegal, Tanzania, Uganda, and Vietnam have already done so. Five more countries have produced interim strategies and ten have started a process that could result in a PRS. The 2005 Review of the PRS Approach (IMF and World Bank 2005) established that PRSs have helped to provide a framework for low-income countries to articulate their development priorities and to specify the policies, programs, and resources needed to meet their goals. In general, the PRS approach has helped countries to (i) focus more squarely on poverty reduction in formulating and implementing their development strategies; (ii) open up the participatory process; (iii) improve public expenditure management systems and better define and protect poverty-reducing expenditures; and (iv) focus more attention on monitoring poverty-related outcomes. In countries where PRSs are reasonably well articulated and govern- ments have assumed a strong leadership role, the PRS approach has started 11 12 Introduction to bring about the intended shift in the relationship between developing countries and their external partners. In these cases, strategies tend to be better linked to budget processes, monitoring arrangements are more effec- tive, and participatory processes have started to be institutionalized. In other cases, however, PRS formulation and implementation have focused more on donor requirements and expectations (external accountabilities) without enough attention to the accountability of governments to domestic constituents. The 2005 review identified a number of areas where sustained efforts by countries and development partners were needed to further enhance the effectiveness of the PRS and strengthen domestic accountability. Of particular importance was the need to build better links between the PRS and domestic processes. Linking the PRS to a medium-term expenditure framework and the budget process encourages prioritization and enhances country ownership and customization by integrating the PRS with domestic decision-making processes (IMF and World Bank 2005: paras. 39­44). Equally important are expanding participation through greater involvement of sectoral ministries, local authorities, and parliaments; and strengthening monitoring and evaluation systems, including identifying intermediate indicators to monitor results. The 2005 review led to a number of follow-up actions in the area of PRS implementation, monitoring, and reporting,1 including the following review of the links between PRSs and budgets based on the findings of nine country case studies. Background As designed, the PRS approach was expected to support domestic account- ability by clearly framing development goals and setting targets for develop- ment results. Increased accountability of public institutions to their citizens was anticipated through a deeper and more critical engagement between governments and national civil societies.2 Links between national strate- gies and the annual and multi-annual budget processes were expected to enable the delivery of development priorities. And adequate institutions for monitoring and evaluation were expected to trigger learning and improved performance. In this context, PRS annual progress reports (APRs) were envisaged as crucial vehicles for accountability, enabling changes in policies or implementation approaches to deliver results. Experiences with PRS implementation in recent years suggest that links between the PRS and budget are often weak and that the account- ability of governments to their citizens could be strengthened, in terms of policy choices, resource allocation, and implementation. APRs that have Introduction 13 built upon existing domestic mechanisms have helped to resolve some of these accountability gaps, while those that run parallel to domestic systems in combination with high aid levels have risked over-burdening domestic institutions and diverted attention from domestic political processes and local capacity. To address these problems and improve the usefulness of the PRS as a policy document that guides spending allocations, the rela- tionships among PRSs, budget processes, and reporting mechanisms are now receiving more attention. The view that PRSs are less likely to be successfully implemented or to achieve development results unless they are linked to budget processes and influence budget allocations is well established (SIDA 2006: 55). The reason for this link is quite simple: many of the policies and strategies embodied in the PRS need to be resourced, which means they must be reflected in the country's budget. Once included in the approved budget, the funds need to be distributed and spent as intended. Thus, in recent years, a good deal of attention has been paid to strengthening budgets, not only to increase their comprehensiveness, credibility, and transparency, but also to make them more outcome-oriented and thus more amenable to linkage with PRSs (Holmes and Evans 2003; World Bank 2006a). The role of the budget as a possible vehicle for enhanced domestic accountability has only recently begun to receive significant attention (World Bank/IDA and IMF 2005; GTZ 2005; GTZ 2006). According to the 2005 PRS Review, the national budget and the processes that led to its implementation are critical vehicles for ensuring country ownership and leadership and for strengthening a government's accountability to citizens (World Bank/IDA and IMF 2005: 1,15,19; GTZ 2005). This reflects a shift in the central focus of discussion of PRS-budget links, away from the budget as a tool to implement a development vision and toward the budget as a policy process with the potential to increase accountability to domestic stakeholders. Objectives, Scope, and Methodology The primary objective of this book is to examine how closer links between the PRS process and the budget could help to strengthen rather than side- line domestic systems, and result in overall gains for domestic account- ability, both in terms of promoting ownership and of generating incentives for performance. The study aims to contribute to understanding of an increasingly important, but relatively unexplored, challenge for the development com- munity and policy makers: the promotion of PRS-budget links as tools of 14 Introduction domestic accountability. Under this overarching objective, it addresses two main questions: · What challenges have arisen in countries where efforts have been made to integrate the PRS with the budget? · What are the lessons of this experience and the potential entry points for actions to strengthen these links? With a view to answering these questions the book reviews the experi- ences of nine low-income countries--Albania, Burkina Faso, Madagascar, Malawi, Mali, Mozambique, Rwanda, Tanzania, and Uganda--as well as four successful higher-income reformers--Australia, Chile, South Africa, and the Republic of Korea. A relatively simple framework of accountability (focusing on ownership and incentives) is employed to assess planning and budgeting processes in these countries. The book reviews reforms that have been implemented to improve PRS-budget links and assesses the extent to which they have successfully increased country ownership and boosted the incentives for good performance. The book pays particular attention to the role of integrated reporting processes as a highly promising avenue for enhancing domestic account- ability in low-income countries. Basing PRS annual reporting on the country's budget reporting would enable financial information and performance infor- mation to be combined, providing the foundation for an integrated assessment of government policies, activities, and actual outcomes. It would have the additional advantage of building on the country's existing systems. The review has a strong focus on country context. The nine low-income case study countries were selected based on a variety of criteria. All of them except Albania receive significant amounts of aid,3 use medium-term expen- diture planning frameworks or program budgets (although at different stages of implementation), and have benefited from sector-wide arrange- ments (sector-wide approaches [SWAps], in a few sectors) that have had a generally positive impact on sector strategy formulation. Despite these commonalities, the countries vary widely as to budget formulation, execu- tion, and monitoring4; the level of performance orientation of the budget; and the institutional arrangements around the PRS and budget process. As to PRS-budget links, only a few countries have employed domestic reporting documents and processes5 for PRS reporting. Most of the developing country case studies were written by country- based authors with a good understanding of the local context, and those for Mozambique, Rwanda, and Uganda were written by authors who have been directly involved in the countries' planning and budgeting processes. Introduction 15 Four short briefs were also prepared on experiences of successful budget reform and development of accountability in the higher-income countries to inform lessons and recommendations. (Annex 1 describes their experience in more detail.) The findings reported from the case studies are not meant to be compre- hensive or to serve as a blueprint for linking PRSs and budgets, but are intended to provide insights on the challenges and entry points for donor support and institutional reform.6 Limitations in the case studies, due to a certain unevenness in coverage, diverging country contexts, and varying degrees of complexity, need to be acknowledged. Structure of the Review Chapter 3 lays out a simple framework based on the idea that ownership of PRSs and budgets by domestic actors, combined with positive institu- tional incentives, can boost domestic accountability. Considering the case study experiences and recent literature, the chapter assesses the extent to which accountability has been enhanced through the PRS and budget in practice and highlights the accountability gaps that typically preclude the forging of closer links between the two processes. Chapter 4 explores how general efforts to link PRSs with the budget have worked in the past. It argues that narrow technical reforms generally enjoy less success than gradualist and integrated reforms, which gain impetus from ownership by core groups in government. Chapter 5 applies these insights to the particular issue of PRS and budget reporting, indicating the special role that budget reporting can play in domestic accountability, provided that incentives to both generate and use performance and fiscal information are aligned. The final chapter identifies a series of entry points for linking PRS and budget processes that is realistic about the constraints but could help to strengthen domestic accountability. 3 Domestic Accountability in the PRS and Budget Processes: Expectations and Reality This chapter introduces a simple framework for explaining domestic accountability in terms of ownership and incentive structures. It then dis- cusses the expectations for, and reality of, accountability mechanisms within the Poverty Reduction Strategy (PRS) and budget processes, drawing on findings in recent reviews of PRS countries (such as IMF-World Bank 2005; GTZ 2005; Entwistle and Cavassini 2005; World Bank IEG 2004) and findings in the nine case study countries. Accountability: Ownership and Incentives Drivers of Accountability In its simplest form, accountability is the responsibility of one party (the agent) to another (the principal) for delivering on a specific task or set of tasks.1 The accountability relationship has three dimensions (Figure 3.1): · Ownership: First, accountability requires the agent to consider itself responsible and willing to carry out the tasks. Meanwhile, the princi- pal needs to be interested in the tasks and demand that the agent carries them out. · Incentives: Second, the principal needs to have adequate powers to ensure that the agent carries out the tasks, or at least to make it in the interest of the agent to do so. Incentive mechanisms, including rewards and sanctions, are therefore crucial for ensuring delivery.2 17 18 Domestic Accountability in the PRS and Budget Processes: Expectations and Reality Figure 3.1 Drivers of Accountability incentiv ship principal demands principal has powers accountability from to ensure that agent es wnero agent for x delivers x strong accountability agent considers it is in the agent's itself accountable capacity interest for x the agent has the to deliver x ability to deliver x Source: Authors. · Capacity: Third, for any accountability relationship to deliver results the agent must have the means and be able to undertake the responsibility. Real accountability mechanisms are complex, however. To usefully apply this stylized framework to planning and budgetary processes, it is necessary to recognize the variety of actors and institutions involved in pol- icy formulation (for example, central and sector agencies within govern- ment, parliaments, civil society groups, donors), and the fact that often the tasks involved can be of a different nature (for example, related to the PRS, to sector policies, or to the budget). Different accountability mechanisms are at work at any one moment, such as · Parliament (principal) holds the executive (agent) accountable for implementing the budget as approved. · Donors (principals) hold the government (agent) accountable for using aid resources in accordance with PRS priorities. · Citizens (principals) hold politicians in parliament (agents) account- able for delivering on electoral promises. · The Ministry of Finance (principal) holds sector ministries (agents) accountable for reporting regularly and reliably on funds spent and results achieved in each sector. Domestic Accountability in the PRS and Budget Processes: Expectations and Reality 19 Unpacking these complexities, and the ways in which they determine incentives, is therefore important for reaching a more systematic under- standing of why accountability relationships are weak in both PRS and bud- get processes and how they can be strengthened. Because much has been written about the need to address inadequate capacity in PRS implementa- tion (World Bank 2006b: 63­66; World Bank IEG 2004: 29­37; World Bank and IMF 2002: 14­33) and public financial management reform (Levy and Kpundeh 2004; Bennet 2004; Anderson and Isaksen 2002). The next two sec- tions draw on the literature and try to take a slightly different angle, paying particular attention to ownership and incentive structures--both in theory and in practice--in PRSs and budget processes. Ownership of the PRS and the Budget While mechanisms to enhance domestic ownership have played a central role in the design of PRSs, experiences with their implementation have been mixed. Relative to budgets, PRSs have tended to evoke stronger ownership outside the government and weaker ownership within it. Asymmetries of ownership have resulted in obstacles to integrating these two systems. Ownership of the PRS The PRS process was designed to be country owned and participatory, supporting the accountability of the government (the agent) to its citizens (the principals). Ownership by the executive was meant to be reflected in the PRS being adopted as the government's main development strategy, and in PRS- related processes being integrated with other domestic planning and budget- ing mechanisms (Entwistle and Cavassini 2005: 10). In turn, civic engagement and participation in the PRS process was expected to widen ownership and consequently enhance accountability, as stakeholders' priorities came to be reflected in policy documents, and policy making became more transparent and inclusive (Tikare and others 2002: 238). Within government, the level of ownership in PRS processes has varied across groups of stakeholders. A 2004 evaluation by the World Bank's Inter- nal Evaluation Group (IEG) found that ownership is strongest in central ministries that are close to PRS preparation and review, and tends to fade in sector ministries, especially those with low PRS priority, and in regional and local administrations (World Bank IEG 2004).3 The PRS initiative tends to be driven by technocrats working in either planning or finance ministries; their degree of ownership tends to be high, given their heavy involvement in the 20 Domestic Accountability in the PRS and Budget Processes: Expectations and Reality formulation of the PRS. Cabinets have an incentive to ensure PRSs are prepared because of the development assistance associated with this instru- ment, but this does not guarantee their sense of ownership. Parliaments' involvement was generally found to be weak; in most countries Poverty Reduction Strategy Papers (PRSPs) were not approved by parliaments (this is also true in our case study countries). Civil society participation in the PRS process has been more variable, depending on the extent to which participation has been institutionalized, as well as on countries' political structures, civil freedoms, and the organiza- tional and communications skills of civil society organizations (CSOs) (IMF-World Bank 2005: 26). This is also borne out by the case studies. In Tanzania for example, CSO involvement is strong in PRS formulation and implementation, through participation in regular government-organized forums to review progress in implementation. In Madagascar, civil society monitoring is formalized through CSO participation in working groups that prepare monitoring reports related to the PRS; however, the impact of this arrangement on Madagascar's budget allocations seems to have been limited--at times because of capacity constraints, but also because civil soci- ety groups often have difficulty getting access to decision makers.4 Ownership of the PRS by stakeholders can also be compromised by the existence of alternative, and often competitive, national development strate- gies. Among the case studies considered here, the PRS is either the sole existing medium-term national strategy for poverty reduction (Burkina Faso, Malawi), or, where other strategies exist, it is widely or increasingly integrated with them (Mozambique, Tanzania, Uganda). The exceptions are Albania and Madagascar, where initially competing documents existed, but efforts are under way in each country to develop a unified, nationally owned strategy reflecting PRS principles. Ownership of the Budget Process The budget process is based on accounting, reporting, and auditing require- ments that should be built into one cycle to ensure good fiscal outcomes and avoid the mishandling of funds (IMF 2001). These various mechanisms ought to extend ownership of the budget process beyond the Ministry of Finance-- enabling government ministries, departments, agencies, and parliament to participate in decisions surrounding the allocation and expenditure of funds and allowing civil society and the public to oversee outcomes and voice their preferences. Relative to the PRSP, budgets tend to have a higher degree of political (as well as technical) ownership, given that they are prepared within the Domestic Accountability in the PRS and Budget Processes: Expectations and Reality 21 government and approved by parliaments. However, the case studies indi- cate that the depth of participation will vary according to institutional structures around the budget process. Table 3.1 and Box 3.1 summarize the extent to which institutional arrangements enable ministries, departments, and agencies to participate in the budgetary processes in some of the case study countries. Clearly delineated budgetary calendars that allow ample time for digestion and discussion of the budget, among other institutional arrangements, will determine the extent to which various actors are able to participate in the process (Table 3.1). Except for Malawi and Madagascar, where informality has tended to prevail, all the case study countries have established budgetary calendars that are broadly followed, and most actors are familiar with the process. The "challenge function" represents another vital element for the promotion of meaningful participation, especially of parliament (Box 3.1). Table 3.1 Participation in the Budget Process Dimension Uganda Mozambique Existence of and adherence A clear annual budget calendar There is a clear budget to a fixed budget exists, but some delays are calendar in the SISTAFE law calendar. often experienced in its and regulations. Apart from implementation. The calendar occasional delays, this allows MDAs reasonable time calendar is generally (at least four weeks from followed, and gives receipt of the budget circular) agencies two months to so that most of them are prepare detailed budget able to meaningfully proposals. complete their detailed estimates on time. Clarity/comprehensiveness A comprehensive and clear Decisions about budget of and political (cabinet) budget circular is issued to ceilings are not approved involvement in the guid- MDAs, which reflects ceilings by cabinet, but delegated ance on the preparation approved by cabinet (or to the minister of finance. of budget submissions equivalent). This approval Cabinet only approves (budget circular or takes place after the circular budget estimates once equivalent). distribution to MDAs, but agencies have submitted before MDAs have completed their proposals and these their submission. have been consolidated by the Ministry of Finance. Source: 2005 Public Financial Management Assessments using PEFA performance measurement framework. 22 Domestic Accountability in the PRS and Budget Processes: Expectations and Reality Box 3.1 Budget Challenge Functions in Case Study Countries The "challenge function" in the budget process is a key component of hierarchical and horizontal accountability. Of great importance in this respect is the space that the bud- get calendar affords to three levels of budget challenge: (i) for the Ministry of Finance to assess ministries' budget proposals; (ii) for the cabinet to make strategic decisions before the detailed budget is prepared; and (iii) for parliament to scrutinize allocations and propose amendments before approval. In the first stage, unsurprisingly the strength of the Ministry of Finance, and in particular the proactive role of the budget department, is important for the technical challenge function, and can play an important role in promoting PRS priorities during formulation, such as, for example, in Mali. Conversely, a combination of the PRSP being prepared by the planning ministry, and a weak budget department, has under- mined the budget challenge and links with the PRS in Malawi. The latter two levels of budget challenge (cabinet and parliament) appear central to political ownership of the budget processes, but they are often collapsed into one. For example, in both Madagascar and Mozambique the budget is presented to the cabinet only 10­15 days before its submission to parliament. This limits the space for meaningful cabinet discussion and decision making. The process is also conflated in Mali, but there the cabinet has more time for discussion. This contrasts with arrangements in Albania, Tanzania, and Uganda, where cabinet deliberates on ministries' and/or sector ceilings long before the budget is presented to parliament. However, what really counts is the level to which the cabinet discusses resource allocations in the context of wider policy choices. The way the budget is presented is important for facilitating such a debate. In Tanzania the budget is pre- sented in a highly technical way, and the budget ceilings presented for approval by cabinet are effectively a fait accompli, and changes are rarely made. In Uganda, the cabinet is presented with a spending options table, which serves as an attempt to present policy choices. Although the formulation of the budget tends to command more political involvement from the executive and parliament than does the formulation of the PRS, budgeting usually remains a largely technical exercise. This may in part explain a generally lower level of ownership outside government: the technical and secretive nature of the budget processes often affords civil society little opportunity for meaningful engagement.5 Asymmetrical Owners in the PRS and Budget Most of the case study countries have a multiplicity of institutions, actors, and interests in the planning and budgeting process, with different actors presiding over a variety of instruments and initiatives, and blurred mandates. In Tanzania, for example, before the Kikwete government took power, no Domestic Accountability in the PRS and Budget Processes: Expectations and Reality 23 fewer than four central government ministries were involved in overseeing different elements of the planning and budgeting processes: planning, public service, finance, and the vice president's office. Many other countries have separate planning and finance ministries. Albania, Mozambique, and Uganda have relatively unified responsibilities for coordinating planning and budgeting, although in both Mozambique and Uganda central plan- ning institutions have recently been (re)established. In Burkina Faso, Tanzania, and Uganda, central policy-making functions at the cabinet level tend to be relatively strong. In the other six low-income case study countries, however, the relatively weak policy-making function of cabinets helps to explain the narrowly technical quality of budget preparation and the overall weakness of the challenge function in the budget process. Figure 3.2 conceptualizes the typical patterns of ownership in the PRS and the budget. These asymmetries of ownership mean that different principals and agents--with different agendas, priorities, and limitations--dominate the PRS and budget processes, respectively. Because the planning and finance ministries--the main owners of the respective processes--are separate insti- tutional entities in a number of countries, there is no automatic incentive for Figure 3.2 Asymmetries of Ownership in the PRS and the Budget The PRS The budget executive executive political parliament political parliament cabinet Finance cabinet Ministr y technical technical Line planning finance planning Mfinance& inistries ministry ministry ministry ministry Providers sector sector ministries & provider ministries & provider societal societal civil civil society electorate society electorate groups groups strong ownership moderate ownership weak ownership Source: Authors. 24 Domestic Accountability in the PRS and Budget Processes: Expectations and Reality strong coordination between the two. Even if civil society is a moderately strong presence in the PRS process, its weak pattern of ownership over the budget will mean that it is unlikely to be an effective source of demand for greater links. And because parliament and the cabinet are relatively weak owners of the PRS process, they are unlikely to focus on PRS priorities when reviewing the budget, even when their role in the latter may be moderately strong. In the many case study countries that have these asymmetries of ownership, it is not surprising that while the focus on PRS may have increased in the budget process at the technical level (i.e., identification of spending, budget lines that broadly fit with PRS objectives), the budget process typically gives short shrift to debate and decision making geared toward delivering (pro-poor) policies. Incentives in the PRS and the Budget A number of countries--in particular those with second-generation PRSs-- have developed results-oriented strategies. Budget rules and transparency mechanisms have raised the incentives to execute the budget effectively. However, in some cases the results orientation of PRSs has led to the prolif- eration of products without significantly raising performance incentives. In budget processes, poor budget formulation, weak enforcement of rules, and, in some cases, continued informality, have limited the extent to which budget regulations and transparency mechanisms could create incentives to execute the budget effectively. Incentives in the PRS Process The PRS is explicitly intended to provide a results-oriented approach to strategic planning and policy making, by formulating a clear framework of policy goals and targets for poverty reduction and monitoring their imple- mentation. Setting up such a system was expected to help open a policy space for dialogue, establish priorities, design programs and policies, create a demand for data that would help set realistic targets, and assess imple- mentation with a view to refining the strategy (Norton 2002: 31). The country case studies offer a number of positive examples of enhanced performance incentives. The definition of clear and measurable poverty reduction goals within government strategies is undoubtedly useful, and the case study countries with second-generation PRSs (such as Burkina Faso, Rwanda, Tanzania, and Uganda) have placed significant emphasis on devel- oping results-oriented strategies, influenced by the strong international focus on development results and achieving the Millennium Development Domestic Accountability in the PRS and Budget Processes: Expectations and Reality 25 Goals (MDGs). This has generated more attention to the use of performance indicators and targets, and to the improvement of information systems. In Burkina Faso, for instance, the PRS is accompanied by a Priority Action Program, which helps to make the PRS operational and includes priority measures and performance indicators. In contrast to these successful efforts to cultivate incentives for improved results, a lack of prioritization and operational detail has remained a persistent feature in some PRSs and annual progress reports. This has been attributed to an inherent tension in the PRS between participation and comprehensiveness on the one hand, and the need for prioritization on the other hand, coupled with diffi- culties in making tough policy choices across competing demands, a lack of capacity to analyze trade-offs, and/or donors continuing to encourage the inclusion of particular projects and priorities. Partly as a result of the lack of operational detail, donors have supported the formulation of derivative products of the PRS, such as prioritized performance assessment frameworks, which seek to make the PRS opera- tional and around which donors can align. They have thereby introduced the risk of diffusing performance incentives, because these monitoring efforts usually are separate from domestic processes and only cover donor- supported interventions. In Burkina Faso, recent progress has been made with the establishment of a harmonized framework for assessing budget support by the group of bud- get support donors. These arrangements involve a common performance assessment framework, which relies mainly on performance indicators drawn from the PRSP and the government action plan for budget reforms. Incentives in the Budget Process Institutional arrangements around the budget are meant to discipline and facilitate decision making and the scrutiny of those decisions by the appropri- ate players (president, cabinet, line ministers, legislators, front-line service providers, and so on) (World Bank 1998: 18). Mechanisms that promote trans- parency (for instance, the public availability of information, or open budget preparation/execution/reporting) ought to raise the incentives for good performance by enabling parliament and the public to question spending decisions and criticize results. There are indications that public financial management in low-income countries has improved in recent years. Over the period 2002­04, heavily indebted poor countries (HIPC) indicators and country institutional and policy assessment data point to a 3 percent average annual improvement in public financial management. This general picture belies large cross-country 26 Domestic Accountability in the PRS and Budget Processes: Expectations and Reality variations: while many countries strengthened their budget systems in just three years--some by significantly more than any plausible margin of error--performance in others declined (World Bank 2006a: 145). A recent Global Monitoring Report, which assesses budget performance based on four criteria developed under the Public Expenditure and Financial Accountability (PEFA) program, likewise concludes that gradual overall progress has been made in recent years.6,7 PEFA ratings were available for six of our nine case study countries. They show that external scrutiny and audit were weakest,8 followed by pre- dictability in budget execution, accounting and reporting, transparency and comprehensiveness, policy-based budgeting, and budget credibility.9 None of the countries received the highest rating (A) for having the multi-year perspective in fiscal policy, planning, and budgeting that is an important aspect of linking the PRS and the budget. Budgets in the case study countries are rarely executed as planned in the annual budget or medium-term expenditure framework (MTEF) documents, breaking the link between approved budgets and actual spending. When budget execution is weak, the emphasis placed on the importance of the PRS-budget link can be called into question, because the budget cannot be considered a reliable instru- ment of policy implementation. Weak budget execution can be the result of poor budget formulation, but a number of other factors can also play a role, such as volatility of revenues and weak enforcement of rules. Where informality is the norm, incentives to exe- cute the budget as planned are poor, because internal controls are inadequate and domestic accountability weak. In some of the country cases (Albania, Tanzania, Uganda) aggregate spending is managed with a high degree of predictability, though even in these countries the budget execution by individ- ual spending agencies is difficult to predict. In Malawi, at the other extreme, the budget is a highly unreliable guide to expenditure in nearly all respects (Rakner and others 2004). Atypically low level of ownership of the budget out- side of government can also undermine incentives for good budget execution. Summing Up: Accountability Gaps in the PRS and the Budget In theory, both the PRS and the budget offer opportunities for enhanced domestic accountability through ownership and incentives for perfor- mance. In practice, asymmetrical ownership relationships and the weakness in some countries of domestically driven incentive structures can hamper the development of accountability mechanisms. Accountability gaps result in fractured relationships between budgeting and planning mechanisms, as conceptualized in Figure 3.3. Domestic Accountability in the PRS and Budget Processes: Expectations and Reality 27 Figure 3.3 Multiple Fractures in Planning and Budgeting PRSP policies and plans budget budget MTEF formulation execution sector strategies PRS and budget reporting Source: Authors. While both the PRS and budget offer scope for enhanced domestic accountability, fractures in planning and budgeting systems pose obstacles for national governments and donors. To overcome them, different actors need to be brought together to strengthen the budget as a process to imple- ment political decisions. The following challenges are central to strengthening the links between PRS and budgets: the prioritization of plans and coordi- nation between planning and budgeting units; the creation of incentives to formulate realistic budgets and execute them as planned; the expansion of ownership of the PRSP at the sector level; and the development of a multi- annual perspective in strategic resource allocation. A critical link in the relationship between PRSs and budgets, which deserves particular attention and is discussed in Chapter 5 below, is reporting. The accuracy, frequency, and timeliness of reporting on the PRS and budget can play an important role in expenditure accountability and performance monitoring. In most country cases, reporting processes could be enhanced to improve the delivery of important information on budgets and plans. Before this in-depth discussion of the role of reporting mechanisms in bridging domestic accountability gaps, Chapter 4 provides insights into recent efforts to strengthen PRS-budget links in our sample of countries. 4 Linking Poverty Reduction Strategies and Budgets From the earliest Poverty Reduction Strategy (PRS) experiences, links between the PRS and the budget were recognized as a likely area of weak- ness (Alonso, Judge, and Klugman 2006). As Chapter 3 has shown, gaps in accountability in both the PRS and budget have limited the effectiveness of both instruments for planning and resource allocation. Several countries have made efforts to put suitable mechanisms in place to link PRSs and budgets. This chapter provides an overview of these efforts, indicating that internally driven, integrated approaches to reform, which focus on getting the basics right, offer the highest probability of success. Mechanisms to Link PRSs and Budgets Fragmented ownership of PRS and budget processes and inadequate incentives to execute the budget as planned represent serious obstacles to the integration of planning and budgeting instruments. Since the intro- duction of the PRS approach, donors and PRS countries have introduced a number of technical instruments aimed at improving the integration of PRSs and budget processes, and at ensuring that PRS priorities are reflected in budget allocations. The initiatives that have taken place in the case study countries include the enforcement of pro-poor spending priorities, the introduction of medium-term expenditure frameworks (MTEFs), and program budgeting. The reforms are at different stages in different countries, reflecting the level of development of public financial management systems. 29 30 Linking Poverty Reduction Strategies and Budgets Priority Sectors and Pro-Poor Spending One of the most common ways in which budgets are being linked explicitly to the PRS is through the identification of priority sectors or programs. These sectors and programs are given priority in resource allocation during budget formulation, and their activities are tracked more closely during implementa- tion. They may also be given special protection against within-year cuts in budget disbursements and identified as a virtual poverty fund. Many of these mechanisms were first introduced to track the spending of additional resources made available through the heavily indebted poor countries debt relief initiative. However, they have since been institutionalized as conven- ient means of reassurance to donors. A summary of approaches to determine priority spending in six of the case study countries is provided in Annex 2. The use of priority sectors has helped to focus efforts on specific areas of expenditure, and to varying degrees it has facilitated shifts in resource allocations to those priorities. However, a preferential focus on priority sectors, or pro-poor expenditures, runs against the concept of the PRS as a comprehensive strategy paying attention to intersectoral complementarities and trade-offs. Tanzania's Strategic Budget Allocation System (SBAS) is in part an attempt to address this problem and to create a comprehensive link between the budget and the PRS (Box 4.1). Medium-Term Expenditure Frameworks Most of the case study countries are implementing reforms related to establish- ing MTEFs. These reforms vary significantly in their nature, scope, and cover- age, and consequently in the extent to which they can be useful in developing a strong link between the budget and the PRSP. False starts in the introduction of MTEFs have created situations where, in countries such as Malawi and Madagascar, development partners have been actively lobbying against their introduction, considering a gradual approach to be more appropriate. The basic aim of an MTEF is simple: to give a strategic dimension to the annual budget process by introducing forward revenue and expenditure estimates, and thus to encourage shifts in resource allocations based on expenditure priorities and an assessment of the medium-term sustainability of expenditures. However, while the aim is simple, the required changes in practices and systems are not. Table 4.1 shows the major components of the MTEFs of Uganda and Mozambique, as assessed through the Public Expenditure and Financial Accountability (PEFA) framework. This gives a general idea of the difficulties in implementing such wide-ranging reforms in environments that are often characterized by limited technical capacity and weak or uneven political commitment to reform. Linking Poverty Reduction Strategies and Budgets 31 Box 4.1 The Strategic Budget Allocation System in Tanzania Tanzania has gone further than any other country in developing a hard and structured link between the PRS and the budget through the Strategic Budget Allocation System (SBAS). The Tanzanian Poverty Reduction Strategy Paper (PRSP) has 3 "clusters" and 181 "cluster strategies," making its structure very different from that of the performance- based budget, which was structured by ministries, departments, and agencies; objec- tives; programs; and output. Although there are some objectives in between the clusters and strategies, it proved difficult to explicitly align the performance-oriented budget to the PRS clusters. The SBAS was developed to solve this problem. It is used to generate ceilings for spending by ministries, departments, and agencies (MDAs) for incorporation in Tanzania's budget guidelines: · SBAS Micro--Ministries prepare budget requests on their own versions of SBAS, "SBAS Micro," by identifying their resource requirements for MTEF objectives, targets, and activities, and linking these requirements to specific cluster strategies of the Mkukuta, Tanzania's Poverty Reduction Strategy, where appro- priate, or to other objectives from the ministry's strategic plan. These requests are submitted to the central Budget Guidelines Committee. · SBAS Macro--The Ministry of Finance uploads the Macro Framework and min- istries' requests into SBAS Macro, and the Budget Guidelines Committee agrees to the necessary cuts to all agency submissions to enable them to fit within the Macro Framework. In so doing, the Ministry of Finance scrutinizes the link between the MDA requests and the Mkukuta. The resultant ceilings are then incorporated into the budget guidelines that are submitted for approval by the Interministerial Technical Committee and the Cabinet. Once approved, these ceil- ings and allocations to targets are returned to the ministries, allowing them to prepare their detailed MTEF submissions in the normal way and upload them into the government's Integrated Financial Management System. A similar system, PlanRep, is under development for local governments, and local authority submissions are due to be incorporated with the national SBAS system in the 2007-08 budget. As illustrated, two recurring features of MTEFs can undermine the PRS- budget link. First, the MTEF is often considered as a separate process from the budget, because ownership of the MTEF and the incentives to use it as a rolling budget framework are weak (Le Houerou and Taliercio 2002). Often MTEFs themselves are not clearly linked to the budget, in particular for the outer years of the normal three-year timeframe, but sometimes even for the budget year itself.1 Second, the costed sector strategies that are needed for a meaningful debate around strategic intersectoral priorities and trade-offs are often absent, or inconsistent with the PRSP. A successful approach to 32 Linking Poverty Reduction Strategies and Budgets Table 4.1 Breaking the MTEF into its Constituent Parts Dimension of MTEF Uganda Mozambique Preparation of multi-year Multi-year aggregate fiscal A medium-term fiscal fiscal forecasts and forecasts and forward framework is prepared functional/sectoral expenditure estimates on an annual basis, with allocations. (based on an economic projections for three years and sectoral basis) are and an indication of prepared on a rolling sectoral allocations. annual basis, and However, the links with explanations are provided. budget allocations for the While year one MTEF following years allocations end up as the are still unclear. budget, forward year projections have proved unreliable in the past. Existence of sector Costed strategies exist The majority of main sectors strategies with for several sectors have sector strategies, but multi-year costing including health, in many cases these are not of recurrent and education, water, fully costed, and are not investment expenditure. and defense. linked to fiscal projections. Links between At a planning level, In very few cases is there investment budgets there is recognition a clear link between and forward of the relationship investment projects and expenditure estimates. between capital implications for future investment and recurrent recurrent expenditures. expenditure. However, In several sectors capacity constraints are investment decisions not well integrated, for do not seem to be example, by factoring in made on the basis the capacity of teacher of strategic plans. training centers or nursing schools to meet the demand. Source: 2005 Public Financial Management Assessments using PEFA performance measurement framework. these problems has been developed in Uganda, where sector strategies are closely linked to resource allocations; the MTEF process in Uganda is also helped by the fact that the first year of the MTEF is clearly the forthcoming budget (Box 4.2). However, the weaknesses associated with the preparation of reliable MTEFs must also be seen in a larger context of poor aid predictability. In Tanzania, for example, donor MTEF projections of external financing Linking Poverty Reduction Strategies and Budgets 33 Box 4.2 Role of Planning and Budgeting Instruments in Uganda In Uganda, the roles of planning and budgeting instruments are relatively clear, having evolved prior to the PRS initiative, facilitated by a unified finance and planning ministry. · The Poverty Eradication Action Plan (PEAP), which became Uganda's PRS, is the overall guide for the development of sector strategies and strategic (intersector) resource allocation in the overall MTEF and budget. · Meanwhile, sector strategies are the basis for guiding more detailed sector-level MTEF and budget allocations. The involvement of sector working groups and their strategies in the budget prepara- tion process has helped to deepen the link between the PRSP and the budget. As sector working groups have to identify high level, key outcome and output indicators, they tend to use PRS targets as a reference. over a two- to three-year horizon have diverged from actual financing by 4­5 percent of GDP, on average, over the last few years. The resulting poor credibility of MTEF projections undermines the potential usefulness of the framework for budget planning and domestic accountability. Program Budgeting Program budgeting is based on the premise that all budget items can be linked to discrete outputs that contribute to the achievement of specific policy objectives, such as those included in PRSs. The introduction of pro- gram budgeting has the aim of increasing the results orientation of budgets, through a restructuring of the budget system that allows for the creation of explicit links from, for example, expenditure on basic health care, to delivery of health services, to outcomes such as reductions in morbidity or mortality indicators. Most PRS countries' budget systems have traditionally had an adminis- trative focus, and have been input oriented, and efforts to enhance the focus on performance have tended to promote reforms to the budget classification system. In some cases two aspects of program and performance budgeting became conflated: · The need for strategic information to policy makers within the execu- tive and beyond on the implementation of policies to facilitate strategic resource allocation decisions; and · The need for detailed information on program implementation by agency and program managers to improve program implementation performance. 34 Linking Poverty Reduction Strategies and Budgets Thus, when combined with the legacy of detailed control of inputs, the introduction of program budgeting has often caused a proliferation of indica- tors and programs. For instance, in Tanzania, the (relatively small) Ministry of Community and Children has more than 90 output targets specified, and meanwhile, given the absence of standard program codes, it is very difficult to get a picture of total allocations to sectors. In Madagascar, the budget is organized around 46 missions and 146 programs, which is potentially sound; the problem comes at the level of objectives, with 4,526 indicators. A contrasting approach has been taken in Mali and Uganda, where results orientation has been introduced without formal performance or program budgeting. Uganda has managed to develop a deep link between planning and budgeting, without changing the budget classification system. The administrative budget has been grouped into sectors, which represent broad PRSP priorities. In addition, grants to local governments are earmarked to sectors and sub-sectors. During the budget process, each sector identifies high-level, key outcome and output indicators and targets, often based on sector strategic plans and the PRSP, and aims to justify its MTEF/budget ceil- ings in these terms. Detailed budgets are then prepared in the traditional way. In Mali, the budget for programs is an annex to the traditional budget that contains broad information on the planned expenditures by ministry and pro- gram, and descriptions by ministries of short- and long-term objectives.2 Although these reforms may not have gone as deep as the introduction of full program or performance budgeting, they have helped to shift decision mak- ing on budget allocations toward policy objectives. Toward a More Effective Link As the previous section indicated, despite the challenges of implementing highly technical reforms, some of the case study countries have linked PRSs and budgets through the use of priority spending, MTEFs, and results ori- entation. The SBAS in Tanzania, a clear link between the MTEF and budget in Uganda, and an increased emphasis on results in Uganda and Mali, rep- resent some of the good-practice approaches to these technical issues, which have been supported by good sectoral cooperation and buy-in. If, in contrast, reforms are mainly driven by donor investments and ownership is limited to a few technocrats (often in the MoF) without taking account of the interests of different sets of stakeholders, such reforms are not likely to generate strong incentives for the behavioral changes that are needed for successful implementation.3 This is often because technical capacity problems are often compounded by lack of widespread ownership Linking Poverty Reduction Strategies and Budgets 35 of the proposed reforms. The implementation of reforms is not merely linked to developing technical capacity in a narrow sense (through training and skill building) but requires a broader and inherently political process with individual, organizational, and institutional dimensions. In Malawi, for example, the government has tried since the late 1990s to remedy weak expenditure management by introducing an Integrated Financial Manage- ment Information System. This initiative has had strong donor support but has been plagued by technical difficulties and frequent resistance from officials in line ministries, who saw the electronic system as a threat to their discre- tionary space. In 2005, the original effort was replaced with an alternative based on the system used by the government of Tanzania. This new system was introduced in five of Malawi's core ministries in November of that year,4 without donor assistance but with strong government commitment in particular from the accountant-general's department. Experience in the higher-income reforming countries featured in this review also reveals the importance of country ownership and an internal drive for reforms. In addition, strong accountability relationships within the executive and between the executive and parliament are a common feature in these countries. As can be expected in countries with limited donor interven- tion, budget reform processes in these reformers were largely internally driven, and resulted from government efforts to deliver on electoral promises (Australia, Chile, and South Africa), while at the same time responding to the external environment (the Republic of Korea and Australia). Inevitably, finance ministries were the main initiators and guardians of the reforms in all countries, although they played these roles mostly through defining the rules and facilitating the engagement of other actors. Cabinets were also heavily involved in all cases, but in Chile and South Africa parliaments and civil society also played a part. Accountability worked mostly in its internal, horizontal dimension within cabinets; the importance of keeping all key actors on board meant that all cabinet members had to perceive that they were sharing the pain or the gains of the reform. In Australia and Korea, spending agencies were given more autonomy and flexibility in return for their compliance. Perhaps one of the most striking features of the successful higher-income reformers is the close institutional connection between strategic planning and budgeting. There were no cases of stated conflict between the entities responsible for managing finances and those in charge of planning, and in most cases a single agency was in charge of both. Reform strategies put their main emphasis on improving budgets as tools for implementing govern- ment policies, by ensuring aggregate fiscal discipline, prioritizing resource 36 Linking Poverty Reduction Strategies and Budgets allocation, and achieving efficiency and effectiveness in spending. In partic- ular, budget reforms were often born out of a need to ensure fiscal disci- pline, whether in response to a regional crisis (Korea) or to create a solid basis for expanding service delivery (South Africa). Reconciling fiscal disci- pline with policy objectives in turn created the need to develop a medium-term budgeting perspective, to allow for better prioritization and clearer con- testability. As a result of these close institutional connections, reform agen- das in the successful higher-income reformers were integrated, in the sense that they covered not just budget formulation but also accounting, monitor- ing, and reporting. In some cases (notably Australia and Chile) they also involved changes in core public administration structures, such as dealing with the flexibility of managers, pay structures, or performance contracts. Reform, in this sense, was much more than identifying specific measures to be implemented; it involved identifying interlinked sets of issues that were key in determining incentives and outcomes, knowing that different reforms would reinforce each other. Other interesting features of the successful reformers include a gradual approach that started with fixing basic deficiencies in public financial man- agement and allowing sufficient time before introducing more sophisticated measures. In all four countries studied, reforms started with basic measures, tackling capacity constraints and existing incentive systems, then gradually introduced more sophisticated systems. Demand for macro-fiscal disci- pline was the starting point in Australia and South Africa. The importance of credible budget information and effective expenditure controls therefore received the most attention; mechanisms for introducing a medium-term dimension came next (Australia and Korea), with a view to enhancing the policy orientation of the budget and turning it into a more useful tool for policy planning and implementation. Performance orientation came last, and in many countries (except Chile) is still lagging behind, sometimes even after 20 years of budget reform. Even in the more successful case of Chile, performance orientation developed gradually, starting from a very simple base.5 Three principal lessons, then, can be derived from these higher-income country experiences. First, domestic ownership of the reform agenda is often crucial and may need to extend beyond the ministries of finance to the cabinet, and, in some cases, to parliament and civil society. Second, an integrated approach to reform rests on strong institutional connections between planning and budgeting. And third, the most effective reforms begin with getting the basics of public financial management right, before moving on to more sophisticated, performance-oriented reforms. Linking Poverty Reduction Strategies and Budgets 37 Each of these lessons is also reflected in the most successful reform experiences in the lower-income case study countries. Among these coun- tries, Madagascar has taken an innovative and--by early accounts--highly successful approach to building capacity, ownership, and incentives at the core of government. A leadership action and management program (LAMP) has resulted in broad ownership of the national development strategy and much closer integration with sectoral strategies and the budget (Box 4.3). LAMP has contributed to the development of a government-owned devel- opment strategy (the Madagascar Action Plan), accompanied by a specific results agreement between the president and his cabinet. Some of the low-income countries among our case studies have begun to better integrate planning and budgeting processes, enabling an institution- ally harmonized and integrated approach to planning and budgeting reforms. For example, the Tanzanian government has recently launched an initiative to harmonize planning and budgeting instruments, following the second iteration of the PRS, which commands deeper political and technical ownership than did its predecessor. In Mozambique, a consensus was reached on integrating the next PRSP within the government's next five-year program, from 2010. Albania is promoting an Integrated Planning System (IPS) (Box 4.4). Summing Up Integrating PRSs and budgets substantively--as arenas of priority setting--is fraught with difficulty. Countries have made significant efforts to link PRSs and budgets through identifying pro-poor spending and introducing MTEFs and results-oriented budgeting. These cases indicate that the introduction of medium-term budgeting and performance orientation are technically very ambitious, and often lack support outside the MoF. Consequently, the impact of these technical innovations has been limited by a failure to address deeper institutional problems. It requires institutional change and high-level politi- cal support as well as support from management and from staff within the administration. Lessons from successful budget reformers seem to suggest that an internal drive for reform--as reflected in strong central leadership, a strong role for the cabinet in the strategic phase of the budget process, and a potential role for parliament and civil society--helps to increase the incentives for perfor- mance and the impact of technical reforms. A lack of institutional conflict between strategic planning and budgeting is also important for aligning incentives for reform. Finally, a gradual and integrated approach to 38 Linking Poverty Reduction Strategies and Budgets Box 4.3 The Leadership and Management Program in Madagascar Coming to power after a protracted political crisis (December 2001­July 2002), President Ravalomanana's government inherited an inherently problematic PRSP process, weak- ened by low ownership in the government. The annual budget, lacking a medium-term perspective, did not reflect PRSP or sectoral strategies. As a consequence, activities focused on short-term goals under a variety of shifting objectives, overstretching the state apparatus and causing confusion. A hierarchical organizational culture also meant that the government bureaucracy was ill equipped to implement medium-term development strategies. On the president's request, the World Bank helped to initiate a Leadership and Management Program (LAMP) in 2003. LAMP aimed to enable the creation of a coher- ent development strategy across government, while equipping officials with the tech- nical expertise to overcome bureaucratic obstacles to change. So far it has included a series of cabinet-level retreats (2003­06), a one-week management training course in Canada by the Institute of Public Administration of Canada (2004), an ongoing leadership training program (2005­07), and capacity building and institutional sup- port to priority sectors. While the first three of these programs have been aimed at ministers and senior government officials, the fourth is aimed at a broader segment of leaders in government. LAMP's tangible contributions to outcomes include · The development and dissemination across government of a five-year develop- ment strategy--the Madagascar Action Plan (MAP)--built on far-reaching con- sultations with all sectors and levels of society. · The creation of a new results framework, La Politique Générale de l'Etat, which sets specific targets for the administration. Under the Politique each ministry defines three to five results that form part of the results agreement between the president and his cabinet members. Implementation of the Politique is tracked on a regular basis and progress and bottlenecks are discussed at the cabinet level. · The introduction of a program to improve the transparency of budget allocations and ongoing support to improve consistency between the MAP, sectoral strate- gies, and the budget. · Support to priority ministries (health, education, transport, environment) to assist them in preparation and implementation of program budgeting. In par- ticular, the Ministry of National Education and Scientific Research has had notable success in identifying priority programs and in ensuring that these are protected when budgetary restrictions are made or cuts are imposed. Source: Heidenhof, Teggemann, and Sjetnan 2007. Linking Poverty Reduction Strategies and Budgets 39 Box 4.4 The Integrated Planning System in Albania Albania's Integrated Planning System (IPS) was the result of a 10-month collaborative process between the government and donors. Numerous planning frameworks, includ- ing the National Strategy for Social and Economic Development (NSSED), the European Union (EU) Stabilization and Accession Agreement, the government's medium-term budget program, and NATO accession, were implemented as separate exercises, each with its own set of priorities, procedures, and reporting requirements, despite efforts to integrate them. Not surprisingly, this produced multiple priorities, overlapping processes, an excessive reporting burden in ministries, and public confusion. The overriding goal of the IPS is not to overlay yet another new planning system but to serve as a unifying framework for existing systems. The annual progress report of 2004 for the NSSED outlines this new planning architecture as the basis upon which the new NSSED will be built. Key elements include · Creation of a political mechanism--the Strategic Planning Committee--chaired by the prime minister, to provide strategic policy and fiscal planning directions; · Establishment of a Government Modernization Program that may eventually include related initiatives, such as functional reviews, e-government, and service quality improvement; · Transformation of the NSSED into a government-wide, strategic planning process, incorporating EU and NATO integration, the Millennium Development Goals (MDGs), the Government Program, decentralization, and other cross-cutting strategies; · Restoration of the medium-term budget plan as the exclusive process through which all cross-cutting and sector strategies compete for funding and are trans- lated into concrete plans, budgets, and expected results; · Establishment of more strategic and comprehensive public investment and external assistance planning and financial management processes; · Restructuring of central institutions with planning mandates: moving NSSED to the Council of Ministers, transferring public investment and external assistance functions from the Ministry of Economy to the MoF, and creating an integrated delivery function (planning, budgeting, monitoring) within line ministries; and · Institutionalization of the government-donor coordinating mechanisms to ensure ongoing collaboration and effective deployment of external assistance. Implementation of the IPS is a daunting task and will confront a range of chal- lenges despite the firm commitment of the new government to its implementation and the strong support of all development partners. If, however, IPS is implemented at a measured pace over the next three to five years, a more coherent and stable policy environment will emerge. For decision makers, the IPS will identify a series of clear decision points. For donors, the IPS will create a single point of initial contact and a demand-driven system where external assistance proposals flow from stated govern- ment priorities. For the public, the IPS will produce a single, annual published plan and corresponding set of results for each ministry. Source: World Bank Country Assistance Strategy for Albania 2006­09. 40 Linking Poverty Reduction Strategies and Budgets reform seems to have worked well. These cases allowed sufficient time to strengthen basic systems, starting with credible and reliable budget execution, before moving to more sophisticated measures. Chapter 5 homes in on the important role of budget and PRS reporting for domestic accountability and the difficulties and possible gains associ- ated with different approaches to their integration. 5 Integrating PRS and Budget Reporting This chapter considers the challenges posed by the proposal to build stronger links between Poverty Reduction Strategy (PRS) and budget report- ing in the interests of more effective domestic accountability. The discussion builds on the description in Chapter 4 of the challenges that countries often face in strengthening PRS-budget links. It argues that opportunities for inte- gration are shaped by the individual and institutional incentives to report and that use of information in decision making and accountability processes plays a critical role in generating these incentives. The Nature of Budget and PRS Reporting A well-functioning, results-oriented reporting system needs to integrate financial information on actual expenditure (inputs) with performance information on outputs and outcomes. If feasible, basing PRS annual reporting on the country's budget reporting would have the considerable advantage of building on existing country systems. Ideally, the budget is consistent with the expression of a country's policy priorities for a given year. As the central process of the public finance system, the budget process is a key instrument for translating national priorities into policy actions, and ulti- mately results. While the PRS goes beyond the budget, because it includes strategies and priority actions that do not require the allocation of public funds, the systems for monitoring government performance in implement- ing budgets could provide a solid basis for reporting on a critical part of PRS implementation. Despite the overall potential gains (in efficiency and learning) to be made from linking PRS and budget reporting systems, the respective structures of these mechanisms do not lend themselves to easy integration. 41 42 Integrating PRS and Budget Reporting As the brief discussion on budget and PRS reporting below indicates, budget reports tend not to be linked to discussions of results, and PRS reports tend to offer little detail on the links between government spending, actions, and poverty outcomes. Institutional separation between budget and PRS reporting mechanisms prevails in most of the low-income case study countries. Budget Reporting In many countries, budget reporting is limited to financial information on expenditures against budget allocations, and--usually as a result of com- pliance with laws and regulations--is organized by implementing institu- tions rather than for management purposes. This is often supplemented by reporting on aggregate revenues and expenditures, which is usually needed for macroeconomic management purposes, including satisfying IMF requirements. All countries have statutory requirements to present consolidated reports on budget implementation to parliament in one form or another, but such reporting is often not accompanied by performance information, so it is of limited relevance for accountability purposes. Uganda is an exception where overall budget performance reporting has been linked to results (Box 5.1). Box 5.1 Budget Reporting in Uganda Between 1998 and 2001, sectors prepared quarterly reports that were compiled into the Poverty Action Fund quarterly reports by the Ministry of Finance, Planning, and Economic Development (MFPED), and combined with a financial statement illustrating disbursements against budget. These reports were discussed with the concerned min- istries and donors. This represented the first attempt at compiling within-year sector reports on financial and physical performance at a national level. But the quality of the reports was relatively poor, with little rigorous analysis and much repetition. Half-yearly budget performance reports (BPRs) were introduced in the early 2000s, representing a move from the performance assessment framework (PAF) to more com- prehensive budget reporting. The BPRs are produced by MFPED. They initially only included information on budget releases, but now include performance information drawn from sector reports. The budget performance reports are superior to their prede- cessors, the PAF reports, but there is no forum for debating them beyond discussions with donors, which limits their accountability role. Source: Tim Williamson, Uganda case study. Integrating PRS and Budget Reporting 43 PRS Reporting Reporting on PRS implementation has evolved in different ways in different countries, reflecting the fact that there has been no blueprint for such reports.1 Governments in PRS countries are supposed to produce an annual progress report (APR) for every year of PRS implementation, to meet one or more of the following objectives (Booth, Christiansen, and de Renzio 2004): · enhance government performance in poverty reduction; · meet donor reporting requirements (for budget support, highly indebted poor countries (HIPC) support, other international financial institution and bilateral support); and · support enhanced government accountability to citizens. APRs were originally intended to be built upon the country's existing arrangements for monitoring poverty outcomes. Among the case study coun- tries, Mozambique and Uganda (and to a certain extent Burkina Faso) use existing reporting processes as a basis for the APR. Mozambique has used the annual report on the implementation of the Social and Economic Plan--a document that is presented to parliament alongside budget reporting, detail- ing the implementation of government priority programs--to fulfill the APR requirements. The Ugandan Ministry of Finance used the Background to the Budget document, which traditionally provided an overview of economic prospects and major budgetary issues. While the Mozambican reporting process was a natural vehicle for annual PRS reporting, building on domestic accountability processes, Uganda's Background to the Budget document was not particularly appropriate for this purpose, and is now being replaced by an annual implementation report on the Poverty Eradication Action Plan (Uganda's PRS), at the government's initiative. In most of the case study countries, including Malawi, Rwanda, and Tanzania, the APR was introduced as a new and separate instrument, not linked to existing processes, often because no appropriate instruments were in place at the national level. Even in cases where the preparation of the APR is highly participatory (for example, in Albania) the primary motiva- tion is still to fulfill donor requirements. However, it is important to note that in Rwanda ownership of the APR process has increased over time. PRS monitoring was originally conceived as including both monitoring of basic service delivery activity and outputs (through collection of routine administrative reports) and analysis of poverty outcomes (using information from household surveys and similar instruments).2 In practice, however, APRs draw on either one, but rarely both, of these sources of information. Those that report on the implementation of government actions and services 44 Integrating PRS and Budget Reporting often draw on sector reports (as in Albania, Mali, Mozambique, and Rwanda), with little discussion of poverty outcomes. Others (as in Tanzania and Uganda) focus largely on poverty outcomes and high-level policies in relation to those outcomes, and draw on survey data, with little focus on government actions. The latter kind have the problem that outcome data do not change signifi- cantly from year to year, because surveys are not usually carried out on an annual basis. In almost all cases, the APR provides few details on the composition of public spending. As a result, the links among government spending, actions, and outcomes are rarely explicit in APRs. In Mozambique, for example, the APR includes information on budget execution only at a very aggregate level, to show government performance on the 65 percent target of total spending that is dedicated to Poverty Reduction Strategy Paper (PRSP) priority sectors. In Rwanda, significant improvements were made in the 2005 APR, which included more detailed information on budget expenditures. The nature of APRs in case study countries is described in more detail in Annex 3. Links between PRS and Budget Reporting In most cases, a disconnect between budget and PRS reporting systems stems from the differences in their respective roles. The two systems are geared to different stages of the results chain, with PRS reporting geared toward out- comes and impact, and budget reporting geared toward inputs--and in a few countries, outputs. The separation of institutional responsibilities for reporting on plans and reporting on budgets is a clear demonstration of this, and exists in many countries. In Mozambique, for example, budget execution reports are produced separately from the reports on the annual Social and Economic Plan, which focus more on performance. The same happens in Albania, where budget reporting is only input based, while results are reflected in the APR. Budget reporting in Uganda, while containing some information on outputs, is confined to the lower levels in the results chain (activities), while progress reporting on the PRS concentrates on the higher levels (outcomes and impacts) and does not cover budget implementation (Figure 5.1). Unsynchronized timetables are often a function of poorly integrated budget and planning processes, which in turn act as further obstacles to integration. In Mali, APRs (and sector reports) are produced after the con- clusion of the budget process. In Mozambique, the timing of central and sectoral reporting instruments is different and does not coincide with the Integrating PRS and Budget Reporting 45 Figure 5.1 Use of Performance Information in Reporting Instruments in Uganda local government & MDA sector national impact PEAP results poverty status outcomes matrix report annual sector reports outputs/ assessment sector MIS MIS reports, activities reports returns league tables ministerial PAF grant FDS reports policy budget reports statements performance reports LG financial CCS inputs/funds statements returns Source: Tim Williamson, Uganda case study. Note: CCS: Commitment Control System. FDS: Monthly financial statements and quarterly results based on budget reports by local government. MIS: management information system. LG: local government. PEAP: Poverty Eradication Action Plan. PAF: performance assessment framework. MDAs: ministries, departments, and agencies. budget cycle. For 2007, the Ministry of Finance (MoF) is trying to produce a single document by the end of July, complying with the central budget cycle. In sum, PRS and budget reporting is rarely presented in a way and at a moment that facilitates its use for policy making. As a result of the lack of insti- tutional harmonization, information is not shared adequately and promptly between planning and budgeting units, between ministries of finance and planning, or with the cabinet and parliament, and the respective roles of sector ministries are not well aligned. Harmonizing reporting processes so that they are more directly linked to budget cycles is therefore an important way to ensure that relevant information feeds into decision making. However, as the next section demonstrates, the challenge of integration is not merely a technical one but depends also on the bureaucratic and political incentives for change. 46 Integrating PRS and Budget Reporting Three Building Blocks From the case study countries, three essential building blocks for integrating PRS and budget reporting emerge: · Regular administrative reporting from various government imple- menting institutions, such as central ministries/agencies and local governments. · More advanced sectoral monitoring and reporting processes, particu- larly in health and education. · Improvements in surveys and statistics on poverty outcomes. Although the three elements for successful integration may be considered a function of adequate capacity, they are, more fundamentally, related to bureau- cratic and political incentives to integrate financial and performance reporting. In many low-income countries, the weakness of such incentives means that one or more of the building blocks are absent, incomplete, or unreliable. The first building block--regular administrative reporting from various government implementing institutions--is strong where agencies and ministries perceive themselves to be participants in and owners of the PRS and budget processes. Here, the scope and quality of reporting often depend on whether delivery has been decentralized to local governments or remains concentrated within line ministries. The second element--more sophisticated reporting from key sectors--has improved largely due to the roll out of sector-wide approach (SWAp) type initiatives and to the need to report on the Millennium Development Goals, which in themselves brought about increased and better coordinated donor funding for planning, bud- geting, and monitoring systems. In the third element--improvements in surveys and statistics on out- comes--there is often a large discrepancy between the generation of infor- mation and its application. The PRSP initiative and the associated poverty monitoring systems have put great emphasis on the need to improve information on poverty outcomes, and surveys are a key source of such information. This means that surveys have been reinvigorated in recent times in many of the case study countries. But because the surveys are usually carried out by a central statistics agency, and only once every few years, data from the surveys are rarely integrated into reporting and deci- sion making beyond the formulation of the PRS. This is especially true for sector-level or budget reporting, which hardly ever takes into account the results of survey analysis. Despite the increased availability of survey data, the collection of routine data on service delivery also remains a key challenge (as in Malawi and Tanzania), undermining the quality of the performance information included in reports. Integrating PRS and Budget Reporting 47 The case study findings are confirmed by a recent study on PRS monitoring systems, which notes severe issues with data collection, in particular rou- tine administrative data, poor coordination, and little evidence that moni- toring outputs are being used either for policy making or advocacy (Bedi and others 2006: 14). External and Internal Incentives to Report Two forms of incentives are particularly relevant to the creation of ade- quate building blocks for integrated reporting: first, incentives to produce information and reports, using both financial and performance informa- tion; and second, incentives to link such information and analysis with policy-making processes, to trigger learning and improve performance. While incentives to supply information are often generated through external expectations, incentives to apply information to analysis depend more on domestic demands. Incentives to Supply Information In most case study countries, donors are cited as a major source of demand for performance information. Domestic stakeholders, including parlia- ment, the cabinet, civil society, and the electorate are relatively weak sources of demand (Figure 5.2). As a result, domestic accountability may be weak, especially as accountability to external stakeholders is often com- pelling. This imbalance is particularly evident in APRs, which have been largely a response to external demands for information on PRS implemen- tation, first linked to the HIPC initiative and subsequently to direct support to government budgets. In a number of case study countries, external demands for information have created incentives for improved PRS reporting mechanisms where domestic accountability is weak. In Tanzania, for example, donors' focus on poverty monitoring has helped improve the supply of data and reports on poverty outcomes. In Rwanda, well-coordinated donor-supported joint sector reviews have helped to improve the sector review process. In Mozambique, donors helped to improve the quality of reporting. Mozambique's first APR, produced in 2003, was of poor quality, but donors, rather than falling back on separate reporting systems, supported and pushed for improvements in the APR. In some cases, donor-driven processes have generated domestic demand for information. For instance, parliaments and civil society in Rwanda and Mozambique employ budget support reviews as a source of information on PRS implementation. Uganda's sector review processes at first relied strongly 48 Integrating PRS and Budget Reporting Figure 5.2 A Typical Imbalance in Incentives to Report to External Stakeholders executive cabinet parliament planning & finance ministries donors sector ministries & providers societal civil society electorate groups strong incentives moderate incentives weak incentives Source: Authors. on donors but are now domestically led, and donors participate alongside other stakeholders. Uganda's experience also shows that larger groups of stakeholders (line ministries, local governments, and civil society) can become effectively involved in the APR review process, expanding opportu- nities for evidence-based decision making. While donor demands for information can galvanize domestic interest in reporting processes, excessive external demands compound the risks of frag- mentation and duplication. For example, at the national level in Rwanda, the report on the annual action plan, joint sector reviews, the APR, and budget support reviews could benefit from more streamlining and further integra- tion. More generally, multiple donor requirements undermine the integration of reporting and the legitimacy of governments' own reporting systems. As mentioned earlier, SWAp-type arrangements can help to improve sec- tor reporting and improve coordination. However, even where SWAps are in use, there are different degrees of coordination among donors and of integra- tion between planning and budgeting instruments. In a number of countries the links between sector processes and the budget are weak, precluding the integration of reporting mechanisms. Uganda is one country whose sector Integrating PRS and Budget Reporting 49 processes are firmly rooted in the budget process. In Rwanda, since the intro- duction of the PRS, substantial efforts have been made to develop sector strategies, with varying degrees of success; in all but a few sectors, the link between these strategies and budget allocations needs to be strengthened. In most of the other case study countries, the links between sector-level processes and both the PRS and the budget have received limited attention. Using the example of Mozambique, Box 5.2 shows how SWAp mechanisms can set up planning and reporting systems that are parallel to mainstream budgetary systems, and thus distract attention from national processes. Although currently weak in most case study countries, domestic demand for information may represent a more promising source of reporting incentives. Rwanda's annual action plan (AAP) process is politically driven and domes- tically owned. Ministers report annually on activities conducted against AAPs at a retreat chaired by the president (Box 5.3). This is an important Box 5.2 Fragmentation and Duplication in the Health Sector in Mozambique In Mozambique, fragmentation and duplication of planning and monitoring instru- ments at the sector level have affected the importance of the PRS (PARPA) and its annual implementation plan. The health sector, for instance, has developed a strategic plan (Plano Estratégico do Sector Saude, PESS) since 2001, linked to a SWAp financing arrangement, with bas- ket funding made available by a number of donors on the basis of a Memorandum of Understanding. The main instrument for operationalizing the PESS is an Annual Opera- tional Plan (Plano Operacional Anual), which details the main activities to be under- taken during a specific year, and the available inputs in terms of different financing sources (the mainstream budget, donor SWAp funding, separate projects, etc.). At the central level the PARPA is operationalized through an Annual Social and Economic Plan and monitored on an annual basis. Because the ministry obtains most of its funding from donors through the PESS- SWAp arrangement rather than through the PARPA-funded national budget, the ministry has tended to see the PESS, rather than the PARPA, as its main policy document. For the same reasons, the Annual Operational Plan (which mainly covers donor-funded activities) is more important than the Annual Social and Economic Plan as the main annual docu- ment linking available resources with program activities and outputs. As a result, the Ministry of Health has tended to view sectoral planning and reporting instruments (i.e., the PESS and the Annual Operational Plan) and national planning instruments (PARPA and the Annual Social and Economic Plan) as competing, rather than as comple- mentary, documents that are part of the same national planning system and serve the same purpose. Source: Paolo de Renzio, Jose Sulemane, Mozambique case study. 50 Integrating PRS and Budget Reporting Box 5.3 The Annual Action Plan Process in Rwanda Rwanda provides an interesting example of a domestically driven accountability process. Around the end of each year, all ministries are required to produce AAPs, in a process that is coordinated by the prime minister's office. A year later, ministries must produce a report on the implementation of their AAPs. Since 2004, a one-week high- level government retreat has been organized (the "Akagera retreat"), at which minis- ters presented their action plans for the current year and are held accountable for implementation in the previous year. The president himself chairs the retreat discussions for the five days of the retreat. Following these discussions, ministries update their AAPs, on the basis of which the prime minister's office submits a consolidated report to the president. Although impor- tant for domestic accountability, these action plans could be better linked to the budget process to enhance their effectiveness. Source: Vincent de Boer, Robert Cook, Rwanda case study. example of intragovernment accountability and ownership, as reporting against Rwanda's AAP is explicitly linked to the implementation of the gov- ernment's seven-year political program, while some government officials indicate that the action plans may be seen more broadly as plans for the implementation of national strategies, including the PRSP. Incentives to Use Information to Improve Performance A key observation from the case studies is that information generated by reporting processes is often not fed into decision-making processes. Despite the fact that much of the information produced is potentially powerful for informing decisions, when reporting is done primarily to comply with donor-related or statutory requirements it will not automatically feed into policy making and public spending decisions. The case studies show that the greatest positive incentives to integrate reporting processes are domestically driven. Several countries have efforts under way to integrate reporting processes, and explicitly link them to government decision making. In both Mali and Uganda, the aim is now to produce the APR in such a way that it can feed into the budget process. In Mozambique, sector reporting related to SWAps is gradually being syn- chronized with central reporting related to the APR and budget cycle. In Tanzania, poverty monitoring systems have explicitly recognized the need to engage with domestic policy-making processes, and now include briefs to parliament and the cabinet. Integrating PRS and Budget Reporting 51 Box 5.4 The Role of Political Commitment and Leadership for Informed Decision Making in Uganda Uganda's first PEAP in 1997 embodied key political priorities of President Museveni's 1995 election manifesto, such as universal primary education. There was therefore strong ownership from the executive. Strong management in key institutions such as the MoF, combined with political commitment, appeared to create the space for evidence-based decision making. Technical reforms, often supported by donors, helped the government take advantage of this space, allowing decisions to be made. However, since then Booth and Nsabagasani (2005, p.7) has observed a narrowing of political space: Against this background, the second revision of the Poverty Eradication Action Plan dur- ing 2003/04 has been an inclusive and reportedly high-quality process, but one that is increasingly nonpolitical. President Museveni himself has become progressively less involved over a number of years, allowing the strategy to become more sophisticated in terms of overall objectives and articulation with sector plans, but more technocratic in terms of ownership. Meanwhile, the attention of opposition politicians. . . . has been increasingly focused on wresting power from Museveni and influencing the third-term debate, not on the details of policy. This has, among other things, reduced the space for Uganda's much-lauded monitoring innovations to operate and have influence. The pro- tection for critical thinking that has been afforded by the MFPED umbrella is less strong than it was. The potential for new evidence . . . has been reduced. As the political space has narrowed, the inadequacy of civil service reforms has also come to light. Although the MoF remains relatively strong, the government of Uganda has not been able to retain or replenish quality management and staff. Source: Tim Williamson, Uganda case study. Uganda's experience illustrates how political commitment to specific pro-poor policies, good management by the political leadership and the civil service, and donor support can combine to create the right environ- ment for evidence-based decision making and thus a results orientation in the budget process (Box 5.4). But it also illustrates how this environment is easily affected by a shift in political interest. In Malawi, stronger leadership in the MoF seems to be opening opportunities for more informed decision making. If political interest is a binding factor, this also brings to the fore the importance of supporting local, politically driven policy-making processes, such as the Rwanda AAP, even if these are not explicitly related to the PRS. Summing Up This chapter has reviewed the issues posed by the current state of reporting processes in PRS countries. Its findings resonate with a number of those 52 Integrating PRS and Budget Reporting of an earlier study on PRS monitoring systems (Bedi and others 2006: 20). In particular, it notes that all too often, reporting systems focus too much on the supply side of information, rather than on the demand side. While in most countries the basic building blocks of reporting systems--such as timely collection of administrative data--are improving, the extent to which these data will contribute to greater accountability depends on where, how, and by whom the information is demanded and used. In systems with weak incentives for the integration of reporting mechanisms, demands for infor- mation are often externally driven and information may not feed into domestic decision-making processes. In this context, a key question is whether PRS reporting is linked to relevant policy-making processes, or to policies where there is political demand for performance. Demand for performance will clearly be stronger in connection with processes that benefit from political "ownership" than with those that do not.3 Experience with successful reporting mechanisms in the higher-income countries and some of the low-income case study countries (Madagascar, Mozambique, and Uganda) emphasizes the importance of core policy-making processes.4 Most of these processes relied on cabinet-level structures such as committees and review bodies, which discussed fiscal frameworks and budget priorities, as well as getting involved in monitoring the execution of the budget and evaluating its performance. These structures ensured the direct engagement and involvement of politicians (mostly ministers, but in some cases also parliamentarians) at key decision points, thus ensuring the necessary buy-in and political peer pressure for "sticking to the path." This engagement allowed the use of information in policy and decision-making processes, with positive effects on the design of programs and sectoral policies. 6 The Agenda Ahead The case studies have revealed a number of challenges to integrating Poverty Reduction Strategy (PRS) processes with budgets. Even after several iterations, which have generally given PRSs a greater degree of country ownership, the PRS is seldom purely and simply a statement of the government's priorities. In several countries, multi-year planning instruments favored by the political leadership continue to function alongside the PRS. Instruments--such as pro- poor budgeting, medium-term expenditure frameworks (MTEFs), and sector- wide approaches (SWAps)--that have been used to link budget allocations to PRS priorities do not entirely overcome this problem. Also, though many PRS countries have adjusted their budget processes to make budgets more outcome oriented and therefore more PRS friendly, the reforms have often had limited effect. A simpler approach to reform, with the limited objective of making budgets more responsive to current policy priorities, appears to have worked better in the higher-income reformers and, in a different way, in Uganda. Even so, given the efforts to increase the outcome orientation of budgets, greater use of budget reporting mechanisms in PRS processes is both desir- able and possible. Efforts to do this must be realistic about the opportunities and constraints that country systems offer. The case studies suggest that dis- connects and incentive problems in reporting systems may be more effectively addressed with an approach that is not too ambitious but is simple and centrally led, and makes use of existing systems while gradually improving them. This chapter draws together the strands of this argument and suggests a way forward. Key Findings The argument of the book supports three principal findings. These con- cern, respectively, the causes of relatively weak domestic accountability in 53 54 The Agenda Ahead PRS countries, the causes of the fragmentation that characterizes country systems, and how PRS-budget integration has been hampered by prevailing approaches to reform. When ownership is contested and incentives are poor, domestic account- ability tends to be weak Accountability, in the context of this study, has been conceived as the rela- tion between principals and agents for the implementation of specific tasks. In looking at the case study countries, it soon became apparent that the objective of strengthening domestic accountability requires a more nuanced understanding of some of the factors that shape the reality of accountability in PRS countries. Ownership by a restricted group of technocrats and senior civil servants in central ministries is often treated as a sufficient basis for strengthening PRS-budget links. However, the case studies confirm that this is insufficient to generate the incentives needed to bridge the various disconnects that are typical of both PRS and budget systems in PRS countries. Heavy donor involvement in government processes bears the risk of undermining account- ability for content, because it can shift the focus of attention toward routine processes and consultations.1 In thinking about PRS-budget links, there is a clear need to consider the level of ownership of different policy instruments and the budget process itself by presidents, cabinet ministers, and parliamentarians. The degree of wider societal ownership, and the extent to which this brings with it an effective demand for delivery against stated policy objectives, is another dimension that needs to be further explored. Policy making, planning, and budgeting are embedded in fragmented processes and institutions . . . Planning and budgeting processes in the case study countries suffer from a series of gaps and fractures that present numerous challenges for integrating PRSs and budgets, at the formulation, execution, and reporting and account- ability stages. These gaps include · institutional fragmentation, with responsibilities for planning and bud- geting, and for recurrent and investment expenditure, borne by different ministries or by different units within central and sector ministries; · fragmentation of reporting processes that is exacerbated by donor requirements for separate reporting on multiple interventions; · lack of results orientation in budget reporting; The Agenda Ahead 55 · duplication, overlap, and inconsistencies in planning instruments, which include Poverty Reduction Strategy Papers (PRSPs), other gov- ernment strategies, and sector plans; · weak budget execution systems and informality, which create gaps between policies, their implementation, and their monitoring; and · information gaps, in terms of missing building blocks for reporting, missing transparency mechanisms to make useful information avail- able to all stakeholders, and lack of application of such information to policy-making processes. . . . that are hard to tackle using mainly technical fixes Donors' main response to the gaps identified above has been to promote a series of technical reforms, ranging from the identification of priority pro-poor spending programs to the introduction of MTEFs and program budgeting. These reforms have encountered many problems in implemen- tation, partly because of serious capacity constraints but also because of poor design and lack of attention to some of the underlying factors and incentives identified above. Little attention has been given to core decision- making and policy processes, to institutions that enjoy higher levels of political ownership, or to the incentive structures that these generate (see also Stevens 2004). Practical Lessons and Entry Points for Reform Lesson 1: Focus on strengthening and harmonizing existing processes and adopt a gradual approach to reform Build on existing processes. Tailor interventions to local context and realities. Each country has a specific set of planning and budgeting institutions and instruments, often predating the PRS initiative, and sets of actors with differ- ent strengths and interests. In each case, therefore, a different constellation of ownership and incentives will shape the ways in which domestic accountabil- ity can be strengthened, and in which fragmentation can be addressed through donor interventions to promote the integration of PRS and budget processes. Adapting and strengthening existing reporting mechanisms, rather than relying on new ones, would be an obvious useful step in this direction. Adopt a gradual approach to reform and allow it enough time to work. Rein- forcing existing processes to increase the focus of budgeting on results requires not only technical capacity but also a change in culture and thinking, in the sense of organizational and institutional capacity development. 56 The Agenda Ahead Experience in higher-income countries shows that successful budget reform processes tend to start simple, evolve gradually, and use an integrated approach that links improvements in public financial management with broader civil service reforms. Lesson 2: Build support from within, through high-level ownership of policies, a challenge function within the executive, and clear sector roles Intragovernmental accountability, starting at the political (cabinet) level, is a powerful tool in promoting reform, as experiences in higher-income countries have shown. PRS countries have given a great deal of attention to opening up the PRS process to participation, particularly by civil society actors. This has raised the public profile of poverty reduction in these countries, and in some cases promoted a higher degree of technical and societal ownership of PRSs. But accountability mechanisms internal to the executive government have been largely overlooked in the PRS approach, even in recent years when the early neglect of parliaments has been corrected to some extent. Depending on the country context, some initiatives that may help to catalyze the improvement of core policy processes and internal account- ability include the following: Introducing or reforming cabinet committee structures to help strengthen high-level political ownership and a challenge function within the executive. This could be achieved by encouraging the establishment of cabinet com- mittees for reviewing and monitoring different areas of policy formulation and implementation, and for reviewing the budget. Such structures may be mirrored in the civil service and the committee structure of parliament. These committees, rather than the cabinet as a whole, would then be formally involved in the PRS, sector policy, and budgetary processes. Also, leadership training programs, such as those that Madagascar started after its crisis, could prove beneficial in strengthening governmental ownership and account- ability (Box 4.3). Building policy links "from the budget up," rather than "from the PRS down." Successful reform efforts have focused on core policy processes, and on turning budgets into policy-oriented tools, rather than starting from long- term plans and translating them into annual budgets. In contrast, efforts at integrating PRSs and budgets usually take the PRSP as a starting point. The common practice of structuring PRSPs in terms of "pillars," "clusters," and "objectives," however, makes them difficult to map in to budgets, which are usually structured by sector and/or administrative units. This applies in most case study countries, even in cases where program or performance budgeting has been introduced. The Agenda Ahead 57 Moreover, when starting with the budget, relatively straightforward efforts to improve the reliability of budget execution are a highly relevant and effective way to improve the policy orientation of the budget process. This should strengthen ownership and enable actors to use the annual budget as a more reliable guide for future expenditures. The example of Malawi shows that such steps can create the momentum for other policy-oriented reforms to follow. Equally important for budget reliability is the predictability of aid, as for example in the case of Tanzania. Introducing a medium-term perspective linked to sector policy objectives creates the foundation for a more focused and budget-oriented discussion around government priorities. PRS policy objectives and priorities can and should still be used as a guide, but they should be structured in ways that are "budget friendly" rather than disjointed from existing budget structures and processes. Uganda provides an interesting example in this respect (Box 4.2). Introducing a strategic phase in the budget process can ensure that there is a phase, before the preparation of detailed operational budgets, in which sector managers and the cabinet review high-level policy priorities and their impact on intersectoral resource allocation. MTEF processes, where functional, can provide an obvious arena for such strategic dialogue. Developing sector-level policy processes. While the PRS should be seen as a mechanism for developing high-level policy priorities, sector ministries and institutions should be charged with establishing processes to elaborate sector policies and budgets and review policy implementation, and their roles should be clearly defined. At the same time, such sector processes need to be integrated with centralized policy making and avoid being driven by sector- specific funding. Lesson 3: Foster incentives for integration: target reporting to decision- making processes Reporting instruments best serve their purpose when they are instigated by, and remain linked to, specific decision-making processes involving specific actors that are likely to demand and use the information. Budget formulation instruments are automatically integrated into the decision-making process that leads to the approval of the budget, and are therefore tied to reporting on budget performance. It follows that strengthening reporting on budget execution, within the year and at year end, should be given priority to ensure that that information feeds into budget documentation. Where the budget formulation process makes reference to the policy objectives of ministries and departments, outcome information could be introduced alongside financial reporting on the budget. 58 The Agenda Ahead APRs are generally produced to satisfy external demands for information and lack a direct link to national decision-making processes. Their usefulness for internal accountability could be greatly enhanced if PRS performance goals could be made more consistent with the goals articulated in the budget. Closing this gap may require a careful review of existing measures and indi- cators. The discussion in budget circles is likely to concentrate on the lower levels of the results chain--that is, outputs and activities--whereas PRS tech- nicians may have a clearer idea about outcome objectives (Figure 5.1). Better targeting of reports, both financial and performance related, to users' needs could facilitate and improve decision making. Table 6.1 provides a simple typology of how reports resulting from different processes produced by government agencies could be targeted to the different audiences that demand them, including donors, finance ministries, parliaments, and presi- dents. It assumes that better alignment of decision-making functions will increase the likelihood of a successful merger of the monitoring processes. Lesson 4: Keep it simple Strengthening the results orientation of budgets does not necessarily require sophisticated technical solutions. As the case studies show, comparatively simple budget reforms can significantly improve the budget's responsiveness to policies, and consequently to the PRS. Such reforms start with making budget execution credible and transparent. A simple rearrangement of the budget, grouping administrative units into broad sectors, can help show how resources are being allocated to broad policy priorities. Similarly, presentation of the recurrent budget alongside the development budget for each adminis- trative unit is an important step in integration. The identification of priority votes linked to the PRS can also emphasize the link with the PRS. Finally, presenting sector performance information (output- and outcome-related) alongside broad sector budget allocations is a simple way to start linking policy performance with budgets. Table 6.2 shows how relevant information on past performance can be included in budget and strategy documents. Getting budget allocations to reflect government's strategic priorities, as the experience of some case study countries shows, does not require full- blown performance or program budgeting. A strong economic classification system, combined with an administrative budget structure detailed enough to identify useful subsectoral spending categories, can be sufficient for policy-oriented budgeting. Once such systems are established, program budgeting in a simple form can help improve the situation, but an early move to full program and performance budgeting is likely to undermine strategic decision making. Table 6.1 Targeting Reporting to Specific Decision-Making Processes Target decision- Relevant information making processes Target stakeholders Strategic Assessment of poverty outcomes drawing on PRSP and sector strategy Ministries of finance and planning Strategic policy survey and other outcome data, linking back formulation and review Sector policy makers documents to the implementation of government processes Parliament policies and strategies. Cross-sector resource Donors Highlights of key policy achievements and allocation mechanisms by outputs by sector in the context of past the Ministry of Finance and sector expenditures and future allocations. cabinet (e.g., MTEF) Annual budget A synthesis of sector reports, highlighting Budget formulation process, President and cabinet documents major outputs and policy actions by sector, and in particular cross- Ministries of finance and planning linked to expenditures against budget. sector resource allocation Parliamentary budget committee A summary of key results achieved by individual mechanisms Budget support donors institutions alongside targets for the coming 59 budget year, including expenditure estimates and summaries of past and planned expenditures by program and major economic item. Annual sector Reporting on outputs, policy actions, and Intrasector resource allocation Ministers and senior civil servants in documents activities in the context of sector strategies mechanisms (sector level sector institutions and budgets, linking to outcome data when MTEF/budget process) Sectoral parliamentary committees available and appropriate. Sector policy making/review Cabinet committees Setting out past performance in terms of forums (e.g., SWAp related) Sector civil society organizations outcomes, outputs, and policy actions at a Sector donors sector level alongside future projections/ targets in the context of past expenditures and future budget allocations. In-year Comprehensive reporting on implementation Intra-institution managerial Managers within institutions institutional of institutional operational plans and decision-making processes Ministry of finance and central ministries documents budgets, including inputs, activities, and Dialogue between sector/ Sectoral parliamentary committees outputs, and relating these back to institution and Ministry Beneficiaries of services where relevant expenditures against the budget. of Finance/parliamentary Donors providing support committees to institutions Operational Source: Authors. Table 6.2 Integrating Performance Information into Budget Documents Main target decision-making Relevant information processes and stakeholders National Highlights of key policy achievements and ­ Cross-sectoral monitoring and resource Strategic MTEF/budget outputs by sector in the context of past allocation mechanisms strategy sector expenditures and future allocations ­ President and cabinet documents ­ Ministry of Finance and Planning ­ Parliamentary committees ­ Donors Annual A summary of key results achieved by ­ Cross-sectoral monitoring and resource national individual institutions alongside targets allocation mechanisms budget for the coming budget year including ­ President and cabinet 60 documents expenditure estimates and summaries ­ Ministry of Finance and Planning of past and planned expenditures ­ Parliamentary committees ­ Donors ­ Civil society Sector Setting out past performance in terms of ­ Intrasectoral monitoring and resource budget outcomes, outputs, and policy actions at allocation mechanisms strategy a sector level alongside future projections/ ­ Cabinet and parliamentary committees documents targets in the context of past expenditures ­ Ministry of Finance and future budget allocations ­ Sector civil society organizations ­ Sector donors Institutional Information on past activities, outputs, and ­ Intra-institutional monitoring and operational expenditures linked to implementation resource allocation mechanisms budgets of operational plans and budgets ­ Managers within institutions (including delivery of services) ­ Cabinet and parliamentary committees ­ Donors providing support to the institution Operational Source: Authors. The Agenda Ahead 61 As budget systems evolve, more ambitious links between a PRS and the budget can be considered, that take into account the types of ownership that are needed at each level. Structuring a PRSP in a more budget-friendly manner, for example, by sector, would facilitate the interface with the budget. If the PRSP were to be considered as a high-level statement of sector policy priorities, rather than a fully elaborated strategy for the whole of government, then sectors would be responsible for elaborating those policy priorities, cost- ing them, and establishing the resource implications. This approach to linking the PRS with the budget would therefore not entail a comprehensive up-front costing of the PRSP, but an iterative process of delegating responsibility to sectors. The first step would be, at a high level, to draw conclusions on the intermediate steps and policy actions that are needed to meet strategic objectives on current evidence, and to examine whether the current budget allocations are broadly right. The second step would be to request sectors to elaborate policies and costing, so as to enable a further refinement of resource allocation, and so on. Such an integrated approach would potentially be more efficient and inclusive. Conclusion The integration of PRSs, budgets, and the corresponding reporting processes has the potential to strengthen domestic accountability and the implementa- tion of pro-poor policies. Yet, to achieve this requires a more thorough awareness of the incentives underlying domestic accountability and country ownership. Among donors, it calls for a renewal and reinforcement of the commitment made following the 2005 PRS Review to build on what exists, and to pay attention to what has political traction as well as positive effects on development at the country level. Findings · When ownership is contested and incentives are poor, domestic accountability tends to be weak. · Policy making, planning, and budgeting are embedded in frag- mented processes and institutions that are hard to tackle with mainly technical fixes. Lessons · Strengthen and harmonize existing processes and adopt a gradual approach to reform. To enhance the policy-making function of the budget in line 62 The Agenda Ahead with PRS priorities, promote a flexible approach to PRS design that responds to the domestic policy environment and political dynamics and provides a structure that is aligned with sector policies as reflected in the national budget; use and strengthen existing systems and reporting mechanisms; and adopt a more coordinated and integ- rated approach to budgeting and reporting that helps to bridge and minimize institutional fragmentation. Such change takes time, so manage the expectations of donors and governments alike. · Build support from within. High-level leadership for reforms; a strong role for the cabinet in the strategic phase of the budget process, to provide a challenge function within the executive, as well as outside; and a clear delineation of central and sectoral responsibilities help to increase the coherence of planning and budgeting systems. Under- stand the core policy and accountability processes internal to govern- ment at cabinet and sector levels before considering how the PRS can interface with and strengthen them. These factors are important in addition to the challenge function by parliament and civil society, which has received more attention in the past. · Foster incentives for integration. Targeting PRS and budget reporting on actual decision-making processes helps to reinforce domestic incentives to produce and use information. Providing easy access to quality information (especially sufficiently aggregated budget execu- tion information in combination with information on activities and outputs) helps decision makers to monitor policy implementation and take corrective measures. · Keep it simple. A strengthened results orientation within the budget need not require sophisticated technical solutions. Simple budget reforms can significantly improve the budget's responsiveness to policies. Such reforms start with making budget execution credible and transparent. Structuring a PRSP in a more budget-friendly manner, for example, by sector, would facilitate the interface with the budget by involving sector agencies in elaborating policy priorities and estab- lishing the resource implications. It could expand ownership and boost incentives for the integration of a greater number of stakeholders, thereby strengthening domestic accountability. Part Two Summaries of Country Case Studies Part Two provides summaries of country reports prepared by Artan Hoxha (Institute of Contemporary Studies, Albania), Sawadogo Malick (Burkina Faso), Veronique Hubert (Madagascar), Philipp Krause (Malawi), Verena Fritz and Florian Lang (Mali), Paolo de Renzio and Jose Sulemane (Mozambique), Vincent de Boer and Robert Cook (Rwanda), and Tim Williamson (Tanzania and Uganda). 7 Albania: Creating an Integrated Planning and Reporting System Formulated on the basis of a broad consultative process, Albania's National Strategy for Social and Economic Development (NSSED) has helped the govern- ment to introduce good practices in setting strategic objectives and identifying priorities. But the NSSED's priority actions are not costed or prioritized, and though its goals are reflected in the annual budget documents, the links between the strategy, the medium-term budget plan, and the annual budget are weak. Reporting on the budget and the NSSED also remain detached from one another: budget reporting focuses on the proper use of public funds, not on their effectiveness in achieving policy goals, while NSSED reporting focuses on inputs and out- puts, rather than the outcomes and impacts of policies and the resources spent on them. The NSSED is only one of several planning frameworks in use in Albania, but donors have recently started to acknowledge it as the government's main policy document. The government has also begun implementing an Integrated Plan- ning System (IPS) that promises to help to streamline the planning of the NSSED and the medium-term and annual budgets. Country Context Since 1997 Albania has been able to maintain macroeconomic stability, keeping growth rates at an average of 6 percent per year and inflation down to 2­4 percent, while reducing the fiscal deficit to 3.4 percent of GDP at end 2005. The country has also made significant progress in privatization and trade liberalization. Albania is a member of the World Trade Organization and has signed free trade agreements with all its neighboring countries. Successful macroeconomic management has been an important contributor to 65 66 Albania: Creating an Integrated Planning and Reporting System growth--itself a critical factor in poverty reduction--including over the last three years. The headcount rate for absolute poverty declined from 25 percent in 2002 to 19 percent in 2005, lifting 221,000 people out of poverty. This decline is faster than witnessed recently in many middle-income countries. Meanwhile, the fight against poverty is far from over. While the rapid growth in incomes was the key contributor to poverty reduction, it was partially offset by increasing inequality, especially in coastal and central areas. Rural poverty remains high, with the headcount still at 24 percent in 2005, compared to 11 percent in urban areas. GDP per capita is US$2,500, and unemployment remains relatively high, as does the proportion of employment in the informal sector. Economic growth in Albania, though rapid, has been constrained by the country's poorly developed infrastructure, poor access to basic services, weak governance, and political instability. Growth since 1990 has been driven mainly by a fast rise in total factor productivity, which is now slow- ing down significantly. Capital accumulation has been negligible, raising concerns about the sustainability of the growth rates needed to alleviate the high poverty level. The trade deficit is equivalent to 21 percent of GDP. Remittance inflows, at 14 percent of GDP, are the major source of financing for this deficit. Furthermore, though Albania has received substantial assistance from a large number of bilateral and multilateral donors over its 15 years of transi- tion, its achievement of middle-income country status implies a gradual reduction in access to concessional financing. Against this background, Albania adopted the NSSED in 2001 as a com- prehensive national development strategy to promote growth and reduce poverty over the period 2001­06. The strategy rests on two pillars: improved governance and strong economic growth. It recognizes private sector devel- opment as the driving force behind growth and poverty reduction and also identifies health care, education, and infrastructure as priorities. The NSSED has not been the only policy framework that guides the gov- ernment's decision making and resource allocation. Albania also puts significant emphasis on the Stabilization and Accession Agreement (SAA) process with the European Union (EU), which involves both economic and institutional reform; on the NATO integration strategy; and on achievement of the Millennium Development Goals (MDGs). These parallel policy frameworks come with their own action plans, reporting systems, priorities, and resource requirements. For a number of years, modest efforts to inte- grate them under the NSSED have not been able to address fragmentation and overlapping, and have failed to provide a clear vision and direction for Albania: Creating an Integrated Planning and Reporting System 67 public policy design and implementation or a clear focus for donor coordi- nation. During 2005 the government made efforts to reconcile these separate agendas by adopting an IPS and building a comprehensive new National Strategy for Development and Integration (NSDI) (2007­13). Linking the PRS and the Budget The NSSED has helped the government introduce good practices in setting policy objectives and identifying policy priorities. It was formulated at the Ministry of Finance (MoF) on the basis of a broad consultative process that was innovative for Albania. The NSSED has increased the focus of the gov- ernment and other stakeholders on setting policy priorities, provided space to civil society for discussing national priorities and policies for poverty reduction and growth, and, to a certain extent, increased donors' awareness of national priorities, providing a focus for aligning their support. But the strategy's priority actions are not costed or prioritized. The NSSED influences budget planning at the moment when sector allocations are discussed and decided. The 2006 budget reflects the broad priorities set forth in the NSSED, including education, health, infrastruc- ture, and social protection, and it gives strong emphasis to strengthening fiscal discipline and transparency, to reducing administrative expendi- tures, and particularly to improving public expenditure management. Beyond this stage, however, the link between the NSSED and the budget has been weakened by several factors: · The MoF, which is the driving force for Albania's Medium-term Budget Program (MTBP) and annual budget process, sets indicative resource ceilings for sectors with limited involvement of the line ministries at the outset. This poses doubts about whether the ceil- ings match the priorities of sectoral strategies. Line ministries do not respect these ceilings, and most of them recurrently bid for more resources. · Budget projections have tended to be optimistic; actual expenditures were as much as 15 percent over budget during 2002­04. Projections of foreign financing, notoriously difficult, show a much wider differ- ence and are the weakest element of Albania's budget planning. · Albania's macro-fiscal framework has often been subject to revisions by the MoF after the line ministries have prepared their budget proposals. (The revised projections often cause the MoF to make across-the-board cuts in the line ministries' expenditure plans to safeguard fiscal stability. 68 Albania: Creating an Integrated Planning and Reporting System Such cuts may endanger the achievement of the NSSED priorities embodied in the initial budget proposals, and they diminish the cred- ibility of the planning and budgeting process.) · The budget is approved at the program rather than the project level, allowing expenditures on individual projects to be amended in the course of budget implementation as long as aggregate and program expenditures remain within their approved levels. While such an approach ensures fiscal stability and is often standard practice in other countries, large in-year reallocations in Albania have under- mined the credibility of annual budgets and weakened links with the MTBP and the NSSED, both at the MoF and line ministry level. · The budget process remains largely input oriented and based on incremental proposals. This approach restricts the effectiveness of expenditures and the achievement of policy goals. The decision-making process for the draft annual budget involves the Council of Ministers and Albania's parliament. The discussions at both these levels generally focus on the adequacy of allocated funds, rather than on providing or debating policy alternatives. During the NSSED implemen- tation, efforts have been made to improve the link from the strategy to the budget, both medium term and annual, but the achievements are far from the desired level. Integrating Reporting on the PRS and the Budget Each of the government frameworks for setting policy priorities--including the NSSED, EU integration, and NATO integration--has different monitoring and reporting requirements. Budget monitoring and reporting has remained quite detached from the reporting on the NSSED. It is regulated by the Organic Budget Law and guidelines issued by the MoF, and is not linked with any other monitoring systems. The annual report on the execution of the budget is first discussed and approved in the Council of Ministers and then presented to parliament in a plenary session. High State Control conducts an audit of the budget implementation every year and submits a report to parliament (and thence the public). Budget monitoring and reporting does not shape the progress reports on the NSSED or inform the strategy's stakeholders on how budget performance has affected sector outcomes. It focuses mainly on macro-fiscal indicators, and would need to radically change its focus to comply with the needs of reporting on the NSSED. While there are no evident technical or political obstacles to such a change, there are no incentives or requirements for it Albania: Creating an Integrated Planning and Reporting System 69 either. An unchallenged tradition of formulating input-based budgets and reporting only on financial status makes the MoF and line ministries unin- terested in measuring and reporting the impact of expenditures on the achievement of policy goals. Neither have the findings of annual progress reports on the NSSED sig- nificantly influenced decisions on the budget or sectoral policy adjustments so far. Three such reports have been prepared since the strategy was intro- duced in 2002. These reports are compiled by an inter-ministerial working group that is chaired by the MoF, and are based on sector progress reports provided by the monitoring and evaluation (M&E) units of line ministries. While these ministries show adequate ownership of the process, their reports largely provide quantitative data on inputs and outputs, with little focus on the evaluation of policies and their impacts. They have highlighted shortcomings in achieving policy goals but made no recommendations on how to follow up. Thus no strong evidence is available on whether budget implementation furthers the achievement of NSSED priority goals. Various useful changes are being made in the area of NSSED reporting and monitoring. The annual progress report on the NSSED is becoming the major mechanism for monitoring and reporting on public policy priorities, and now incorporates reporting on the EU integration initiative and on progress toward the MDGs. This is being deepened and institutionalized in the context of the IPS. Line ministries are beginning to cost and prioritize their policies to ensure a better link with the MTBP and the annual budget, and four ministries are undertaking a pilot program to assess the impact of NSSED policies in their sectors. Ownership and Accountability Budget reporting in Albania is quite strongly linked to the domestic account- ability of the executive, as evidenced by the audience for such reporting, noted above. By contrast, as in many other countries, reporting on the PRS is mainly the domain of the government and donors, with parliament retaining only a consultative role. A striking feature of the NSSED annual reporting process is the number and variety of actors involved. The interest of most of these stakeholders is driven by the influence that the NSSED annual progress report (APR) is perceived to exert on the MTBP and annual budget preparation, and their expectation that the NSSED will to some extent affect the government's expenditure plans. Experience shows, however, that in practice the impact of the APR has been limited. Local governments participate in preparing the APR as their way to ensure that the policies embodied in the NSSED are 70 Albania: Creating an Integrated Planning and Reporting System appropriately applied at the local level, and subsequently to receive sub- stantial budget allocations. Civil society provides the NSSED annual reporting process with exper- tise and with contacts at the grassroots level. The government uses civil society networks in different regions to disseminate the NSSED APR and conduct M&E exercises. The interest of civil society groups in participat- ing in NSSED annual reporting is to increase the transparency and accountability of policy decision making, influence policy formulation, strengthen the public pressure on policy formulation, and make the case for specific spending priorities. Donors have recently started to acknowledge the NSSED as the govern- ment's main policy document, around which to streamline their future program and resource allocations. Up to now, their tendency to stick with their own priorities has not been countered by any clear signal from the govern- ment that the NSSED is the only platform for orienting donor support. The long list of priorities of the NSSED and the lack of costing and prioritization of policies has made alignment with the strategy even more difficult. To strengthen the incipient trend, the NSSED needs to become unequivo- cally the main document that orients the government's dialogue with donors, fully incorporating policy frameworks such as the European Partnership Action Plan, pursuit of the MDGs, and sectoral and cross-cutting strategies. The Integrated Planning System The IPS that the government endorsed in 2005 entails a number of changes that will affect the entire process of planning and formulation of the NSSED, MTBP, and annual budgets, better aligning the calendars and making a clear definition of roles and functions of each document to avoid overlapping. The IPS is neither a new nor a separate planning system, nor is it an attempt to amalgamate existing processes within a single process. It is a broad plan- ning and monitoring framework that aims to ensure that the government's core policy and financial processes function in an integrated manner. These core processes are · The National Strategy for Development and Integration (NSDI), which establishes the government's medium- to longer-term goals and strategies for all sectors. · The MTBP, which requires each ministry to develop a three-year plan to deliver program outputs to achieve its policy objectives and goals within the ministry's expenditure ceiling--as set out in the govern- ment's fiscal plan. Albania: Creating an Integrated Planning and Reporting System 71 The IPS is seen as a significant contributor to enhancing the govern- ment's ability to deliver the EU integration agenda, for three reasons: first, because EU integration will be at the center of the NSDI and properly reflected in every sectoral and cross-cutting strategy amended or developed as part of the NSDI process in 2006 and 2007; second, because for key decision makers in the two key interministerial committees the IPS process will highlight where priorities need to be set and implementation monitored; third, because policies--including commitments made within the Interim Agreement (Albania), SAA, and Plan for Implementing the Accession--have not been adequately linked into the budget development process. The IPS will strengthen the link between policy and budget by ensuring that the MTBP process is informed by the sector strategies developed through the NSDI (which in turns reflect EU integration commitments) through the work- ing groups on strategy, budget, and integration that are now operating in all ministries of the government. The IPS will need three to five years to yield its desired results. It will have seven core products: · an integrated planning calendar will be issued to ministries at the beginning of each year covering the major requirements and dead- lines for all core policy and financial processes; · combined training programs will be delivered wherever feasible; · each year, a ministry integrated plan will be produced containing the major commitments for the core policy and financial processes; · a single external assistance strategy document will be produced annually that identifies all priority areas where external assistance is needed; · a single ministry monitoring plan will be negotiated annually with each ministry, encompassing all key outputs and indicators to be monitored by the center; · an IPS information system will serve as the primary system for inte- grated reporting on progress toward achieving the outputs and indi- cators contained in the ministry monitoring plan; and · ministry annual reports will present the results achieved vs. the targets set for all processes. Conclusions and Lessons Suggestions for Improvement in Albania To avoid the need for regular expenditure cuts within the year, the macro- fiscal projections made in the early phases of MTBP and budget planning 72 Albania: Creating an Integrated Planning and Reporting System should be based on the previous year's budget implementation figures rather than on the planned budget for the current year. The budget process--from its formulation in line ministries to its imple- mentation and reporting--needs to focus on results. To be able to link the NSSED with the MTBP and budget processes requires strengthening the insti- tutional and technical capacities of line ministries. Program policy reviews and program expenditure planning that clearly link the NSSED with the MTBP and budget exercises need to become obligatory in all line ministries. To avoid having to change line ministries' projects after the budget is approved, a public investment management system with robust procedures for project identification, appraisal, selection, implementation, and monitor- ing should be consolidated as part of the MTBP process and supported by all donors. The list of approved projects should be attached as indicative to the draft annual budget when sent to parliament. Reporting on fiscal performance is important, but needs to focus on results. Budget reporting, especially the annual budget report, should show the extent to which expenditures have served the achievement of policy goals set forth in the NSSED. High State Control should also assess whether the use of funds complies with policy priorities. These issues need to be addressed by Albania's new Organic Budget Law, currently being revised, and enforced so that links between policy priorities and budget allocations can guide the entire budget process, from formulation through approval to implementation. To ensure that resource allocations are consistent with priority goals, the key requirement is to introduce meaningful performance indicators that link the sector expenditures undertaken during a year with the achievement of NSSED policy goals. This is supposed to be the objective of the newly introduced Program Policy Reviews and Program Expenditure and Investment Plans. Individual policy goals of NSSED need to be costed. Such information is crucial to ensure coherence between these goals and the resources available in the medium-term and annual budgets. Costing and prioritizing the actions specified in the NSSED will allow M&E to be more results oriented. The indicators used for M&E need to measure outcomes, as well as inputs and outputs. Major resources need to be committed for measuring the impact of the line ministries' policies and expenditures. Only then will the consultation process become more substantive and will the annual progress reports on NSSED be able to provide policy makers with enough informa- tion to adjust policy goals if necessary. Permanent civil society partners need to be identified, to institutionalize the involvement of civil society in policy making. Changing the current Albania: Creating an Integrated Planning and Reporting System 73 practice to allow discussion of the NSSED in parliament would institution- alize the involvement of the legislature in the NSSED process, and requiring the NSSED to be approved by parliament in plenary session would confer national ownership on the strategy and thereby significantly increase domestic accountability for its implementation. Because donors have only recently started to acknowledge the NSSED as the government's main policy document they have not substantially aligned their support with the priorities of the strategy. In this regard the adoption of the IPS, which brings together the function of planning and coordination of foreign aid with the function of preparing the NSSED, is likely to increase the predictability of donor funding of the NSSED policy priorities. The actions, indicators, and targets associated with the IPS should be brought under a single M&E matrix, which will help to reduce the burden of reporting, which is huge for the limited human capacity that is available. Lessons for Other Countries · Requiring the PRS to be approved by parliament is important to ensure national ownership of the strategy and increase domestic accountability for its implementation. If shortcomings in achieving policy goals are highlighted through the APRs, they need to be fol- lowed up with clear action plans to address them. · Integrating all strategies and parallel policy frameworks into a single document increases the chances of achieving a better link between strategic planning and budgeting. Adoption of parallel policy frame- works undermines the PRS process, and creates confusion and report- ing burdens for the entire administration. Streamlining all donor assistance to support PRS priorities reduces the demands associated with parallel accountability. Establishment of a single institutional structure in charge of formulating policy priorities and coordinating external assistance is a good way to foster domestic accountability. · Costing and prioritizing PRS priorities ensures a better link with the medium-term and annual budget processes. Formulating results- focused budgets that highlight the effectiveness of expenditures is necessary to counter a tradition of input-based budgeting based on historical data. · The PRS-budget link is undermined if ownership is limited, compro- mising the reliability of the budget formulation process. · Establishment of a robust system of public investment planning and programming integrated within the medium-term budget process should help improve the rationale for selection of projects, and 74 Albania: Creating an Integrated Planning and Reporting System consolidate the large number of small projects into larger projects that are clearly linked to the sectoral priorities included in the PRS. · Unless the budget reporting system is linked to results it will be impos- sible to evaluate whether PRS goals are achieved. Accountability demands with a narrow focus on fiscal performance give no indication of whether policy goals have been achieved. 8 Burkina Faso: Aligning Donor Procedures with National Systems Burkina Faso's experience shows that it is difficult to integrate the budget and Poverty Reduction Strategy (PRS) processes when the key structures responsible for managing the two processes do not report to the same authority. Budget program- ming and planning tools (medium-term expenditure frameworks [MTEFs], action plans, and program budgeting) can be potential integration vectors if they are well mastered and effectively used in decision making, but they need to be harmonized and integrated into both the budget process and the PRS process. The case also shows that donors can play a positive role in strengthening domestic accountability. Where national institutions are still too weak to provide a real challenge to the executive, the role of development partners becomes critical. Aligning donor proce- dures with national systems could simplify the choice of reporting requirements and, indirectly, augment the power of stakeholders in domestic governance. Country Context Burkina Faso has about 42 percent of its population in poverty (2006). Economic growth and macroeconomic performance depend heavily on agriculture, and especially on revenues from cotton, the main export crop, which are subject to fluctuations in weather and world market conditions. Recent macroeconomic performance has been fairly satisfactory, and per capita GDP growth was expected to be as high as 4.6 percent in 2005. Public spending is fairly well managed and there are no arrears in external debt payments. The government budget deficit was 5.4 percent of GDP in 2005, and a weakness of the country's public finances still results from the poor mobilization of revenue; the tax burden was only 11 percent of GDP in 2005. 75 76 Burkina Faso: Aligning Donor Procedures with National Systems Since 1991 Burkina Faso has had a democratic system based on free elections and a constitution that provides for the separation of powers among the executive, the National Assembly, and the judiciary. The ruling party dominates the political process. Though in 2002 the opposition won a large number of seats in parliament for the first time, it remains weak and divided. A decentralization process was launched in 1998. Decisive steps occurred recently with elected rural councils and the establishment of 13 regional governorates. Official development assistance (ODA) from a wide range of bilateral and multilateral donors is equivalent to nearly 30 percent of Burkina Faso's GDP. The IMF and several nongovernmental organiza- tions are also involved. Donors have been increasing their use of budget support, but 70 percent of the ODA still takes the form of project and pro- gram financing. The first PRS, adopted in 2000, is Burkina Faso's unifying framework for national policy. It forms part of a larger package of reforms, including long- term development plans for health, education, and rural development (adopted in 2004), the Program to Strengthen Budget Management Capacity (2002), and the National Good Governance Policy (2005). A national anti- corruption policy is being drafted. Local Poverty Reduction Strategy Papers (PRSPs) are in preparation. The revised version of the PRS, issued in 2004 for the period 2004­06, is accompanied by a Priority Action Program (PAP) incorporating a range of sector program investments that contribute to poverty reduction. Linking the PRS and the Budget The integration of the PRS and budget processes, which are managed by different ministries, is in its preliminary phase and is still weak. The PRS process is managed by the Ministry of the Economy and Devel- opment (MEDEV) and does not enjoy the same level of institutionalization as the budget process. The revised PRSP for 2004­06 and the accompanying PAP are the working documents for planning. The credibility of the budget process is weak. Because of domestic revenue shortfalls, the Ministry of Finance and Budget (MFB) has the right to adjust and reduce line ministries' budget allocations to be able to meet the country's deficit criteria. This regulation has discouraged line ministries from preparing responsible program budgets. Burkina Faso's budget preparation and execution are highly centralized. However, some improvements have been made in this area (crédits Burkina Faso: Aligning Donor Procedures with National Systems 77 délégués), and efforts are ongoing to support the deconcentration of budget execution under the PRSP process. The MFB is the key actor and exerts considerable power in budget matters over the sectors and peripheral structures. MEDEV is responsible for preparing investment budgets as part of the budget preparation process. The two ministries had to collaborate in revising the PRSP, but their collaboration often proved difficult and led to coordination problems. Parliament plays a secondary role in the budget process, but parliamentary commissions hold hearings in which they question representatives of the executive. Civil society organizations are not active in the process. Budget preparation is well established and institutionalized. The pro- cess begins around March each year with the preparation of the MTEF by MFB. Subsequently MFB produces the budget circular, which is signed by the president of Burkina Faso and signals the preparation of the Budget Law for the year. The next phase is the macroeconomic framework projec- tions, and this is followed by the intersector and intrasector budget allocation phase led by MFB and the spending ministries. A final arbitra- tion is made by the Council of Ministers. The last phase of the budget process is the examination and adoption of the budget proposal by the National Assembly. Integration between the PRS and the budget is most evident in budget preparation and planning. The nine priority sectors identified in the PRSP receive priority in the budgetary process, through the MTEF. In principle, several recent reforms and measures promote the integra- tion of budget planning with the PRS: · The Program to Strengthen Budget Management Capacity includes provisions relevant to PRS goals, such as the introduction of sector MTEFs in education and in health, and the continuation of budgetary decentralization in the regions. · Significant recent efforts have been made to align macroeconomic framework projections with the PRS and to harmonize the annual reporting on the PRS with the MTEF process. · The resources available to Burkina Faso under the heavily indebted poor countries (HIPC) initiative and resources from budget support have been integrated into the Budget Law. This integration has encouraged more transparent and predictable resource allocation for sectoral ministries and the MFB. · Largely at the impetus of donors, several attempts to develop more results-oriented approaches to budget planning are incipient in the 78 Burkina Faso: Aligning Donor Procedures with National Systems health and education sectors, though to make faster progress on them will require changes in civil service incentives. · Work in the health sector may provide a model for integrating pro- gramming and budgeting within sectors. The MTEF in use in the sector integrates the budgeting of the additional expenditure programs directly with the poverty reduction objectives that are embodied in sector action plans. Budget programming and planning tools (MTEF, action plans, and pro- gram budget) can be potential integration vectors if they are well mastered and effectively used in decision making, but they need to be harmonized and integrated into both the budget process and the PRS process. The impli- cations of this change at the institutional level and for capacity building were underestimated in Burkina Faso. Two factors still discourage a stronger alignment between budgeting and the PRS: · Sector ministries' access to their own external financing, indepen- dently negotiated with donors, reduces the impact of the budgetary process on these ministries. Donors are partially responsible. Many donors are sector oriented and do not leave the government enough flexibility to decide where to use their money. · The actors in the budgetary process (sector ministries, civil society organizations [CSOs]) do not all have the same information about the regulations and modalities for allocating public resources. This situa- tion creates a bias in favor of the so-called protected social sectors and no incentives for the other sectors to perform. Integrating Reporting on the PRS and the Budget The alignment is not yet perfect between reporting on PRS implementation and on budget execution, and at the sector level better links are needed between the PRS annual report and sector reviews. Budget monitoring in Burkina Faso is essentially a four-pronged process, dealing with the execution of current expenditures and investment expenditures, treasury operations, and, since 2000, the efficiency of public expenditure as measured by public expenditure reviews. Budget reporting uses the format of the line-item budget, which can be displayed by adminis- trative units and spending ministries. Budget reporting is mainly perceived as an accounting tool rather than a political or results-oriented instrument. A revision of budget monitoring tools--notably the government's financial operation table and the periodic budget execution tables--was Burkina Faso: Aligning Donor Procedures with National Systems 79 considered to take into account the tracking of social expenditures for poverty reduction and HIPC funds. The main changes are the inclusion of a separate account dedicated to HIPC funds and the eligible sectors of the PRS in budget laws. No formal PRS monitoring and reporting schedule is binding on all actors. Since the PRS was revised in 2004, a new institutional mechanism for monitoring its implementation has been set up. This relies in principle on reporting by sector and thematic committees that bring together mem- bers of government and donor representatives. In practice, the committees have not provided the needed reports, and the directorate general of economy and planning fills the gap. The resulting reports are then vetted and approved by the committees. These reports are used as the basis for the annual progress report (APR) on the PRS and for the implementation report on the PAP. The main users of the PRS APR are donors, particularly those that supply budget support. Each year, as the basis for their disbursement decisions, donors carry out a joint evaluation of PRS implementation and of fiscal management based on a performance matrix agreed with the government. Within the government, the MFB uses the APR to prepare the MTEF and as a tool for mobilizing budget support. For sector ministries, the PAP annual report constitutes an important source of information because it covers all key government reforms undertaken during the year, provides a sector policy evaluation framework that could not have been instituted within the normal administrative mechanism, and obliges ministries to produce information necessary for evaluation. Nonetheless, the report does not seem to play a major role in sector performance reviews. And the health and education ministries have instituted their own performance review frameworks. Ownership and Accountability By law, budget execution in Burkina Faso is subject to administrative, judicial, and parliamentary controls: · Administrative control is the responsibility of the financial controller. · Judicial control is carried out by the Audit Office (Cour des Comptes), which has the legal authority to investigate public accounts. Each year it approves the management accounts assesses the proper execution of the national budget for the previous years. · Parliamentary control is greatest at the stage when parliament debates and approves the Budget Law; otherwise, as noted above, 80 Burkina Faso: Aligning Donor Procedures with National Systems parliament plays a minor role in the budget process. It has no power to sanction, and therefore has limited powers to act. In addition, members of parliament lack enough information, especially on the performance of projects and ministries, to challenge the government's decisions. Strengthening the capacities of parliament's finance commission could improve the quality of the draft budget review. PRSP implementation, along with the Program to Strengthen Budget Management Capacity, has helped to improve transparency in the presenta- tion of public priorities and in public finance management. PRSP-related documents are now available to CSOs, which are also invited to participate in annual review meetings. Budget transparency has improved with the intro- duction of a computerized expenditure chain. However, weaknesses persist: · The degree of independence of the various structures to control the executive branch has been discussed, with no satisfactory solution found. There is an independent oversight body for public resources-- the Cour des Comptes, which investigates public accounts--albeit with limited capacity and a generally weak presence. · The poor dissemination of information, especially on budget execu- tion, but also that arising from numerous evaluation and control reports, seems to be a key problem for internal transparency. Clearly, the administration has no tradition of disseminating budget infor- mation. The availability of information, and the opening up of the budget and PRSP processes to the public and to citizens, are thus still limited. · In cases of noncompliance, formal sanctions are poorly applied. Over- sight bodies and inspectorates can impose financial sanctions in case of mismanagement, but the application of such sanctions is very often left to the discretion of political leaders, hampering good governance. The annual progress reporting on the PRSP, with its demands for results- based information, has clearly contributed to the promotion of reporting on outcomes rather than simply inputs and outputs, and can be seen as a precursor to a results-oriented culture. But thus far, the impact of informa- tion generated by the PAP report on sector performance remains limited, and the issue of performance is not yet at the center of discussions within the government or even parliament. Three trends encourage the adoption of results-oriented incentive systems: · The development of sector development plans in education and health, with associated programming and planning frameworks (action plans, Burkina Faso: Aligning Donor Procedures with National Systems 81 MTEFs). This has promoted the development of monitoring indicators for these plans. · The use of budget support as an aid modality. The matrix that budget support donors use to make decisions on disbursements relies on the same performance indicators as the PRS. · The development of new financing mechanisms by donors through the setting up of "common baskets" in health and education. However, such initiatives remain limited. Given its financial clout, the MFB has the potential to play a more important role in encouraging perfor- mance-based management, notably through examination of program bud- gets for sector arbitrations. The MFB has created a directorate within the general directorate of budget that will be responsible for program budget and MTEF implementation. Annual reporting on the PRS has strengthened the internal accountability of administrative systems, and has reinforced the prime minister's role in coordinating and monitoring government action. But it has done little to strengthen accountability systems vis-à-vis parliament and civil society. Parliament does not discuss the PAP annual report, but receives it for information. A growing demand for information by CSOs could strengthen government accountability. Civil society is still trying to position itself with regard to the PRS. CSOs receive, through the National Civil Society Unit, documents and invitations concerning the annual review. But they feel overwhelmed by the workload, and most of those consulted for this study feel that there is not yet a real partnership between the state and CSOs within the framework of the PRS. CSOs' participation in the process is due more to donors' insistence than to strong government willingness. Where national institutions are still too weak to provide a real challenge to the executive, the role of development partners becomes critical. By align- ing their support with national mechanisms for monitoring the PRS and budget management, donors to Burkina Faso create a strong impetus for integrating the two processes. Aligning donor procedures with national systems could simplify the choice of reporting requirements and, indirectly, strengthen the power of stakeholders in domestic governance. Conclusions and Lessons The strong message that emerges from this study is that the integration of the budget and PRS processes is much more likely where (i) efficient budget man- agement systems coexist with operational tools for planning/programming, 82 Burkina Faso: Aligning Donor Procedures with National Systems execution, and control; (ii) good sector policies are combined with perfor- mance and measurement systems and indicators; and (iii) operational inter- nal accountability systems exist. Suggestions for Improvement in Burkina Faso · Better integrate the reporting and planning instruments used in the budget process and in the PRS by improving their quality and use in decision making, and better harmonize sector programming and planning tools (the MTEF and sectoral MTEFs, program budget, action plan). This integration of tools should be done by translating the national poverty reduction objectives into sectoral objectives. · To help incorporate the priorities and ambitions of the PRS in the cur- rent budgetary framework, promote better collaboration between the MFB and the MEDEV. · Strengthen budget execution by improving the predictability of resources for sector ministries, by pursuing reforms on the deconcen- tration of budget execution, and especially by advancing budget decentralization. · Reinforce internal ownership of the PRS and sector strategies by further aligning the PRS process with inter- and intrasector monitor- ing systems. This could be achieved through better integration of poverty monitoring at the sector level and better coordination of joint partner reviews in the education and health sectors and the annual PRS/PAP reviews. · To improve the links between annual PRS reporting and joint sector reviews, improve the quality of information discussed in the APR, by drawing on more analytical studies (public expenditure reviews, impact studies, expenditure incidence) and strengthening dissemina- tion to major country stakeholders, civil society, and parliament. · Revise the incentive systems for implementing the PRS through the budget by better informing key actors (sector ministries, civil society) about the influence of the PRS review and the MTEF on the allocation of resources. Transparent criteria and modalities should be defined for evaluating policies and allocating resources. The overall incentive system within the administration should be addressed in the context of the ongoing reform of the public service. Incentive systems that are tied to donor budget support modalities and conditionalities should be redefined, creating greater synergy and complementarity with sector support. Burkina Faso: Aligning Donor Procedures with National Systems 83 · Donors should seek to align their aid modalities and requirements, including for monitoring and reporting, with those of national systems. They should also consider supporting capacity building for national institutions (parliament, Audit Office, and especially the Budget and Finance Committee) and supporting civil society organi- zations in the production, publication, and dissemination of reports from independent studies. Lessons for Other Countries · It is difficult to integrate the budget and PRS processes when the key structures responsible for managing the two processes do not report to the same authority. Budget programming and planning tools (MTEF, action plan, and program budget) can be potential integra- tion vectors if they are well mastered and effectively used in decision making, but they need to be harmonized and integrated into both the budget process and the PRS process. The implications of this change at the institutional level and for capacity building must not be underestimated. · Countries should institute an overall budget programming process that covers all resources (including HIPC resources) and integrates poverty reduction concerns. · The PRS cannot be integrated with the budget unless there is an effi- cient budget execution system that allows resources to be allocated to poverty reduction priorities. Predictability of resources is essential for aligning budget execution with programming and constitutes a positive incentive for sector managers. · The search for better integration can affect the format of budget pre- sentation. For example, functional presentation facilitates integration. · Because it stimulates demand for information on performance, annual progress reporting on the PRS can have a beneficial effect on domestic accountability systems and on performance monitoring. This effect, and the ownership of the PRS process, are facilitated when the PRS monitoring mechanism strengthens the existing review and monitoring systems without replacing them. · Greater attention should be paid to the incentives that affect the implementation and alignment of the budget and the PRS. Without suitable and transparent incentives, one cannot expect actors to be results oriented. Development partners can influence these incentive systems by sharing information and reducing the transaction costs of 84 Burkina Faso: Aligning Donor Procedures with National Systems aid delivery for governments. A credible request for information from donors or from government decision makers is in itself a strong incentive for results orientation. · External accountability systems should strengthen internal account- ability systems through aligning themselves with national systems and institutions. 9 Madagascar: Toward More Effective Government Leadership After decades of economic mismanagement followed by a protracted political crisis, Madagascar now has a clear strategic focus in the Madagascar Action Plan (MAP)--the country's new Poverty Reduction Strategy (PRS)--combined with a systematic monitoring and evaluation framework. Key ministries are responsible to the president's office for results-oriented business plans, and the culture within the administration is changing for the better. Several reforms have been introduced to improve the link between the PRS and the budget, but more work is needed to realize their benefits and achieve consistency among the MAP, sectoral strategies and busi- ness plans, and the budget. PRS and budget reporting have been distinct, separately managed processes, and neither the budget reports nor the progress reports on the PRS have been used to influence policy or budget allocation decisions. For both the PRS and the budget, the main accountability relationship tends to run between central government and donors, partly because of the weak technical capacity of parliament and the fragmentation of civil society. Country Context Madagascar is one of Africa's poorest countries, with a GDP per capita of around US$290 in 2004. Three out of four people live below the national poverty line. The effects of low incomes are compounded by very limited public services, particularly in health, education, and transport. The coun- try's roads are so poor that four out of five people have no reliable transport services and one in three have no road access at all. For some 30 years, Madagascar pursued a socialist development strat- egy aimed at achieving economic self-sufficiency. Then, in the late 1990s, 85 86 Madagascar: Toward More Effective Government Leadership structural reforms sought to create a favorable environment for private investment and begin integration of Madagascar into world markets. Growth picked up substantially, reaching 4.6 percent a year in 1997­2001, but the impact on poverty was limited, and popular dissatisfaction led to the ouster of the government in the general election of December 2001. The incumbent president contested the election result, however, and a seven- month political crisis followed. The economic and social consequences were severe: during 2002, real GDP contracted by 13 percent, food prices rose by 25 percent, thousands of jobs were lost, and the poverty rate rose from 69 percent to 80 percent. Once the political crisis was resolved, the incoming president, Marc Ravalomanana, faced a demanding population, a weak tradition of popular involvement in government affairs, and rampant corruption within an ineffi- cient public administration. At his request, an innovative training program for the entire government leadership team, provided with donor support, helped to develop leaders' individual capacities and enabled them to create a common vision for Madagascar's development. Behavioral and organiza- tional changes within the administration have begun yielding concrete results, most notably in improved rice production and in transport services. The PRS that the new government inherited had emerged in 2001 from a participatory process involving parliament and civil society, but it was seen in civil service circles as largely a means to obtain international debt relief, and lacking in operational focus. The new government updated the Poverty Reduction Strategy Paper (PRSP) in 2003. The successor strategy--the MAP--was prepared in a participatory way, with workshops organized in all 22 regions of the country. It was presented to parliament in late 2006 along with the 2007 budget and was formally launched by a national work- shop in November 2006. Madagascar's fragile macroeconomic situation still limits the govern- ment's freedom of action. Since the political stand-off, GDP growth has averaged more than 7 percent a year, but with the annual population growth rate at 2.9 percent, real GDP per capita is still below the pre-crisis level. The year 2004 was marked by two devastating cyclones, a sharp depreciation of the exchange rate, and high inflation (27 percent). The ratio of government revenue to GDP has stagnated at around 10.5 percent (2004). More than 70 percent of Madagascar's investment budget is financed by aid, from a host of international organizations and donors. Formal steps were not taken to coordinate aid until October 2005, with the signing of a Partnership Agreement with the country's general budget support donors: the European Commission, the World Bank, the African Development Bank, and France. Madagascar: Toward More Effective Government Leadership 87 Linking the PRS and the Budget The MAP, the new PRS, provides a year-by-year roadmap for 2007­12 toward achieving the country's long-term vision of economic and social development, Madagascar Naturellement. It includes eight commitments: (i) responsible governance; (ii) connected infrastructure; (iii) educational transformation; (iv) rural development and a green revolution; (v) health, family planning, and the fight against HIV/AIDS; (vi) high growth econ- omy; (vii) cherished environment; and (viii) national solidarity. Each focus area contains quantifiable goals to be achieved by 2012; the government is in the process of disaggregating these goals into annual targets for the public administration. When the current government came to power, neither the PRS nor the majority of sectoral strategies were adequately reflected in the annual bud- get, which did not take a medium-term perspective. Inconsistencies among the budget, the PRS, the various policy documents, and sectoral strategies were worsened by the lack of results orientation and/or performance moni- toring in the administration. To improve results orientation and monitor the implementation of policies and programs, the government in 2003 introduced a matrix--the Politique Générale de l'Etat (PGE)--that defines annual business plans for all ministries and departments. Under the PGE, each ministry defines three to five sought-for results to be featured in a results agreement between the president and his cabinet members. Implementation of the business plans is tracked on a regular basis, and progress and bottlenecks are discussed at the cabinet level. As a next step, the government intends to integrate the differ- ent activities into a coherent results-based management framework. In 2005, the government introduced several important reforms designed to better link the PRS with the budget, improve the transparency of bud- getary allocations, and further strengthen results orientation: a medium-term expenditure framework (MTEF); a program budget structure based on the three main axes of the PRS; and the alignment of annual progress reporting on the PRS with the government's fiscal year. The link between the programs and activities of the PRS and the budget process is not yet well established, however, for several reasons. The MTEF used for budgeting is disconnected from Madagascar's macro framework and may not be sustainable. The budget process suffers from a lack of timely preparation and coordination between line ministries and the Ministry of Finance (MoF), which has the sole responsibility for budget preparation. The line ministries' weak institutional and technical capacities, the absence of budget negotiation within the cabinet, and the absence of a budgetary 88 Madagascar: Toward More Effective Government Leadership conference limit the involvement of other ministries in budget decision making. Responsibility for developing the PRS is separate from responsibility for the budget; at present the monitoring of the MAP is ensured by the prime minister's office in collaboration with the presidency. Achieving consistency among the MAP, sectoral strategies and business plans, and the budget remains a high priority area for the government. The budget for 2007 reflects the MAP priorities for the first time, and efforts are in progress to better align the business plans with program budgets and with the MAP. Budget credibility poses another challenge in Madagascar. As in many other low-income countries, donors play the dominant role in determining the amount of budget resources available, and inflows of general budget support may fluctuate widely from year to year, making realistic budgeting difficult. The last Public Expenditure and Financial Accountability Assess- ment (2006) reports a 15 percent deviation between Madagascar's approved and executed budget in 2003. In 2005, expenditures had to be reduced during the budget year without reference to the priorities of the PRS, and the same problem reappeared in 2006. Integrating Reporting on the PRS and the Budget Budget and PRS reporting have been distinct, separately managed processes in Madagascar. Linking physical achievements and output indicators to financial execution data is not possible at this stage. Neither the budget reports nor the progress reports on the PRS are used for policy orientation or budget allocation decisions. The audiences for budget and PRS reporting differ. Budget implementa- tion reports go to general budget support donors and to the pilot ministries as well as to the MoF. Annual progress reports (APRs) are made publicly available on government Web sites and distributed widely within the administration and to parliament, but they are used mainly by the donor community; the performance assessment framework matrix used by donors of budget support is aligned with the actions and indicators spelled out in the PRS and reported in the APRs. Budget monitoring and reporting is the responsibility of two entities: the Treasury, which produces regular aggregate balance statements geared to the IMF, and the ad hoc Comité de Suivi des Finances Publiques (CSFP), which was established to monitor budget execution by six pilot ministries. CSFP comprises representatives from the MoF, pilot line ministries, and the general Madagascar: Toward More Effective Government Leadership 89 budget support donors. It meets every four months and reviews two types of reports: · Program budget monitoring reports from each of the six pilot ministries (education, health, agriculture, justice, transport, and environment). Most of these reports rely on manual compilation of national and local government data. Their form and quality differ among the ministries, making aggregation very difficult. · Four monthly aggregate reports from the MoF, which track financial budget execution by line ministries (but not program budgets). The PRS reporting system dates from the introduction of the first PRS in 2001. It is only partly consistent with the monitoring framework in the sectors, which use the PGE matrixes of actions and achievements. A new monitoring and reporting system for the MAP is presently being developed. The best aspect of the PRS reporting system is that, today, several forums and participatory mechanisms are reporting on public program implementation. This is a major achievement, given Madagascar's political history. Monitoring of PRS implementation is the responsibility of the Secrétariat Technique de l'Ajustement (STA), advised by a monitoring group, the Cellule Technique de Suivi, both attached to the MoF,1 with STA acting as the technical secretariat of the group. The Cellule Technique de Suivi has about 30 members that are largely from within government (implementing line ministries, the national statistical institute, and the presidency), though it also includes some parliamentarians and members of civil society and the private sector. The Cellule Technique de Suivi uses four instruments to monitor the PRS and prepare annual progress reports on it: · Regular household surveys to provide representative statistics on core output and outcome indicators. · Reports from thematic working groups of public sector staff that are organized by STA according to the broad axes of the PRS. · Periodic regional workshops, also organized by STA, that bring together representatives of donors, national and local governments, and regional civil society. The workshops collect information on regional activities as well as feedback on PRS program implementation. · Reports on macroeconomic developments from the MoF. Two APRs on the PRS have been issued: one in mid-2004 covering July 2003 to June 2004, and one in 2005 covering January to December 2004. The APRs are quite long and complex documents written in French. They present 90 Madagascar: Toward More Effective Government Leadership general budget information as well as a few core indicators linked to the budget process. They do not analyze the pilot ministries' budget reports, and they do not compare planned with realized activities or analyze the reasons for the reported developments in core indicators. Ownership and Accountability The MAP was collaboratively developed and is broadly owned. As part of the leadership and management program, the president and prime minister chaired a series of cabinet-level retreats, each of which was followed by retreats for representatives from the private sector, civil society, and acade- mia, to help define the overall vision and development strategy for Mada- gascar and to identify means of getting results, solving problems, and learning across ministries. The MAP's broadly based vision is understood and shared at both the political and technical levels of government, and the MAP is generally referred to in government writings and speeches as a key reference document. Up to now, however, the main accountability relationship for both PRS implementation and budget execution has been that between central govern- ment and the donor community. For the PRS, the limited role of parliament and the fragmentation of civil society have hindered the emergence of domestic accountability mechanisms. The main responsibility for PRS reporting rests with the STA, which has no line authority over PRS implementation. The APRs on the PRS are not pub- licly reviewed or debated; indeed, because they do not link information on PRS monitoring to information on budget execution they are not very useful as a tool for overseeing the use of public funds. Though the Cellule Technique de Suivi of the PRS has some parliamentarians among its members, the weak technical capacity of parliamentarians in general limits their demand to fol- low the PRS or budget implementation. And a tradition of limited overall oversight of the executive branch by the legislature may limit parliament's access to the information it would need to monitor government performance closely. Civil society groups are represented in the Cellule Technique de Suivi and regional working groups, but the fragmentation of civil society restricts its role in PRS monitoring. Domestic accountability for budget execution and results has been limited. Within the government, a challenge function has been lacking. Responsibili- ties and accountabilities regarding budget execution and the achievement of results were not clearly established until 2003, and as noted above, the busi- ness plans and results agreements under the PGE are not yet integrated with the budget process. Furthermore, the line ministries' statistical Madagascar: Toward More Effective Government Leadership 91 information systems have been relatively weak; without an integrated financial management system, these ministries and the MoF often have discrepancies in budget execution data, which dilute the accountability of line ministries for implementing their programs. Government has also faced little pressure from parliament and the public for accountability on the budget. Parliament's role in the budget process has been limited to adopting the final budget without many questions on the content (the political opposition is in the minority). Madagascar's Loi de Règlement, which contains the complete budget execution information and is presented to the Court of Auditors and to parliament, has been completed only after long delays, making timely external control difficult. However, the government is now making important efforts to reduce the delay. Civil soci- ety is not yet a player in the budget process. General public access to budget reporting is limited. Conclusions and Lessons · Government has introduced several important reforms designed to strengthen the link between strategy and budgeting, but more needs to be done to realize their full benefits. A program budget structure has been adopted and the budget for 2007 reflects for the first time the objectives and priorities of the MAP. A limiting factor, though, is that the MAP objectives have not yet been fully costed. The MTEF now used for budgeting is disconnected from the macro framework and may not be sustainable. The MoF has sole responsi- bility for budget preparation, and the involvement of other min- istries in budget decision making is limited. Annual business plans for each ministry are now embodied in results agreements with the president's office, but these agreements are issued after the annual budget is issued, so these do not influence the budget. Thus strategy is still only partially aligned with decision making on resource allo- cation or on budget implementation. · The monitoring and reporting mechanisms of the budget and of the PRS are not integrated and their audiences differ. Neither the budget reports nor the progress reports on the PRS are used for policy orien- tation or budget allocation decisions. · The main accountability relationship for both PRS implementation and budget execution has been that between central government and the donor community. Parliament and civil society play only a marginal role. 92 Madagascar: Toward More Effective Government Leadership · All this said, positive change is clearly afoot within the Malagasy administration, and for the first time, a clear strategic focus is com- bined with a systematic monitoring and evaluation framework. The changes thus far bode well for achieving efficiency and accountability in the implementation of the PRS. Suggestions for Improvement in Madagascar · Sector strategies should be fully aligned with the MAP; a results- oriented monitoring framework for the MAP should be developed that outlines annual measurable targets for every MAP objective and commitment. · Key performance monitoring indicators need to be identified that are common to the MAP, sector strategies, and program budgets. · Both parliament and civil society organizations need to be strength- ened and provided with tools so that they can fulfill their role as major stakeholders and actors in the social accountability process. Donors and government should pursue their efforts to strengthen the civil society, and broaden them to include parliament more actively. · Regular sector-by-sector reviews of progress on the MAP should be organized, with the participation of civil society, donors, parliament, and government. Parliamentary sector committees and their members should become constant active participants in these reviews, and the executive should provide them with the necessary information in a timely manner. · With donors of general budget support, it will be important to estab- lish a common framework of disbursement conditions that is closely linked with the MAP and sector strategies. The discussions for this purpose should include civil society stakeholders and parliament. In the medium run, an annual results-oriented policy document provided by the government could serve as the basis for general budget support disbursement conditions. Lessons for Other Countries The integration of a country's PRS and the budget can be improved by · developing sector strategies and results agreements within the framework of the national strategy; · adopting coherent budgeting instruments as well as a clear process for political arbitrage of the budget; Madagascar: Toward More Effective Government Leadership 93 · improving budget formulation at the sector level by developing results-based sector strategies that are costed and prioritized; · making the entities in charge of the PRS and budget accountable to the same government authority; · developing in public institutions a management culture oriented to achieving results, as well as enhancing their technical capacity to perform; · establishing strong coordination among donors, along with harmo- nized aid procedures, both to facilitate the coherence of the budget and to encourage public sector reporting; and · clearly defining the responsibilities of line ministries within the PRS and budget process, to promote reporting and accountability. 10 Malawi: Building a Credible Budget for Better Accountability A change in government in 2004 improved Malawi's prospects for evidence-based decision making after years of budget mismanagement and poor tracking of expen- ditures. The new government has worked to create a new Poverty Reduction Strategy (PRS) that reflects its own priorities, and is building the foundations for solid management of public finances. The link between the PRS and the budget should be considered one of Malawi's priorities for reform. Thus far, this link has been very weak, partly because the link between the budget itself and its implementation has been weak and partly because decision making on the PRS has been too far removed from decision making on expenditures during the budget year. Given Malawi's daunting capacity constraints, the fragmentation of the policy cycle and the insti- tutional responsibility for budgeting and the PRS, and the informality that has long pervaded the operations of the public sector, technical issues are not the most important obstacle to reform. Strong administrative and political champions that are willing and able to ensure sustainable and politically feasible change over the medium term need to be identified and supported.1 Country Context Malawi is one of Africa's most densely populated and one of the world's least developed countries, with a per capita income in 2004 of US$170, and more than half of its people living in poverty. The country is severely affected by the HIV/AIDS epidemic. Per capita income has been falling since 1998, and aid provides more than one-fourth of gross national income. Persistently high fiscal deficits pushed up Malawi's public debt to 95 96 Malawi: Building a Credible Budget for Better Accountability 208.6 percent of GDP in 2005, and high interest rates and inflation have discouraged investment. Malawi is still an emerging democracy. The most recent change in govern- ment, in 2004, was only the second democratic transition since the 30-year reign of President Banda. The political landscape is clearly dominated by the executive; the party system is still evolving and the precise role of parliament is not settled. Informal networks and traditional paternalistic structures still underlie the formal structures that have been set up since the establishment of a multi-party system. Thus, it would be unrealistic to expect a proper system of checks and balances to be at work. Instead, there is a mixture of traditional and emerging democratic patterns of behavior that over time could evolve into a more established democratic polity. Malawi is just emerging from a period of very weak management of public finances and little capacity or willingness of government actors to reform. Although the legal framework for planning and budgeting has been fairly sound all along, a substantial gap still exists between formal rules and norms and widespread informal practices. Capacity constraints are endemic throughout the public sector. Another pervasive feature of Malawi's public administration is that central govern- ment cohesion is very weak, resulting in fragmented ownership of instruments, processes, and reforms. At the administrative level, diverse administrative actors presume to speak for the whole government, yet see their authority limited by overlapping claims. The recent imposition of fiscal discipline shows that such problems are not insuperable, but their existence is of paramount importance. Linking the PRS and the Budget Responsibility for strategic planning and budgeting in Malawi is fragmented: · The Ministry of Economic Planning and Development (MEPD) is in charge of formulating the government's long-term strategies and is the source of all three recent strategic planning documents: the first Poverty Reduction Strategy Paper (PRSP)--the Malawi Poverty Reduc- tion Strategy (MPRS), for 2002­05; the Malawi Economic Growth Strategy (MEGS), which the new administration adopted in 2004 to complement the MPRS; and the upcoming second Poverty Reduction Strategy, the Malawi Growth and Development Strategy (MGDS), which is scheduled to become effective in fiscal year 2006­07. Malawi: Building a Credible Budget for Better Accountability 97 · Overall budget preparation, both on the recurrent and the develop- ment side, is the responsibility of the Ministry of Finance (MoF). · Many sector ministries use separate planning instruments; they negotiate their projects directly with donors and do not necessarily report them to central ministries, making it difficult for the center to track or control their spending. Planning in sector ministries does not follow a common standard, despite strong government and donor efforts to improve harmonization among donors and align donor activities with central government priorities. · Malawi began to outline the legal framework for decentralization, under which substantial responsibilities and resources would be given directly to the districts. Among other new responsibilities, the district administration is now directly in charge of primary education and health care.2 However, most provisions of the newly decentralized sys- tem have not yet been implemented. Local spending units depend almost entirely on resources from the center, and are accustomed to spending their small allocations on the most immediate needs, quite independently of the line items in their budgets. To harmonize local planning and budgeting with the national PRS seems a particular challenge, given that familiarity with the strategy is extremely limited among officials at the local level. In the past decade there was virtually no link between budget planning and budget implementation. Neither at the macro or micro level did the budget document provide a good forecast of actual expenditures through- out the year. Weak central control, weak technical systems, and pervasive informality all served as causes. The important, if partial, exception to this pattern has been protected pro-poor expenditures (PPEs), which have been tracked separately from the rest of the budget. However, the reporting only covers funding toward PPE activities; actual expenditures cannot be tracked. In 2005, PPEs accounted for about 23 percent of total government expenditures. They include fund- ing toward the broad priority areas laid out in the Malawi Poverty Reduc- tion Strategy Paper (MPRSP). The impact of the PPE tracking has been limited, because the activities included were changed from year to year, making comparisons difficult. Overall funding for PPE-related expendi- tures remained stable or increased broadly in line with planned allocations through the years. PPEs were an important priority for the government because the requirements for both the heavily indebted poor countries 98 Malawi: Building a Credible Budget for Better Accountability (HIPC) initiative and the Poverty Reduction Strategy Paper (PRSP) process emphasize that poverty-oriented expenditures should be protected. The change in Malawi's political leadership and direction has begun to change the incentives around the budget. Fiscal discipline has been restored for the first time in about a decade. In general, sector allocations now remain within their ceilings, and previous informal strategies to acquire additional funding during the year seem to have ceased to be successful. In the second budget cycle of the present government, actors in the public sector are aware that their spending wishes need to be included in the annual budget to be realized. The new salience of the budget thus creates a new incentive to integrate planning processes into the budget cycle. But the country's fragmented policy cycle, with divided institutional responsibilities and no clear sequential relationship from planning to execution to reporting, is an underlying factor that strongly constrains efforts at integration. A previous attempt to integrate the PRS with budgeting failed. When the MPRS was drafted in 2002, the intention was to facilitate links from the strategy to the budget, and the Ministry of Planning took important technical steps to make this possible. But several factors essentially disabled the integration process. First, the budget division of the MoF has such severe capacity constraints that it could not have constructively discussed the policy choices of sector ministries or related them to the implementation of the MPRS. Second, the previous government frequently made ad hoc spending decisions that sharply limited the relevance of budget planning for budget implementation. Third, sector ministries may not have known what would be required to comply on a technical level with the plans outlined in the MPRS. All in all, the MPRS was too far removed from expenditure deci- sion making during the budget year to play the central role that donors and the MEPD had envisaged for it. Decision makers now face the same need to integrate planning with budgeting in the context of the Malawi Growth and Development Strategy (MGDS). The MEPD has compiled an expenditure plan to accompany the MGDS, but the costed activities are not coded to be compatible with the classification system of the budget, and officials at the MoF noted that they do not yet know how they could identify activities in the expenditure matrix and translate them into the format of the budget. Unlike the MPRS, which was widely criticized for having been donor driven and largely donor owned, the MGDS enjoys the full ownership of the government and is backed by the cabinet. Both the Minister of Planning and the MoF have publicly endorsed it. Yet at the technical level the ownership of the strategy Malawi: Building a Credible Budget for Better Accountability 99 outside MEPD is still in doubt, even though the sector ministries provided most of the technical inputs for the new strategy. To resolve the numerous remaining questions about how the MGDS should be integrated into the normal budget cycle will require a broader sense of ownership of the strategy across the government. Renewed efforts are under way to tackle technical difficulties that stand in the way of linking budget planning and implementation. A new elec- tronic financial management system is being introduced across the central government. If successfully implemented, this will become a powerful tool to better control the implementation of the budget, and also potentially to assist in linking planning with budgeting. The accountant-general's depart- ment has also taken drastic steps to improve cash management and thereby the control of expenditures. It has been very challenging politically to enforce budget discipline over the last 18 months, and the government deserves credit for this achievement. To sustain the progress, it is now necessary to improve bud- get planning and the micro-level implementation of the budget. Officials in central ministries are aware of this challenge and seem committed to start working on the next steps. Integrating Reporting on the PRS and the Budget Malawi has three separate reporting and monitoring processes: · Budget reporting by the MoF. · PRS monitoring (through annual progress reports [APRs] and a com- prehensive review) by the MEPD. · Poverty monitoring (based on multi-annual surveys) by the National Statistical Office. In addition, many sector ministries have direct relationships with donors and use their own reporting mechanisms, as does the Ministry of Health under the sector-wide approach for health. During the first half of this decade, all three of the main reporting systems were generally weak. The main reasons were shortages of technical, human resource, and institutional capacity, combined with low demand for reliable information within the public sector. Budget reporting is likely to improve strongly, now that the accountant- general's department has begun to implement the new electronic financial management system. According to several reports, the president and the MoF demand quantitative information about government activities, and monthly 100 Malawi: Building a Credible Budget for Better Accountability reports are discussed in cabinet. While this by itself does not transform the state of affairs, and any changes depend on the sustained commitment of political leaders, there is now a much better chance for reform initiatives to succeed in improving reporting systems. All the PRS monitoring reports have suffered from the insufficiencies of Malawi's reporting and monitoring systems, and they contain little analysis that would systematically link results information to government activities or resource flows. The MoF does not use these reports during budget prepa- ration or budget planning. Government officials see the APRs mainly as a requirement for donor funding, and donors and government officials agree that the only recipients who strongly demand delivery of these reports are in the donor community. Malawi's monitoring and evaluation (M&E) arrangements lacked the necessary foundations for robust reporting during the lifetime of the MPRS. And, apart from the MEPD, which has championed an ambitious Master Plan for Monitoring and Evaluation without managing to overcome the technical and institutional obstacles to its implementation, no other domes- tic stakeholders have strongly backed a government-wide M&E system. The government and donors have stressed their commitment to implementing the M&E Master Plan. If it can be implemented, the plan could substantially improve the availability and quality of data on the outputs, outcomes, and impacts of government policies. No performance reporting instrument plays an important role in bud- geting at present. The APRs on the PRS did not play a significant role during budget preparation by sector ministries, budget negotiations with MoF, or deliberations in parliament. The only notable link between the APRs and the reporting on the budget has been the tracking of PPEs. Funding for these expenditures is tracked separately and PPE reporting data are discussed by domestic actors such as the Malawi Economic Justice Network. The poor link between the PRS and the budget should be considered one of Malawi's priorities for reform. Ownership and Accountability Before the present government took office, the space for evidence-based decision making in Malawi was very small. First, the political leadership made its decisions based on short-term political considerations, and because all levels of government were to some extent involved in mismanaging funds, few individuals were interested in upsetting such practices. Second, Malawi: Building a Credible Budget for Better Accountability 101 the absence of a challenge function at the central ministries removed much of the opportunity to use reporting information at a technical level at the center of the executive. Third, because external formal institutions, such as the accountant-general and parliament, had very limited capacity to absorb and use reporting information, domestic forces could not easily push the executive to change its ways. Domestic accountability remains weak in Malawi because actors that could challenge the executive lack the capacity to do so effectively. If infor- mation systems provide better information, it is more likely to be used by both the political leadership and its critics. This would be an important step toward stronger accountability relationships not only between the executive and external actors but also within the executive itself. The accountant-general has a key role to play in improving the account- ability of the executive. In principle, Malawi's National Audit Office (NAO) could be the watchdog that ensures the credibility of the budget and holds spending units accountable if they mismanage funds. Unfortunately, NAO is underfinanced and understaffed, and the evidence suggests that the executive does not take its activities very seriously. Parliament is among the weaker actors in the system. It does not meet in regular sessions, has no permanent administration to support legislative business, and does not even have its own building. Relations between the executive and parliament have been rocky since the present government came to office. Up to now, parliament has played little role in the PRS; members were consulted in three regional meetings after the first draft of the MGDS was ready, but were not involved in the drafting prior to these meetings, and the government does not plan to seek a formal vote on the strategy in parliament. Neither is the budget itself debated in parliament; because parliamentary committee members rarely have the necessary pro- fessional background, or time or resources, to analyze the details of the budget, there is only a limited debate on specific budgetary issues. The Malawi Economic Justice Network has assumed a role in promot- ing civil society oversight of public expenditure, particularly at the local level. Apart from its advocacy work on budgetary issues and the PRSP, the network prepares summaries of the budget intended for its civil society affiliates and local governments. There is clearly great scope for increasing budgetary literacy among many actors in public finance and the network is doing valuable work to improve it. Because the capacity of the executive is weak, and that of other stake- holders even weaker, donors face a difficult challenge: they need to help strengthen the government's capacity to set policies and implement them, 102 Malawi: Building a Credible Budget for Better Accountability but at the same time push the government to be more accountable toward other domestic actors--most importantly the legislature. This cannot be done without also strengthening actors outside the executive, and encour- aging more responsible behavior. Conclusions and Lessons The strong effort by the MoF to control aggregate spending levels, if sus- tained, will strengthen the incentives to improve budgeting and formal reporting mechanisms. Evidence from many successful budget reformers shows that fiscal constraints are often the initial driving force behind reforms. Successful reforms to control expenditures can create the momentum for more sophisticated changes to follow, including a more results-oriented budget process. Suggestions for Improvement in Malawi · The duplication of functions among different actors should be reviewed and reduced wherever possible. The government should strongly consider reviewing the processes and structures of the MoF and the MEPD, with a view to unifying the capacity to formulate and imple- ment expenditure policies in strong, well-placed administrative units. · The budget department, in particular, must play a leading role in integrating the PRS and budgeting. Strengthening its capacity should be a top priority for government and donors. · A formal budget calendar should be adopted in law, ideally as an amendment to the Public Finance Management Act. This calendar should properly integrate strategic and investment planning and donor funding cycles. This legislative change will not by itself ensure improved budget management, but it could serve as an important foundation of further strengthening. Once passed into law, it would be a mandatory norm that all actors, including the MoF, could be mea- sured against. · Donors should support the implementation of the MGDS by making it the basis for their interaction with sector ministries. Sector plans should reflect the priorities spelled out in the strategy and donors should enhance their demand for information about results at the sector level. · The government could enhance the role of parliament in the PRSP process, for instance by delivering progress reports (such as the APRs) Malawi: Building a Credible Budget for Better Accountability 103 to parliament. A permanent administration of parliament would have to be created in the medium term to support informed budgetary deci- sion making by the legislature. · The NAO should be strengthened so it can take up its assigned role as a key actor in a system of domestic accountability. Lessons for Other Countries · Success in linking the PRS to the budget depends not only on techni- cally sound instruments but even more so on the right institutional setup. Even where formal instruments are sound, if ministries lack the capacity or incentives to use them, even the best strategy will likely fail to be implemented. · The budget department must play an important role in the PRS process. To exclude it means creating a firm obstacle against success- ful implementation of the PRS. · A medium-term expenditure framework (MTEF) is not by itself a solution for linking the PRS to the budget cycle; the budgetary process must itself become oriented to the medium term to support an MTEF. In many countries, this may not be the most important reform priority for making the budget a credible policy instrument. · Without credible budget execution, reporting on a PRS cannot work properly, because it would have to rely on external systems to function, thus diverting attention and resources from more fundamental reforms. · Political will from the top leadership is necessary but not sufficient to create demand for information from government actors. Demand can only be created over the long run, by changing the formal and infor- mal incentives at the operational level of public administration. · Strengthening domestic accountability is an inherently political and long-term process that involves changing the political economy and incentives of public expenditure management. For a government to be accountable to its citizens, functioning internal accountability inside government is a crucial prerequisite. Donors and governments should be aware of this and not encourage the impression that there are quick fixes. · Donors can use their leverage with government actors to encourage better results reporting, but their activities must be sensitive to the ability of domestic actors to use such information. For instance, if sector-wide approaches are in use, their monitoring requirements should be aligned with the national strategy. 11 Mali: Linking Policies and Medium-Term Budgeting Mali's experience emphasizes that developing domestic accountability around the budget and the Poverty Reduction Strategy Paper (PRSP) is a complex process involving a range of actors, and needs time and effort even in a broadly propitious-- i.e., democratic--environment. The links between Mali's budget and Poverty Reduction Strategy (PRS) are still weak, and PRS and budget monitoring have evolved separately both institutionally and with respect to planning and to reporting on implementation. Certain initiatives and entry points point to a potential for the development of greater integration over time. The most important is ongoing work on medium-term planning, which could eventually translate the PRS into the annual budget. Ownership and demands for accountability are growing in parliament and within civil society. Donors need to adjust their behavior and internal incentive structures to be less concerned about their own particularistic priorities, and more about promoting effective domestic structures and account- ability relationships.1 Country Context More than two-thirds of Mali's 12 million population live below the national poverty line, and the overall poverty rate has remained roughly constant over the last decade. The economy depends heavily on agriculture, and since 2000, drought and locust infestation have caused some bad harvests, food shortages, and volatile overall growth rates. Falling prices for cotton, one of Mali's major exports; rising fuel prices; and troubles in Côte d'lvoire have posed additional 105 106 Mali: Linking Policies and Medium-Term Budgeting challenges. The situation improved in 2005, however, and by the end of that year the IMF found that Mali's macroeconomic situation was satisfactory. Reforms in public financial management have been ongoing for several years, and have resulted in considerable progress. Tax revenue rose from 14 percent of GDP in 2000 to 18 percent in 2004. A new plan for further comprehensive reforms of public financial management was adopted in April 2005, but its implementation has been slow thus far, partly because of funding constraints. Medium-term expenditure plans have been devel- oped in the priority sectors of the PRS, with donors' assistance; they are now being extended to further sectors and a general/top-down medium- term expenditure framework (MTEF) is being developed. Mali is on a trajectory toward institutionalizing democracy, and is one of only eight continental African countries rated as "politically free" by Freedom House, but substantial risks to this trajectory remain. An important element of the democratization process has been the efforts made to decentralize government from a highly centralized starting point. Political decentralization has outpaced fiscal decentralization, and the control of local-level institutions such as primary schools and health centers and their staff is still centralized, but fiscal decentralization is slated to proceed over the coming years. Mali's aid receipts per capita varied considerably between 1990 and 2005 but have been rising in recent years and now amount to about 9 percent of GDP. Like that of many low-income countries, Mali's aid environment is fragmented, and the country has at least 19 bilateral donors. Donor harmo- nization and alignment are at an early stage; they are most advanced in health and education, where sector-wide adjustment approaches and sector budget support are in use. Linking the PRS and the Budget When Mali developed its first PRSP in 2002 it could build on the experience with a previous poverty strategy, the Stratégie Nationale pour la Lutte contre la Pauvreté, which had been adopted in 1998. A second-generation PRSP, to be adopted in late 2006, is now being developed. In Mali the link between the PRS and the budget is still quite weak. Budget formulation is clearly dominated by the Ministry of Finance (MoF), although the National Development Planning Department, line ministries, parliament, and donors participate. In a recent institutional reorganization, Mali sepa- rated the Planning Department from the MoF, for political reasons. The new Mali: Linking Policies and Medium-Term Budgeting 107 structure poses a problem both for the budget and for the PRSP process, because it divides responsibilities and segregates the planning know-how, and because cooperation across ministries is not well established. Mali's plans become more detailed and operational as the planning period shortens, and the annual budget is the most detailed of them. In principle, the PRS covering a five-year period is translated into the annual budget through medium-term planning. In practice, Mali's two instruments for medium-term planning, the budget for programs (budget­programmes, BP) and the recently introduced sectoral MTEFs (cadres de dépense à moyenne terme, CDMTs), do not fulfill this role. The budget for programs is an annex to Mali's annual budget. It provides information on the ministries' objectives, programs, and activities for the next three years and it contains some indicators, but its definition of programs is still marked by conceptual difficulties and it makes only sum- mary links from programs to expenditures. The CDMTs have been introduced through a bottom-up process in selected ministries only, and a top-down process is now being developed. Decision makers at every stage of the budgeting process are aware of the objectives set by Mali's PRS and the strategy's priorities have a certain effect on resource allocation. But because only a small part of any state budget can be changed from one year to the next, a complete alignment of the long- and medium-tem strategy documents and actual resource allocations in the bud- get will take time to achieve. Moreover, Mali's PRSP is so all-encompassing that the goals stated there could justify almost any expenditure. Once the budget is prepared and voted on, PRS priorities have no specific effect on its implementation. Once resources have been assigned to a purpose, they cannot be used for anything else. While from a managerial point of view this may reduce efficiency, it assures control. Unlike in some other countries, there is no special protection of PRS-relevant expenditures. Generally, budget execution has not posed a problem in recent years but in the case of shortfalls, wages and salaries receive first priority in spending. Mali's civil service has no tradition of results-oriented management. The normal legally binding budget is not results oriented and is not designed to be so. Sector ministries have some incentives for medium-term planning and results orientation, having found that using MTEFs enables them to receive more funds from donors in priority sectors (by showing funding gaps) and that successful programs attract more funds, whether from internal or external sources. The budget for programs introduces some elements of results orientation into the budget process, but it would need to be developed further to achieve its full potential. 108 Mali: Linking Policies and Medium-Term Budgeting Integrating Reporting on the PRS and the Budget Reporting and monitoring are areas where the PRS and budget have remained particularly separate. One structure, comprising the finance and administration departments of ministries, the Planning Department, and the National Budget Depart- ment, is responsible for the budget process, which in its control and audit phase also involves the Inspection des Finances and the Section des Comptes. The structure for monitoring the PRS is comparatively weak and consists principally of the planning and statistics units in the priority sector ministries and the office responsible for the Poverty Reduction Strategy Paper (Cellule CSLP), which is located in the MoF and has limited capacity. The monitor- ing of PRS implementation appears to be owned chiefly by donors and the public administration, and less so by government and nongovernmental stakeholders. Civil servants at the technical level express the will to use annual reporting on the strategy to learn and improve programs. The "binding" budget is a traditional input-oriented line-item budget, and as such it is not well suited for monitoring performance. Two progress reports have been produced on the implementation of the PRS, while various efforts to monitor budget execution focus mostly on the formal compliance of expenditures with legal and administrative rules and regulations. Reporting on the implementation of the budget for programs or the CDMTs could eventually provide a bridge from budget implementation to strategy implementation if these instruments are developed further, though Mali's weak statistical capacities and lack of baselines for use in monitoring would pose problems in this regard. Ownership and Accountability Although the PRSP as an instrument is seen as having been imposed by donors, when Mali already had an anti-poverty strategy--the 1998 Stratégie Nationale pour la Lutte contre la Pauvreté (SNLP)--the current version of the PRSP appears to be owned by the government and the public administration, and to some extent by civil society. Government and the administration have the primary ownership of the budget. Domestic accountability focusing on the budget and the PRSP is limited but clearly evolving. Up to now, parliament's official involvement in the accountability process has been restricted to the budget, and here it has not yet challenged the government. However, parliament is to play a formal role in formulating and endorsing the second-generation PRSP, which will cover the period 2007­11. The tools and processes of parliamentary scrutiny are still Mali: Linking Policies and Medium-Term Budgeting 109 weak. Civil society's involvement in monitoring either the PRS or the budget has remained very limited thus far, but there are plans to expand its role in both. The public shows strong interest in poverty reduction but thus far its demands do not take the PRSP as a reference point. Public interest in either budget or PRS monitoring is weak, although more indirect forms of monitor- ing and accountability via national and local elections exist and are exercised. External accountability processes have caused more data to be produced, but during the first-generation PRS, Mali's donors have done little to promote domestic accountability, and may even have damaged it by overriding Mali's own original strategy. Learning is taking place among all stakeholders, but donors still make uncoordinated demands regarding the indicators to be used for monitoring, and tend to prioritize their own specific needs over the need to enable effective domestic accountability. Conclusions and Lessons Although the links between Mali's budget and PRS are still weak, certain ini- tiatives and entry points point to a potential for greater integration. The most important among these is the ongoing work on medium-term planning, which could eventually translate the PRSP into the annual budget. Ownership and demands for accountability are growing in parliament and within civil society. Suggestions for Improvement in Mali An integration of reports on the budget and the PRS could be pursued using the budget for programs or CDMTs, or a combination of both. If the CDMTs fulfill their function and become sufficiently program oriented, they and the annual implementation reports could largely take over the role of the PRS progress reports. To accomplish this would realistically take the lifetime of the 2007­11 PRS, and would require a concerted effort to upgrade the role of program budgeting practices. In the context of the fiscal decentralization that is in progress, it is impor- tant to strengthen the involvement of local government in implementing, reporting on, and monitoring the PRS. Formulating regional plans for imple- menting the PRS could help to deepen and widen accountability. Regional plans tend to be less abstract than a national strategy and hence can offer better entry points for locally organized groups and their immediate concerns. At the regional and local levels, results of development efforts can be directly observed by the people affected, who can help to hold those responsible to account. It is on this level too that the population can develop a sense of owner- ship for the PRS. 110 Mali: Linking Policies and Medium-Term Budgeting Reuniting the Planning Department and the MoF would give the office responsible for developing the PRSP (Cellule CSLP) access to the technical planning capacities of Mali's administration. Concentrating the competence for PRSP and the budget could lead to better coordination between the two. If this is not viable, given the political dynamics that caused the separation of the PRSP and budget, an alternative might be to strengthen their institu- tional links and the incentives of planning and budgeting agencies to collab- orate through the Council of Ministers--for example, by establishing regular high-level consultations among the MoF, the Department of Planning, and other key ministries, to focus on PRSP-budget links. Actions would also be beneficial in the following areas: · Further efforts to establish an independent external audit body and to develop its capacity so that it can undertake effective performance audits. · Making support available to parliament and civil society groups as these seek to expand their technical capacity to monitor bud- gets and PRS implementation (but carefully avoiding dominating the process). · Supporting the production of public information suitable for radio broadcast. More than half of Mali's adult population has had no schooling, and, particularly in rural areas, public information needs to take a very different form, and be differently distributed, from that demanded by donors or even urban civil society groups. · Support learning with regard to well-structured participation and con- sultation processes. The rather unstructured participation processes for monitoring the first PRS led to frustration, and there appears to be agreement among actors--including civil society groups--that these processes need to be better structured to ensure both that stakeholders have a voice and that work can proceed effectively. Donors need to become more constructive in their relationship with Mali. They need to become less concerned about their own particularistic priorities, and more concerned about promoting effective domestic struc- tures and accountability relationships. They should continue to use a mix of instruments including general budget support on the one hand and technical assistance on the other hand, while increasingly aligning and harmonizing their activities. They need to focus on enabling change over the long term, and to avoid micromanaging processes, imposing demands that conflict with those of other donors, and prioritizing their own needs over those of Mali and its administration. Mali: Linking Policies and Medium-Term Budgeting 111 Lessons for Other Countries Mali's experience suggests that · Developing domestic accountability around the budget and the PRS is a complex process involving a range of actors, and it needs time and effort even in a broadly propitious--i.e., democratic--environment. Participation is only one element, and requires careful structuring (and resources) to be effective and sustainable. Special attention should be paid to intragovernmental accountability, especially inside the adminis- tration, which has been the most effective accountability relationship in the Malian case. · Mali's medium-term planning processes constitute important links between the PRS and the annual budget. But this relationship does not mean that a fully fledged, state-of-the-art MTEF is needed. The important point is that the MoF and the sector ministries should engage in a dialogue on resources, policies, and policy priorities. · The institutional setup for budget and PRS planning and reporting is crucial for bringing the two areas of reporting closer together. Whether institutions compete or cooperate depends not only on the overall institutional framework of government (especially cabinet/ "center of government" structures) but on the incentives and con- straints that affect the monitoring and reporting processes. · Developing a budget system oriented to programs and their perfor- mance, and making it work across government, is a challenging process. How to do this in poor countries that have strong resource and capacity constraints requires greater attention from the develop- ment community. Because the simpler traditional budgeting systems that are in place work, at least partly, there may be advantages in designing PRSPs so that they are easy to translate into the budget. · Donors can encourage results orientation by using potential addi- tional funding as an incentive. However, this has a cost in terms of reducing the links of the national monitoring systems to the domestic budget and weakening the associated accountability relationships. · Long-term assistance is needed to support the introduction of new instruments and institutional reforms, because changes in these areas are highly political, require context-sensitive understanding, and take considerable time. Because successful and sustainable reforms depend on government having a strong management and implemen- tation capacity, coordinated forms of support for capacity development and the implementation of reform plans at the center of government are important. 12 Mozambique: Promoting a Unified Process Mozambique is a highly aid-dependent country that has upgraded its public finan- cial management systems through a phased reform program, attempting to link resources with objectives and gradually addressing some of the imbalances created by aid fragmentation and donor intervention. Though it has made much progress in planning and budgeting, Mozambique still suffers from some serious distortions in domestic accountability because of its heavy reliance on external assistance, the fragmentation of instruments used by donors, and the extent of donor involvement in central policy processes, including the Poverty Reduction Strategy (PRS) and the budget. Generally, the improvements of the past five years are very encouraging, although their sustainability and consolidation will crucially depend on political stability and continued donor support. Country Context Since its civil war ended in 1992, Mozambique has been one of the world's fastest-growing economies, albeit from a very low base. Agriculture, trans- port, manufacturing, and tourism have recovered from the devastation of war. Real GDP grew by about 8 percent between 1998 and 2004, thanks to significant political and macroeconomic stability and market-oriented reforms. Large inflows of foreign aid have also played a key role, as has investment in a number of mega-projects including in aluminum and gas production. As a result, poverty in Mozambique has decreased sharply, from nearly 70 percent of the population in 1997 to 54 percent in 2003. 113 114 Mozambique: Promoting a Unified Process Mozambique became a showcase of the growth and poverty reduction benefits of programs inspired by the "Washington Consensus," with the implementation of a series of market-based reforms under the guidance of the World Bank and International Monetary Fund (IMF). It was also one of the first countries to benefit from debt relief under the heavily indebted poor countries (HIPC) initiative, supported by a PRS (Plano de Acção para a Redução da Pobreza Absoluta, PARPA) approved in 2001. The government's medium-term strategy, reflected in the first PARPA, is based on maintaining macroeconomic stability, encouraging the private sector, promoting investment, rehabilitating infrastructure, and developing human capital. The second PARPA, recently finalized, covers the period 2006­09 and was formulated with the help of a series of sectoral working groups with participation from government, donors, and civil society. It puts more emphasis on economic growth and its importance for poverty reduction than the first PARPA. Since the late 1990s, Mozambique has undertaken a series of reforms to help public agencies deliver services and promote growth and develop- ment. These include the strengthening of planning and budgeting systems through the introduction of a medium-term fiscal framework (Cenário Fiscal de Médio Prazo, CFMP) and an integrated financial management information system (Sistema da Administração Financeira do Estado, SISTAFE). Mozambique's good performance as a reformer has attracted signifi- cant goodwill from the international community. Aid dependence, which had peaked at 87 percent of gross national income (GNI) right after the end of the civil war, had fallen to below 30 percent of GNI by the late 1990s, and is now about 15 percent of GDP. Donor support still supplies more than half of total public spending and most of public investment. Of the official aid to Mozambique (US$1.3 billion in 2004), almost one-fifth takes the form of general budget support directly channeled through the treasury, with the rest distributed among sector programs and fragmented and often unreported projects. Mozambique's aid landscape is remarkable for the sophisticated architec- ture of an agreement through which 18 donors provide budget support based on a memorandum of understanding enshrining reciprocal commit- ments, which in turn are based on biannual joint reviews and reciprocal performance assessments. Government performance is monitored through a performance assessment framework matrix that forms the basis for policy dialogue.Asimilar matrix exists to monitor donor behavior and commitments, for example on harmonization and predictability. Mozambique: Promoting a Unified Process 115 Linking the PRS and the Budget Figure 12.1 summarizes the main planning and budgeting instruments that make up the processo único, a unified process that the government has been trying to implement consistently over the past few years. The overarching long-term planning instrument is the five-year plan (Programa Quinquenal do Governo, PQG), which is approved by parliament after the beginning of each legislature, and outlines government priorities for the five-year term.1 Five-year plans in their present form tend to be very long lists of undertak- ings, without any clear program prioritization and without links to resource frameworks for their implementation. Lower-level medium-term planning instruments provide the link between these plans and annual operational plans. PARPA has been used to provide better focus to the five-year plan, coming to a better definition of priority areas and programs, as well as providing more analysis and substance to the overall objective of reducing poverty, which features prominently in the five-year plan. Apart from the PARPA, a number of sectoral, provincial, and district strategic planning exercises take place, but these tend to have only weak links to the central processo único. Most of them are designed either to promote intersectoral coordination at the provincial or district government level or to elicit programmatic support from donors for sector-wide approaches (SWAps) in specific sectors. The instrument that should in principle link the five-year plan and the PARPA to Mozambique's annual planning and budgeting process is the CFMP. This framework is used to forecast the overall resource envelope, and to allocate spending among sectors for the medium term according to the policy priorities set out in the five-year plan and the PARPA. In practice, however, this link is not robust. Several factors undermine the significance of the CFMP exercise, including the fact that the technical information on which the CFMP is based does not allow for significant policy trade-offs to be assessed, and that a large amount of resources received from donors at the sector level remain off budget.2 Until recently the CFMP was prepared inside the Ministry of Finance (MoF). But the latest CFMP has benefited from greater sectoral involvement, and was discussed and approved by the Council of Ministers (CoM).3 To translate medium-term policies and strategies into annual opera- tions, the two instruments used are the state budget (Orçamento do Estado, OE) and the social and economic plan (Plano Económico e Social, PES). The Figure 12.1 Mozambique: Planning and Budgeting Instruments government's 5-year programme (PQG) agenda 2025 medium-term programming annual operational plans integration of objectives & action plan to reduce targets + absolute poverty (PARPA) economic & social plan (PES) coordination of monitoring prioritization harmonization of 116 process of resources sectoral policy & strategy integration of sectoral & provincial strategic plans sector objectives & targets consistency of budgeting with policy integration of global resource actions and objectives sectoral policy & envelope & allocation strategy among sectors macroeconomic & sectoral scenario medium-term expenditure framework (CFMP) expenditure state budget (OE) limits Source: Sulemane 2005. Mozambique: Promoting a Unified Process 117 state budget contains all revenue and expenditure information, and is accompanied by a statement that outlines the main fiscal policy initia- tives and explains how these affect fiscal aggregates and revenue and expenditure projections. The social and economic plan describes the state of the economy over the previous year and lays out the main priorities of government policy for the following year. It discusses the evolution of general macroeconomic indicators and sector priorities more specifically. On budget policy, it offers a single chapter that only covers aggregate fig- ures on revenues and expenditure. It also highlights the percentage of public spending that is devoted to the priority sectors identified in the PARPA, with a view to assessing government performance in relation to Mozambique's agreed target of devoting 65 percent of total expenditure to priority sectors.4 Having two separate documents--the state budget and the social and economic plan--that detail budget policy and program objectives separately constitutes one of the main limitations of the Mozambican system for integrat- ing budget, planning, and PRS content and reporting, and for orienting budget management toward results. In most of the country's sector ministries, two separate administrative units formulate the state budget and the social and economic plan. Furthermore, the decision made by the government to split the Ministry of Planning and Finance into two ministries could mean that the same duplication will now happen at the central level, despite cur- rent efforts at promoting coordination and integration. Integrating Reporting on the PRS and the Budget At the local government level, Mozambique has no significant building blocks for integrated reporting. Provincial and district governments are decentralized units of central government, implementing sectoral programs that are defined at the central level.5 Urban municipalities have full political and financial autonomy, but they control only a small percentage of overall government expenditure,6 and each is legally accountable only to its respec- tive municipal assembly, sending consolidated accounts and reports to central government only for information. Sector instruments are the main building blocks of the PRS and budget reporting system in Mozambique. Sector decision makers use two parallel sets of instruments when monitoring and reporting on their plans and budgets: · central systems driven by medium-term instruments such as the five-year plan, the PARPA, the CFMP, and the annual social and economic plan 118 Mozambique: Promoting a Unified Process · sector-specific instruments linked to SWAp arrangements, with sector strategic plans that are operationalized through annual operational plans, often in accordance with a separate calendar and systems that are dictated by whatever specific agreement the sectoral ministry has reached with its main donors. The sectors' reporting mechanisms are equally separate. The centrally driven Balanço do PES (the annual report on the implementation of the social and economic plan for the previous year) is produced in parallel to sector- specific reports that feed into annual reviews of SWAp arrangements, again normally according to different calendars and formats. This use of parallel sets of instruments has created difficulties in the inte- gration of reporting mechanisms in recent years--including difficulty in providing the necessary incentives for sector managers to supply good information for compiling the social and economic plan at the central level. Nevertheless, the gradual improvement of the central mechanisms, along with a shift in donor practices toward increased harmonization, is reducing the limitations and contradictions.7 Mozambique's poverty-related surveys have limited value as instru- ments for reporting on budget and policy implementation. The household surveys are generally recognized as having been of a high standard, but their usefulness as a monitoring and reporting mechanism is limited; the analyses of the resulting data so far have not assessed whether and how the changes in poverty are linked to the implementation of government policies, and do not consider the relationship between poverty reduction and the country's strategies as contained in the PARPA (Isaksen and others 2005). Surveys undertaken by civil society are mostly qualitative rather than quantitative, and because they are more forward looking than evaluative they are of limited value as monitoring instruments. The Budget Framework Law and the SISTAFE Law on an integrated financial management information system have laid the ground rules for a more comprehensive and predictable system for reporting, in particular on budget execution. Since 2001­02, the MoF has produced regular quar- terly budget execution reports (relatórios de execução orçamental, REOs) within six weeks of the end of each quarter, and year-end consolidated accounts (conta geral do estado), which are submitted to the Supreme Audit Institution (Tribunal Administrativo) by the end of May of each year for the previous year's accounts. Although the comprehensiveness and reliability of these reports is sometimes questioned (for an assessment, see Lawson, de Renzio, and Umarji 2006), the fact that they are compiled regularly and Mozambique: Promoting a Unified Process 119 on time and made available is a significant step forward. This is partly the result of the new legislation, but it also reflects agreements that the gov- ernment has signed with donors providing budget support to allow for monitoring and reporting on the use of resources that are channeled through the treasury. As regards PRS reporting, the government has insisted throughout the PRSP process that PRS reporting mechanisms such as the annual progress report should not duplicate existing ones. The annual report on the social and economic plan (BdPES), which is compiled by government and sent to parliament in February of every year, was the obvious choice for reporting on PRS implementation. However, the numerous shortcomings of the BdPES made it necessary to create a parallel document for reporting on the first two years of the PARPA, 2001­02, and this subsequently led to the establishment of a parallel monitoring mechanism, the performance assess- ment framework (PAF), as discussed in greater detail below. A gradual alignment is taking place between the PARPA and the social and economic plan, in the sense of linking more directly the priorities iden- tified in the PARPA with the activities covered by the social and economic plan and reported on in the annual report on that plan. The missing links between this and other monitoring systems are partly a consequence of the factors noted above, stemming from the fragmenta- tion in sector instruments, and from the incentives faced by sector officials whose funding comes mostly through sector-specific donor interventions. This reality is also reflected in the lack of detailed budget execution analysis in the BdPES, and the lack of explicit links between policies implemented, results achieved, and funds spent. The part of the BdPES dealing with budget policy is limited to very aggregate information on outturns for general budget categories, or for the PARPA priority sectors as a whole, without further breakdown. Ongoing reforms are meant to tackle a number of these problems and to strengthen the link between public spending and development results. The pressure on the government to do so comes from two main quarters. First, the parliamentary Commission for Planning and Finance (Comissão do Plano e Orçamento) is increasingly urging the government to use public funds more effectively: it regularly presents government with comments and requests on state budget and social and economic plan proposals, ques- tioning numbers, providing evidence from field visits, and pushing the government to provide better and more comprehensive information. Second, pressure comes from donors, especially through the so-called G-18 group of donors that provide direct support to the government budget. 120 Mozambique: Promoting a Unified Process Two instruments that were introduced with the memorandum of under- standing on general budget support have contributed to a shift not only in donor-government relations but also in the use and effectiveness of govern- ment reporting instruments: · The PAF matrix has proved to be an effective mechanism for focusing both the government's and donors' attention on a limited set of agreed indicators. Since 2005, the PAF matrix has appeared as an appendix to the social and economic plan document that is submitted to parliament, and it is expected to be subsumed in the new PARPA monitoring and evaluation matrix. · The joint review process, which takes place twice every year between the government and budget support donors, has allowed for a clearer focus on the government planning and budgeting cycle as a basis for common policy discussions and for overall monitoring and reporting. In the past few years, budget policies and PRSs have become better linked, as the result of the new legislation on public financial management combined with the new arrangements related to general budget support, which have gradually improved the reporting mechanisms. The quality and coverage of both the REOs and the BdPES have been improving. Adding the PAF matrix as an annex to the social and economic plan has allowed donors and the government to widen their policy dialogue, bring parliament more clearly into the picture, and improve civil society involvement in sectoral working groups. Furthermore, the gradual evolution of the PAF as the main monitoring mechanism, along with the focus on bringing all aid within the budget, has provided stronger incentives for sector managers to take part in centrally driven policy and reporting processes. The inclusion of sectoral indicators in the PAF, to allow performance to be monitored in priority sectors as defined in the PARPA, has shown sector managers the benefit of being involved in the joint review process, to the point that sectors have been pushing to have "their" indicators included in the framework, because this gives them more visibility and legitimacy even when negotiating addi- tional or separate donor funding.8 Integration of reporting instruments has been promoted by the fact that the system consistently relies on existing domestic processes and instru- ments. But it has been limited by the fact that REOs are produced separately from the implementation reports on the BdPES. The REOs focus almost exclusively on detailed expenditure outturns, assessing the degree to which Mozambique: Promoting a Unified Process 121 budgets have been executed as planned. The BdPES reports all information regarding the implementation of planned activities, but it gives expenditure data at a very aggregate level, to assess whether the 65 percent target of expenditure in PARPA priority sectors has been achieved. It is difficult to see how these reporting instruments could be integrated without radically rethinking the nature of planning and budgeting instruments themselves, for example, by merging the state budget with the social and economic plan. The current fragmentation limits the results orientation of the reporting system, compounding the difficulties that the budget classification system encounters in linking expenditure to programmatic objectives. On the positive side, the arenas for discussing performance information have multiplied, even though they are still fragmented and are putting con- siderable strain on an already overstretched civil service. Apart from the twice-yearly parliamentary analysis of the REOs and the BdPES, the joint review has become a very active forum for discussion, involving more and more civil society actors, despite their limited capacity to engage. The limi- tation of such discussions is that they tend to focus either on the consistency and coverage of budget numbers, or on outputs and outcomes, rather than on linking the two together in meaningful ways. Conclusions and Lessons Main Issues in Mozambique · Poor integration between sectors and central agencies, and between planning and budgeting institutions and mechanisms, makes for owner- ship problems in sector ministries, and a lack of results orientation. · This is partly due to the fragmentation in donor interventions, includ- ing the large proportion of aid that remains outside the budget, thereby increasing the autonomy of sector ministries and undermining central mechanisms for planning and budgeting. The decision in early 2005 to separate the planning and finance ministries may exacerbate some of these problems. · The lack of adequate systems for results orientation is also partly related to the budget classification system. This does not allow for a significant shift toward a program budgeting approach, and thus prevents clearer links from developing between expenditure and outputs or outcomes. · Some ongoing reforms, aimed at bringing more aid within the purview of the budget and at improving performance monitoring processes, are gradually strengthening central mechanisms and addressing 122 Mozambique: Promoting a Unified Process fragmentation, but because they rely heavily on donors' involvement and impose a big administrative burden, they risk being too invasive to be sustained. · The nature of Mozambique's political system, especially the lack of serious challenge functions within the cabinet and in parliament, does not easily promote accountability and integration. Further hopes for increasing domestic accountability could rely on civil society actors, the academic community and research institutes, and the media, despite their current lack of adequate involvement and capacity. · Generally, the important improvements of the past five years are very encouraging, although their sustainability and consolidation will crucially depend on political stability and continued donor support and their commitment to implement the Paris Declaration principles. Lessons for Other Countries · Insist on not duplicating reporting mechanisms, but rather focus on improving existing ones. Early on, Mozambique adopted REOs and BdPESs as donor reporting instruments for general budget support related to the implementation of the PRS. This helped to strengthen domestic processes and promoted integration. · The pros and cons of different aid modalities, including their side effects on the incentives for integration and fragmentation, have a fundamental influence on a country's ability to strengthen its budget and PRS processes and instruments. Countries should try to build the right links between the budget and the PRS from the start, for example, by correcting the incentives that are generated where donors maintain a strong focus on individual sectors, or by not allowing off- budget funding to persist. · Coherent, comprehensive arrangements for policy dialogue with donors are very important, especially in aid-dependent countries. When governments take the lead, and donors collaborate construc- tively, incentives can quickly change for the better. 13 Rwanda: Domestic Accountability as a Driver of Integration Rwanda's Poverty Reduction Strategy (PRS) process has established most of the right budget links in principle, but these have not yet achieved their full potential. Significant progress has been made in elaborating strategic plans, which are a key tool to link the PRS with the budget. Difficulties are caused by a long-standing separation of planning and budgeting and by problems with aid alignment, as well as by discrepancies between budget formulation and execution. Reforms are now in place to tackle most of these issues. Rwanda's experience shows that having regular frank debates within the government is one of the most effective ways to improve performance. The government is now driving the demand for better reporting and accountability for the use of funds, in particular through a high-level annual retreat that focuses on ministries' accountability for the implementation of strategies. Country Context Rwanda is one of the world's 10 most densely populated countries, and, based on a household survey in 2000, 60 percent of its people live below the poverty line. Per capita income is US$230 and the population of about 8.2 million is growing at 2.7 percent annually. Years of political instability and a tragic genocide in 1994 devastated Rwanda's economy and its institu- tional capacity, including that for public financial management. The 1994 genocide was ended by the Rwandan Patriotic Front (RPF), which removed the previous government. A nine-year transition period cul- minated in presidential and legislative elections in 2003 and a return to a multi-party political system. The elections produced a resounding victory for the governing RPF and the current president, Paul Kagame, who thus assumed a seven-year term of office. 123 124 Rwanda: Domestic Accountability as a Driver of Integration Official development assistance (ODA) accounts for more than half the government budget, and, at US$44 per capita, is substantially higher than the average for Sub-Saharan Africa (US$28). More than 70 percent of the ODA is project and program based, but general budget support accounts for about 27 percent and is growing. The largest donors of budget support are the European Commission, the UK Department for International Develop- ment, and the World Bank. Rwanda's long-term aspirations are spelled out in Vision 2020, which envisages that Rwanda will attain middle-income status (per capita GDP of US$900) by 2020 through a transformation from an agrarian to a knowledge- based society. The Office of the President and Cabinet developed Vision 2020 between 1998 and 2000, and though it has never been officially validated or translated into Kinyarwanda it is a very popular and highly "owned" docu- ment of which all citizens are aware. As the name suggests, it contains broad orientations rather than operational details. The PRS provides medium-term orientations for the national develop- ment objectives, which are supposed to be further detailed in sector strate- gies and district development plans. A medium-term expenditure framework (MTEF) then links these plans to the annual budget. The interim Poverty Reduction Strategy Paper (PRSP), prepared in November 2000, provided the starting point for a broad participatory and consultative process that culminated in the publication of the PRSP in 2002, covering the period 2002­05. The PRSP centers on six priority areas for public action: rural development and agricultural transformation, human development, eco- nomic infrastructure, governance, private sector development, and institu- tional capacity building. Though considerable weaknesses exist in its planning, budgeting, and policy review, Rwanda has a high-paced reform agenda and professes its willingness to address identified weaknesses. It is now elaborating its second- generation PRSP, providing an opportunity to tackle inefficiencies in the PRS process from within the PRS document itself. Linking the PRS and the Budget Rwanda has three phases of review processes with various levels of reporting requirements and processes that could be further streamlined to reduce the burden on government: · a process around the budget, with strategic issues papers, a budget framework paper, and a budget execution report; · a process around the PRS, consisting of Joint Sector reviews and the PRS annual progress report (including the policy matrix); and Rwanda: Domestic Accountability as a Driver of Integration 125 · a process around the government's seven-year political program, with annual action plans and implementation reports. All these processes have compatible broad objectives but need to be better integrated and streamlined. The most important challenge is to link both the reviews and action plans to budget execution information. Rwanda's PRSP process incorporates most of the right budget links, with a well-thought-out budget calendar (Table 13.1). Rwanda's seven-year politi- cal program is a political statement based on both the long-term Vision 2020 and the medium-term PRSP. The president chairs an annual high-level cabinet retreat where sector priorities and progress are discussed and an opportunity for cross-sectoral discussion is provided. The retreat is informed by reporting on the implementation of activity-focused annual action plans (AAPs) prepared in the previous budget year. Results from the ministerial retreat provide the context for joint sector reviews that are conducted at a technical level, include a variety of stakeholders (i.e., nongovernmental organizations [NGOs] and donors), and have a clear focus on outputs and results. The joint sector reviews in turn feed into the joint budget support review with donors (initiated in 2005, and not reflected in Table 13.1). Table 13.1 Rwanda: Annual Budget Calendar Government seven-year Annual budget PRS political program Jan Macroeconomic review Reports on annual action plans implementation (previous year) Feb High-level government retreat Mar Budget execution report on previous year Joint sector reviews & Apr Budget call circular update of strategies May Strategic issues papers 1st draft PRS APR and provisional MTEFs Jun Budget consultations Jul Budget framework paper Final PRS APR Aug Cabinet discussions Sep-Oct Nov Annual action Vote in parliament plans submitted to Dec PM's office Source: Authors. 126 Rwanda: Domestic Accountability as a Driver of Integration However, the integration of planning and budgeting could be further strengthened. A long history of separate institutional responsibilities has separated planning from budgeting. Within the Ministry of Finance and Economic Planning, two different units are responsible for planning and budget formulation, and at the local level the situation is similar for line ministries and districts. At the beginning of the public expenditure cycle, strategic planning is generally geared more toward developing aspirational visions than to laying the foundation for the budget process. At the other end of the cycle, the policy review processes are only slowly starting to look at budget execution infor- mation; they often fail to generate any specific recommendations for the budget, and indeed they have limited formal mechanisms for influencing it. The link from Rwanda's PRS to the budget process relies crucially on sound sector strategies. Enormous progress has been achieved in elaborating such strategies, but some sectors still lag behind; their plans do not take account of projections of available resources, and they contain no proper costing of desired activities and outputs, and little analysis on prioritization. The relevance of sector strategies for the budget process is not always appreciated, partly because of the fragmentation of Rwanda's planning and budgeting processes. As a consequence, the potential benefits from Rwanda's MTEF as a tool to link plans and budgets are not fully realized. The MTEF has remained largely a budgeting rather than a planning tool. Substantial aid inflows remain off budget, compounding the lack of inte- gration between planning and budgeting. In 2003, for example, an amount equivalent to almost 40 percent of the development budget was not cap- tured, while overall donor disbursements amounted to twice the amount of the development budget. A number of efforts are under way to create systems to more effectively capture external aid and ensure its alignment with government priorities. The separation of planning and budgeting units throughout government, as well as a line-item approach to budgeting and presentation of budget, have undermined the PRS-budget link at a technical level. There is an increased move toward output- and program-oriented budgeting, but capacity remains weak. The weak capacity of ministries to carry out the required planning and analysis, and the historically weak incentives to demonstrate value for money spent, further compound the problem, sometimes causing differences between budget formulation and budget execution. Reforms are in place to tackle most of these issues, with a new Organic Budget Law, efforts to integrate the recurrent and development budgets, and a new aid policy document completed and validated in 2006. Dramatic Rwanda: Domestic Accountability as a Driver of Integration 127 reforms are under way in decentralization, which are expected to change the whole context of planning and budgeting from top-down to bottom-up, although many of their details have yet to be finalized. Integrating Reporting on the PRS and the Budget PRS and budget reporting have evolved considerably since 2002. As well as annual progress reporting on the PRS, Rwanda also has a proliferation of other reporting processes, including joint sector reviews, a public financial management and budget support review (undertaken jointly by the govern- ment and budget support donors), and the AAP reports on the government's seven-year program. All contain overlapping content. The multiplicity of reports makes both analysis and feedback to deci- sion making somewhat disjointed. Furthermore, the reports focus on outputs rather than outcomes, and they lack meaningful budget analysis, limiting a good comprehension of the planning­budget formulation­budget execution chain. Three annual progress reports (APRs) on the PRS have been prepared. The APR was intended to satisfy most external demands for accountability but it falls short of deriving policy recommendations on the basis of sector progress. A range of stakeholders contribute to producing the APRs. Donors have been active throughout in recent years, producing detailed joint com- ments on the APR, but civil society participation has been mixed and the future role of NGOs in the APR process is not clear. From donors' perspective, joint sector reviews that the government undertakes with donors and other stakeholders in the course of preparing the APR are critical. These reviews assess the past year's progress in specific sectors, provide input into the APR, and also help to shape the budget formulation process. Budget monitoring in most parts focuses on the proper use of funds, rather than tracing the outcomes of spending. Public expenditure reviews, which have recently been introduced in all sectors, provide a direct entry point to link budgets with the achievement of strategic objectives, and they will need to be firmly integrated into the joint sector review process. Like the budget execution report, the report by the Office of the Auditor General thus far restricts itself to the proper use of funds. The reporting on the AAP associated with the government's seven-year program does not fit directly into the PRS-budget cycle and is not well known among external stakeholders. However, these reports arguably provide the best internal scrutiny of the implementation of national plans, 128 Rwanda: Domestic Accountability as a Driver of Integration and therefore they have the potential to lend support to the PRS process. The performance-oriented budgeting measures emerging from these reports may take a few years to work smoothly, but they are likely to directly draw together planning and budgeting processes in an unprece- dented way. Ownership and Accountability Rwanda has few mechanisms that require spending agencies to justify their budgets and demand accountability for money spent in pursuit of national development objectives. Demand for reporting in the sector ministries has been limited, creating few incentives for developing clear performance indicators or participating in joint sector reviews. Progress in strengthening strategic plans and forums for discussion has been important to create internal demand. And sustained, fully aligned donor support has been an important common factor across the better-performing sectors. At the local government level, the signing of district performance con- tracts with the president has led to a focus on indicators that can be mea- sured and evaluated. More work is needed to refine these indicators to reflect objectives and to provide a fair basis on which to evaluate the dis- tricts' performance. But the integration of these performance aspects into the planning phase gives the potential to progress from visionary planning to strategic budget planning. More broadly, the performance indicators pro- vide a clear framework to establish domestic accountability at a level directly relevant to citizens. Parliament does not yet play a clear role in holding the government to account for the results of development strategy. It does not officially receive the APR or a summary of its findings and hence is not in a position to ask questions on this report. Parliament sees its most important role as verifying whether financial inputs were properly used to generate outputs (efficient use of funds), and to a lesser extent whether the right outputs were produced to yield desired outcomes (effective use of funds). However, it recognizes the need to take this approach, and is pondering options for getting the necessary support to do so. Domestic civil society has a low capacity to analyze PRS and budget information, and its place in the APR process is currently unclear. Donors show strong demand for the APR but, probably because these reports still fall short of their expectations, they have not reduced their other reporting requirements. The growing proportion of aid being channeled in Rwanda: Domestic Accountability as a Driver of Integration 129 the form of budget support will increase development partners' interest in the PRS-budget link. The government is now driving the demand for better reporting and accountability for the use of funds. Its high-level annual retreat may have had an important impact here, with its focus on accountability for the implementation of strategies. The comprehensiveness and timeliness of reports against AAPs, and political commitment to the resolutions coming from the retreat, attest to the strength of the AAP as a monitoring tool for internal government stakeholders. Conclusions and Lessons Rwanda's PRSP has succeeded in setting out widely agreed national objec- tives and has significantly improved the content of policy debate in Rwanda. But though it envisaged clear mechanisms to link strategy with the budget, these mechanisms still need to be improved. Sustained progress has been made in strengthening strategic plans and MTEFs, but systematic annual strategic reviews resulting in budget recommendations have yet to become firmly entrenched in the annual activity calendars of line ministries. Suggestions for Improvement in Rwanda · To further integrate planning and budgeting processes, clarify what each of the current array of planning and reporting instruments is expected to do. Such a review should be mindful of the government's limited capacity. Better integration of the various tools may go a long way to reduce the overall workload. · Reconsider the role of the PRS APR. In Rwanda, the need for account- ability may be better served by the joint sector reviews and the joint review of budget support. As government appears to be moving toward a shorter "Statement of Priorities" to draw together the con- clusions for the budget process that emerge from the joint sector reviews, it may be advisable to reduce the frequency of the APR to every two years (or reshape it as a mid-term review of the current iteration of the PRS), with a view to providing the overarching analysis that previous APRs have failed to present. · View the process around the AAPs as a more general mechanism to plan and monitor the implementation of government programs, thus gradually aligning it with the APR process. The annual operational 130 Rwanda: Domestic Accountability as a Driver of Integration action plan developed in the education sector shows how the AAP can easily be integrated into the budget process by adopting the same program/sub-program/output/activity structure as is used to prepare the budget. · Support the move to performance-oriented budgeting by providing technical support to the Parliamentary Budget Committee so it can better analyze the appropriateness of the budget and its implementation. · External stakeholders should strive to further strengthen the joint sector review process. The cluster system (a forum bringing together all stakeholders in a particular sector chaired by the lead ministry and co-chaired by a lead donor) provides a direct entry point for develop- ment partners to support such a holistic process, far beyond the specific reporting needs of individual donors. The joint education sector reviews provide a good example of how such efforts can be managed. Lessons for Other Countries · Avoid institutional duplication. This may sound obvious, but may not be simple given the large number of stakeholders involved. Coun- tries with PRSPs have tended to develop multiple and sometimes parallel instruments geared to the needs of specific audiences, often leading to a depletion of already weak national capacity. In Rwanda the APR, joint sector reviews, budget support reviews, and public expenditure reviews all serve important purposes, but a proliferation of tools undermines the value of each one, while draining the capacity to do any of them properly. · Use performance orientation to establish clear objectives in the plan- ning phase and to link budget lines to objectives to provide a natural force to bring the planning and budgeting processes more closely together. In Rwanda, strengthening the MTEF in both its presentation and usage (for example, by submitting the outer years to parliament) is critical to this process. Output reporting is a logical starting point for performance orientation, building on existing reporting tools and presenting data that irrefutably link budgets to their usage. · Donors' support for domestic annual review processes, rather than periodic follow-up on their own specific projects, will help to strengthen incentives for better reporting on the use of resources to achieve national objectives. Performance-oriented reporting is likely to help here by throwing light on projects that appear to lie outside the government's stated strategies. Rwanda: Domestic Accountability as a Driver of Integration 131 · External and domestic accountability mechanisms are not necessarily at odds, although Rwanda's experience shows that having regular frank debates within the government is one of the most effective ways to improve performance. Alignment of domestic and external accountability is desirable to the extent that it prevents the prolifera- tion of tools, reducing the possibility of conflicting strategies or actions and the inefficient use of scarce resources. 14 Tanzania: Developing a Structured Link between the PRS and Budgets Tanzania's first Poverty Reduction Strategy (PRS) was developed in the context of multiple competing planning processes and though its priorities initially received preference in the budget, this practice was quickly eroded. Under the second-generation PRS, several initiatives seem likely to strengthen the links between planning and budgeting. In particular, Tanzania has gone the furthest of any country to develop a comprehensive structured link between a PRS and resource allocations as part of the budget process. But the PRS has not succeeded in unifying the separate planning processes of ministries and sectors, although there are now internal initiatives promoting greater integration. On reporting, Tanzania's experience highlights the difficulties of integrating budget and PRS reporting processes when routine data systems and budget reporting are both weak. Demands for accountability are also weak in Tanzania and a key impetus comes from donors in the context of general budget support. Country Context In recent years, Tanzania has achieved 6 percent growth in GDP, main- tained inflation below 5 percent, and improved the business investment climate. Economic stability is a direct result of tight budget controls, which have benefited from cohesive political support. Agricultural growth has lagged behind that of other sectors, keeping a high proportion of the population in poverty, but recent improvements are encouraging. Since 2000, public services in education, health, and road maintenance have significantly expanded. 133 134 Tanzania: Developing a Structured Link between the PRS and Budgets Tanzania is a multi-party parliamentary democracy. For two terms fol- lowing the first multi-party elections in 1995, Benjamin Mkapa, leader of the country's dominant party, was president, and presided over the recent period of stability. The leadership was subsequently taken over by Jakaya Kikwete in 2005, and CCM again dominated the parliamentary and presidential elections that year, taking 86 percent of the directly elected parliamentary seats. The executive is powerful relative to parliament, and an influential technocratic civil service has emerged, which plays an important role in policy formula- tion and implementation. Tanzania depends on official development assistance for about 40 percent of its public spending. More than half this aid is program aid--general budget support plus debt relief under the heavily indebted poor countries initiative. Donors use common modalities for budget support, including linking their disbursements to a common performance assessment framework (PAF) and an annual review process. They are also moving away from fragmented pro- jects to common basket funds in sectors such as education, health, and local government. Nonetheless, uncoordinated project-based aid, which bypasses the budget process to varying degrees, continues in many sectors. When the first Poverty Reduction Strategy Paper (PRSP) was adopted in 2000, it continued a history of planning for poverty reduction that reflected Tanzania's socialist roots. Other reforms were also in progress at the time. Technical administrative reforms, notably in tax administration and the accounting function, were backed by significant political support and resulted in much better fiscal discipline. Reforms around the planning and budgeting process included the introduction of a medium-term expenditure framework (MTEF) and public expenditure reviews. Decentralization reforms, still in progress, are devolving the bulk of basic service delivery to local authorities. Linking the PRS and the Budget The first PRSP was developed in the context of multiple competing planning processes spearheaded by different cross-cutting institutions, including the Ministry of Finance (MoF), the president's offices responsible for planning and public service management, and the vice president's office. In addition, many line ministries were undertaking their own initiatives in strategic planning and budgeting. The PRSP helped the government to establish a formal set of strategic objectives and priorities for poverty reduction, and the seven priority sectors that it identified became the focus of budgetary allocations for poverty Tanzania: Developing a Structured Link between the PRS and Budgets 135 reduction. Early in the PRS period, budget allocations shifted strongly toward those sectors. However, under pressure from ministries whose budgets were not boosted by the PRSP, the budget shares of the priority sectors actually declined toward the end of the PRS period. Few explicit links between the PRSP and the budget evolved at this stage. In particular, there was rarely an explicit link between the targets and results specified in medium-term expenditure documentation and the strategic plans and objectives specified in the PRSP. The second, ongoing, version of Tanzania's PRS, the Mkukuta, was pre- pared in a more inclusive and collaborative fashion. It is explicitly a more comprehensive framework, focusing on growth as well as poverty reduction, and replaces the focus on seven priority sectors with three broad clusters of priority outcomes: growth and reduction of poverty, quality of life and social well-being, and governance and accountability. Recent initiatives are likely to improve the integration of the PRS with other strategic planning and with the budget, and to improve institutional coherence. The president's office responsible for planning has been made a ministry, and the new ministry has been given responsibility for the PRS, with the transfer of the Poverty Eradication Division (PED) from the vice president's office. Acknowledgment of the fragmentation of planning and budgeting initiatives has led to a joint Harmonization Initiative by the concerned institutions. Most notably, the Mkukuta has been embraced by the MoF and at the technical level it is emerging as the main focus for resource allocation. To ensure that spending agencies link their budget allocations to poverty reduction outcomes, the MoF has developed and introduced a computer- ized Strategic Budget Allocation System (SBAS). In this system, Tanzania has gone the furthest of any country to develop a comprehensive structured link between a PRS and resource allocations as part of the budget process (Box 4.1 in Chapter 4). Significant progress has been made in budget execution, and public finance outcomes relating to budget execution are relatively strong, com- pared with those in other countries studied. But Tanzania's budget is still rarely implemented as planned, pointing to a lack of realism in departmental budgeting and a lack of political commitment to adherence to the budget. This indicates that much remains to be done in improving the allocation phase of the budget. The challenge for both the SBAS and the Harmonization Initiative, if they are to foster the deeper technical and greater political ownership of the budget that is needed, will be to evolve from purely technical initiatives into 136 Tanzania: Developing a Structured Link between the PRS and Budgets instruments that facilitate consultative decision-making processes and thus can deliver strongly owned strategies and budgets. The change initiatives in progress do not emphasize the role of policies and strategies at the ministry, department, or agency level in deepening the links between the PRS and the budget, and it is questionable whether the Mkukuta is yet a driver for harmonizing and unifying the sector and ministry strategic planning process. The latter is still spearheaded by the president's offices responsible for public service management. Symptomatic of these problems is a need to improve the policy relevance and operational efficiency of funding allocations. Budget documents are overly detailed and obscure the issues for decision making; there is an undue proliferation of budget performance indicators; and the MTEF is unreliable as a planning tool. Often, too, there are inconsistencies in figures in the SBAS, the MTEF, and the integrated financial management system. Budgets at the local level are even less reliable than the national budget. Although planning and budgeting processes in local governments are bottom-up, they are not shaped by knowledge of available resources from central government, so it is almost impossible for them to be realistic. These problems are significant because many Mkukuta outcomes rely on local gov- ernments to deliver basic services. Integrating Reporting on the PRS and the Budget Tanzania's experience highlights the difficulties of integrating budget and PRS reporting processes when routine data systems and budget reporting are both weak. In addition, it emphasizes that without a budget reporting system linked to results, it is difficult to establish whether a PRS is being imple- mented as planned. This difficulty applies to agency-level reporting as well as to overall reporting on budget performance. The paucity of routine infor- mation in many sectors reflects the lack of performance orientation in government institutions, and a lack of coherent demand for performance information from stakeholders either within or outside government. In the context of the first PRS, beyond reporting on priority expenditures, budget reporting systems were not aligned with the PRS and did not relate expenditures to results. Associated review processes were not well established. Today, budget reports do not help track progress in budget performance in the context of Mkukuta implementation. Working groups and reviews relating to the three clusters of Mkukuta objectives do useful work, but do not serve the need to hold specific units of government responsible for specific outcomes. Tanzania: Developing a Structured Link between the PRS and Budgets 137 Incentives for good government performance are weak in Tanzania, so the demand for information on results, and the responsiveness to this infor- mation, are also weak. The cabinet and parliament do not actively demand performance information, and Tanzania's lack of a cabinet committee struc- ture means that opportunities to scrutinize performance are limited. This raises the importance of improving the ownership of plans and budgets at the technical level, and of maintaining review processes that foster demand for implementation. Another key weakness is that information is not provided in a way that is relevant to decision making and that thus can foster accountability. For exam- ple, information is generally unavailable to shape cabinet deliberations on how different expenditure choices affect program outputs and policy outcomes. Donor dialogue around the PRSP has fostered demand for information on outcomes. In the absence of adequate routine reporting systems, the PAF used for general budget support has acted as a proxy for performance reporting, and also fosters demand for performance. Reviews of general budget support, which is now better aligned with the PRS, are another source of demand for data on the performance of the strategy. Given the diffusion of accountability and the weak political demand for performance from the bureaucracy, the emphasis placed on sector reviews in the context of the dialogue on general budget support is important. These reviews are a potentially important entry point for generating demand for per- formance; in them, ministries are held clearly responsible for specific results. In conjunction with the first PRS, a Poverty Monitoring Master Plan was established in 2001; this was being revised in the light of the Mkukuta at the time of writing this report in early 2006. The move of the PED, and primary responsibility for coordinating the monitoring functions of the government and for the Mkukuta, to the new Planning Ministry is important here. Current thinking in PED suggests that the agency will pursue a more integrated approach to monitoring government performance, placing greater emphasis on linking the APRs on the Mkukuta to reporting from the strategic planning and budgeting processes, as well as the outputs of sector reviews. Ownership and Accountability Tanzania is typical of many developing countries in its dominant executive, weak parliament, and high donor demands for accountability as a result of a heavy dependence on aid. Technocrats dominate decisions on planning and budgeting. No organization or constitutional entity in Tanzania effectively challenges the efficiency of the budget process at either the technical or the 138 Tanzania: Developing a Structured Link between the PRS and Budgets political level. Power is concentrated in the executive, with limited space for parliament to alter budget proposals. Ownership of the first PRS was not particularly deep, but the strategy was perceived (at first) to be legitimized by the allocation of budget resources to its priority sectors. Ownership of the Mkukuta, which was pre- pared with broader and deeper consultation, is much stronger. The current technical initiatives to integrate planning and budgeting and SBAS are internally driven, boding well for ownership, but they need to be balanced with initiatives to promote political decision making. Ownership of the budget does not appear to be strong, as evidenced by the gap between the budget and actual spending; this probably reflects the shallow consultative processes on the budget, narrow space for political decisions, and the rigidity entailed in sticking to Mkukuta priorities. Given Tanzania's technocratic budget process, powerful executive, and the dominance of the ruling party in parliament, members of parliament have lit- tle propensity to scrutinize or criticize the government's performance. Civil society participation is, however, increasing in the policy processes associated with the Mkukuta and individual sectors. For their part, members of parlia- ment see nongovernmental organizations as providers of development and not as promoters of accountability. The greatest contribution to the use and strengthening of domestic accountability systems has come from the trend toward general budget sup- port. Tanzania's case clearly shows how general budget support can help raise the profile of and support the upgrading of domestic accountability systems--through donors' engagement in budgetary processes and public finance management systems--and can help create space for civil society involvement in accountability processes. A separate annual review process and PAF is still in use for general budget support, although it is increasingly aligned with the PRS. The general budget support review process has helped to maintain the pace of reform in public finance management, and has pro- vided performance incentives more generally in the context of Mkukuta implementation. The PAF used for budget support has played a positive role in demanding performance, as domestic processes have been built. Though much progress has been made in developing common systems of accountability for different forms of aid, the continued widespread use of project-based funding and basket funds, with their own separate account- ability requirements, distorts the incentives faced by spending agencies, and results in multiple accountability channels. There are clear examples in which donors' own accountability require- ments have reinforced domestic accountability, but in other cases donors' Tanzania: Developing a Structured Link between the PRS and Budgets 139 demands have crowded out or undermined domestic accountability. For example, though donors have helped to open up the space for civil society engagement in policy processes, their representatives in practice tend to outweigh those of civil society in the policy dialogue. Furthermore, donors often have preferential access to information, and often policy documents are prepared in English and not Kiswahili. Conclusions and Lessons Suggestions for Improvement in Tanzania · Address key weaknesses and gaps in the budget process: -- treat the strategic plans of ministries, departments, and agencies as central to the PRS-budget link, and ensure that the links from plans to resource needs are clearly reflected in the justifications given for allocating resources; -- give priority to the efficiency of budget allocations, including examining the balance between funding of central agencies and that of service providers; -- use previous years' MTEFs, rather than ministry requests, as the basis for setting budget allocations; -- improve the relevance of performance information in budget for- mulation documents by reducing the number of performance indi- cators and providing information on their status and future targets; and -- simplify the presentation of the budget to make it more relevant to decision makers. · Deepen participation in the preparation of strategic plans and budgets to foster ownership. · Develop routine reporting systems for ministries, departments, and agencies (as is currently proposed) and for the budget as a whole (as is not). Root these reporting systems within domestic decision-making processes, such as sector reviews, and within formal cabinet and parliamentary structures. · Help to improve domestic accountability by providing support to parliamentary committees, developing clearer rules for civil society engagement, and improving collaboration within government. Donors could support their demands for accountability by moving from common basket funding to sector or general budget support, aligning the PAF with the Mkukuta matrix, and limiting their involvement in sector-level dialogue. 140 Tanzania: Developing a Structured Link between the PRS and Budgets Lessons for Other Countries · Where planning and budgeting instruments compete, it is difficult to elaborate a clear PRS-budget link. · Any PRS-budget link is undermined if ownership and hence the reliability of the budget is weak. · Sector and agency strategic plans are important for articulating a clear link to PRSP priorities and agency budgets. Strategic plans are a key potential instrument in elaborating how sectors and agencies intend to achieve PRSP outcomes and implement cluster strategies. · Integration initiatives at the technical level need to be balanced with those that promote political decision making, especially where, as in Tanzania, the bureaucracy is very influential in budget decision making. · The contrast in ownership between the two iterations of Tanzania's PRS supports the hypothesis that broader frameworks, and deeper participatory processes, help to deepen ownership. Where different planning instruments competed, the priority lent to resource alloca- tions in the first PRS also helped to foster greater attention to and subsequent ownership of the second PRS. · Weaknesses in routine data systems and budget reporting make it very difficult to integrate budget and PRSP reporting processes. · Designing reporting so that it focuses on outcomes is attractive, but this needs to be accompanied by routine reporting on performance lower down the results chain and assigning clear responsibilities for results. This has not happened in Tanzania, where routine service delivery reporting is weak and results are attributed to collections of institutions. In this context, sector review processes are an entry point for creating incentives for performance at lower levels. · Donor dialogue can stimulate demand for performance information, but not guarantee that the information will be used in decision making. · Weaknesses in domestic demand for performance and accountability undermine the prospects for integrating planning and budgeting. · General budget support has helped to raise the profile and sup- ported the upgrading of domestic accountability systems, but the continued widespread use of project-based support undermines these positive effects by distorting incentives, and results in multiple accountability channels. · Ultimately, political priorities have the overriding influence on policy. Donor interventions are only likely to foster greater accountability if they are either integrated with, or support, improvements in domestic accountability. 15 Uganda: Building on Existing Systems for Monitoring Results Uganda was a pioneer of the Poverty Reduction Strategy (PRS) approach, and the Ugandan "model" was a major force behind the introduction of the global PRS initiative by the IMF and World Bank. Uganda's PRS, the Poverty Eradication Action Plan (PEAP), is well integrated with domestic systems and quite strongly linked to the budget, through a pragmatic combination of instruments and processes. Uganda also has a relatively sophisticated system for monitoring its PRS and budget. Despite recent concerns over governance issues, the unusual degree of domestic ownership and early integration with domestic systems mean that Uganda's experience with the PEAP provides an opportunity to learn about entry points in strengthening domestic accountability in PRS processes. Country Context One of the world's poorest countries, Uganda has achieved sustained eco- nomic growth and poverty reduction over nearly two decades. More than half of its people were living in poverty in 1995 but the proportion has now fallen to 31 percent in 2006. Strong fiscal discipline and macroeco- nomic stability since the early 1990s have helped to keep the fiscal deficit and inflation relatively low, and economic growth has averaged more than 6 percent a year over the last 15 years. Increases in per capita income have been slowed by rapid population growth, which averaged 3.2 percent a year between 1990 and 2002. Uganda has achieved large improvements in primary school enrollment rates, safe water supply, and access to health facili- ties. HIV/AIDS remains the leading cause of death, but HIV prevalence has 141 142 Uganda: Building on Existing Systems for Monitoring Results fallen from around 18 percent of the population in the early 1990s to around 6­7 percent today. Uganda depends heavily on aid. Since the late 1990s, aid inflows have averaged more than 10 percent of GDP and 50 percent of rapidly increasing public spending. Much of the aid takes the form of budget support. Develop- ment partners rely to a great extent on the government's own systems of planning, budgeting, and reporting. The National Resistance Movement government led by Yoweri Museveni has been in power for 20 years--first within a democratic but "no-party" system, and since 2005 within a multi-party system embodied in a revised constitution that also allows President Museveni to stand for a third term. The government has pursued public sector reforms oriented to reducing poverty, and relatively innovative approaches to the management of aid. Its relations with the international community were very good for a long time, but have recently been marked by tension over governance concerns, espe- cially in the run-up to the 2006 elections. In the late 1990s and early 2000s, a coalition of interests among the presidency, technocrats, and donors helped to create a strong demand for reorientation of government policies toward poverty reduction, and for implementation of those policies. Now, however, the executive has turned its attention away from core poverty reduction policies, weakening this coalition, diluting the demand for government performance, and reducing the room for evidence-based decision making. The collaborative decision- making processes established earlier are proving valuable in maintaining the demand for strategy implementation. Linking the PRS and the Budget Compared with other countries, Uganda has developed important links among its poverty reduction policies, strategies, and public expenditures. Complementary reforms in budgeting and planning, combined with political support for poverty reduction and rapid increases in public resources, allowed the government to make significant shifts in budget allocations toward pro-poor policies. The 1997 PEAP was Uganda's first effort at a comprehensive strategy to reduce poverty. The subsequent two iterations of the PEAP (2000 and 2004) have also acted as Uganda's Poverty Reduction Strategy Papers (PRSPs). The PEAP is prepared by the Ministry of Finance, Planning, and Economic Development (MFPED) on the basis of a consultative process, and is perfor- mance oriented. The 2000 PEAP specified a set of priority poverty monitoring Uganda: Building on Existing Systems for Monitoring Results 143 indicators, and the 2004 PEAP went further; its implementation matrix specifies performance indicators alongside policy actions. Political ownership of the PEAP was helped early on by the fact that it embodied key new political priorities such as universal primary education and raised the profile of other basic services such as primary health care and water and sanitation. In addition, when the first PEAP was being prepared, in the mid-1990s, the government was also introducing budgetary reforms such as a medium-term expenditure framework (MTEF) and then output- oriented budgeting, both of which helped to add a strategic dimension to resource allocation. In the late 1990s, sector strategies were developed in key sectors such as roads, education, and health, along with processes for reviewing sector performance. Such sector-wide approaches (SWAps) are now present in most but not all sectors. All these reforms benefited from an unusual degree of domestic initiative and strong leadership from the unified MFPED. This, combined with their high-level political backing, contributed to the development of consistency between planning and budgeting instruments and processes (Figure 15.1). Three main levels of strategic planning have emerged: · National: the PEAP is the single overall national planning document. It influences priorities for spending not only among, but within, sectors, where strong sector-level policy and planning processes encourage a link between plans and budgets. Figure 15.1 Uganda: Planning and Budgeting Instruments strategy/plan budget MTEF, national budget framework poverty eradication paper; background to the budget; action plan national and the annual budget sector MTEF and sector strategic plans sector sector budget framework papers local government: local government: budget framework LGs papers, annual budget & work plans; & development plans; central A government: corporate plans central government: ministerial policy MD statements, annual performance plans Source: Authors. 144 Uganda: Building on Existing Systems for Monitoring Results · Sectoral: sectors develop strategic plans in the context of their own SWAps. · Institutional: local governments have been preparing three-year rolling development plans since the passage of the Local Government Act in 1997. The annual budget process has two stages: · Strategic resource allocation, linked to sector strategic plans through the MTEF. Sector working groups set out their medium-term budget strategy for inclusion in the overall national budget framework paper, which MFPED presents to the cabinet for approval and then to parliament for comment. · Preparation of the annual budget and presentation to parliament by MFPED. Parliament then discusses line ministry budget proposals in the form of ministerial policy statements. Local councils are respon- sible for approving local government budgets, which are often accompanied by an annual work plan covering activities for the financial year. Both the budget and sector processes have been highly consultative and participatory. Sector working groups, which bring representatives of gov- ernment institutions together with civil society and donors, lie at the center of Uganda's sector policy, strategy, and budget making. Their contributions to the budget framework paper set out the medium-term budget strategy for the sector, with measurable performance targets and resource allocations for the government bodies active in the sector. As well as helping to align the technical ownership of strategies and budgets across different units of the government, the reliance on sector working groups has helped to deepen the PEAP formulation process and broaden the scope of public sector, civil society, and donor involvement. The PEAP is relatively well integrated with domestic systems. At the practical level, the links between the PEAP and the budget take effect through three initiatives, all relatively simple and largely based on existing budget systems: · Grouping administrative units sector-by-sector in an MTEF. This allowed a demonstration of the link between PEAP priorities and broad sector allocations, and helped to guide overall expenditures in line with the PEAP. The link was made more explicit with the inclusion of a long-term expenditure framework in the third PEAP (2004), which set out long-term sector allocations. Uganda: Building on Existing Systems for Monitoring Results 145 · The Poverty Action Fund, which was established as a means of channeling additional resources from the heavily indebted poor countries initiative and individual donors toward PEAP priority programs and sector-wide approaches. The Poverty Action Fund has helped to highlight pro-poor spending priorities within sectors in the context of the MTEF (though, some observers argue, it has promoted unbalanced implementation of the PEAP, due to the early focus of the Poverty Action Fund on the social sectors). · Output-oriented budgeting has not yet been implemented, and awaits the implementation of the results-oriented management framework. If it is used alongside the sector-by-sector presentation in the MTEF, sector performance indicators and targets will be set out in sector and ministry budget submissions. This would help to strengthen the link between budget allocations and results, and consequently the link between budget allocations and the objectives and targets in the PEAP. These initiatives fall short of full program and performance budgeting, but if they are all implemented they promise to deliver a relatively clear link between pro-poor policies and the budget. As noted above, the political ownership of the PEAP and sector strate- gies is now reduced. This is partly because politicians have shifted their attention away from policy content to the political transition process and partly because the cabinet and parliament have not taken part in formu- lating these strategies. The political ownership of the budget seems to have remained stronger than that of the PEAP (as shown by predictability in budget execution), perhaps because the cabinet and parliament are more deeply involved in allocating resources than in selecting priorities for the PEAP. Linking PRS and Budget Reporting Uganda's instruments and processes for monitoring and reporting on PEAP and budget implementation are less well integrated than the instruments and processes used for strategy and budget formulation. Prior to the original PEAP little monitoring or reporting was done for either strategy or budget implementation. Since then, monitoring systems for both have evolved rapidly at the national, sector, and local government levels. Early budget reporting was partial, focusing on programs within the Poverty Action Fund. Since the early 2000s, comprehensive reports on budget 146 Uganda: Building on Existing Systems for Monitoring Results performance have been prepared twice a year by the MFPED. They now provide information on budget disbursements and output performance, where information is available, but not actual expenditures. Local govern- ments report on results and expenditures regularly. However, routine budget and performance reporting tends to be fractured and is not comprehensive. The monitoring of PEAP implementation and poverty outcomes has become increasingly sophisticated. A poverty status report, assessing trends in poverty outcomes, has been prepared every two years since 1999. A PEAP Results and Policy Matrix that sets out progress and future plans has been in use since 2004. This is an important innovation that systematically links policy actions to performance indicators for high-level outcome objectives, but it is still not linked to expenditures. It has started to be used as the framework for the government's dialogue with budget-support donors, but not yet as a broader reporting tool. The first consultative Annual PEAP Implementation Review process took place in March 2007. The PEAP Results and Policy Matrix, combined with the introduction of the Annual PEAP Implementa- tion Review, could potentially help to integrate the reporting on the PEAP implementation with that on budget implementation, but this relationship is not yet explicit. Reporting on the poverty strategy needs to better address the needs of decision makers, by focusing on the appropriate level of detail and by link- ing information on results to information on expenditures. The reports on PEAP implementation, such as poverty status reports, have tended to focus on the impact and high-level outcomes of the strategy, and not to relate these back to public expenditures. Poverty-monitoring information on high- level outcomes is not the type of information that the public or politicians are likely to demand the most, or find useful for decision making. To facili- tate decision making, reporting on PRS implementation needs to track more systematically what happens at lower levels of the results chain, such as service delivery outputs--although linking these to outcomes is also impor- tant. Also, a failure to link higher-level PEAP (and sector) reporting to expenditures undermines the usefulness of this reporting for judging the efficiency and effectiveness of spending. Uganda's practice of linking reporting to consultative decision-making processes helps to broaden the demands for good performance and to strengthen incentives for delivering it. Sectors and their associated review processes have been a key pillar of reporting, and broad stakeholder involvement has helped to foster demand for information. But the lack of cross-sector policy review mechanisms beyond the budget has left little role for overall reporting on the PEAP. Uganda: Building on Existing Systems for Monitoring Results 147 Incentives and Accountability As donors have increased their provision of budget support, their demands for the executive to be accountable have become increasingly aligned with the government's own decision-making processes. Donors have helped to build processes that have provided more space for domestic accountability, but they have often inadvertently dominated the policy dialogue. And it is those reporting instruments of interest to donors that tend to get strength- ened, rather than those of interest to local constituents. General budget support (GBS) donors' commitment to using the PEAP Results and Policy Matrix to monitor progress in performance shows how GBS systems can be aligned with domestic systems. It remains to be seen whether the same can be done with the Annual PEAP Implementation Reviews and the dialogue relating to GBS. At the sector level, at least, the experience shows that external support can help to build policy processes that evolve into domestically driven processes. Donors' use of government systems can help to create the space for domestic accountability to develop, yet donors need to be careful not to crowd out domestic voices in the dialogue. There are more fundamental areas of concern. Early in Uganda's reform process, strong management in key institutions such as the MFPED, com- bined with political support, helped to create strong incentives for imple- menting the PRS and sector policies. The current political environment is undermining such incentives. And a failure to adequately address key issues in public service reform may further weaken them. In this context, the domestic ownership of PEAP-related reforms, as well as accountability systems, is likely to become more important if the govern- ment is to continue to deliver against its ambitious poverty reduction agenda. Parliament--which has shown significant interest in the budget in the past and now has the added dimension of a multi-party system--has emerged as a potential entry point for strengthening incentives and accountability. Conclusions and Lessons Uganda's experience emphasizes the value of integrating the PRS and budget, but also the potential vulnerability of such reforms to changes in the domestic political environment. Compared with those in other low-income countries, Uganda's PRS and budgetary processes are grounded more in domestic systems. This has helped to ensure a clearer framework for decision making, a lesser degree of 148 Uganda: Building on Existing Systems for Monitoring Results competition between planning and budgeting instruments, and greater integration of systems than in countries where the PRS has been externally driven. Strong leadership in a unified Ministry of Finance and Planning helped in both the drive to reform and its coherence early on, although there are now signs of fragmentation. Suggestions for Improvement in Uganda · To build on the relatively strong PRS-budget link, establish PEAP objectives that are more consistent with the sector structure of the budget, and align sector resource allocation with those objectives in the medium- and long-term expenditure frameworks. · To increase focus on the efficiency of spending, improve the economic presentation of sector and agency budget frameworks and annual estimates; and introduce a provision for linking central ministry, department, and agency operational plans to the budget and inte- grating them into the ministerial policy statements that are presented to parliament for discussions on the budget. · To maintain and broaden stakeholders' involvement in decisions around the PEAP and the budget, and to deepen political ownership, involve parliament in either deeper discussion or approval of PEAP and sector strategy documents. · To ensure that the proposed annual PEAP review process has domes- tic political ownership, focus on informing cabinet decision making; link policy implementation to the budget; and introduce committee structures within the cabinet to review performance against policies and the budget. · To improve performance reporting, ensure that future annual sector and PEAP reports examine the achievement of results in the context of budgeted and actual expenditures; tailor the presentation of infor- mation to facilitate decision making; and upgrade those processes and instruments for which domestic demand is strongest, such as the ministerial policy statements discussed by parliament. · Introduce sector review processes in sectors where they do not exist, and formalize the requirements for sector performance reporting. · To foster accountability, nurture demand for accountability where it is weak, such as by introducing collaborative processes around the PEAP; strengthen mechanisms of accountability to the public at differ- ent decision-making levels (national, sector, local government); and encourage civil society organizations to work with parliament as well as directly with ministries. Uganda: Building on Existing Systems for Monitoring Results 149 · It is also important to allow the institutions responsible for promoting accountability to act with independence. Recommendations contained in their reports and those of the various commissions of inquiry into mismanagement of public funds need to be implemented to improve accountability and to ensure value-for-money of public expenditures. · Donors should actively try to support domestic voices in the policy processes, focusing on the quality of the processes and restraining their own engagement in them. Lessons for Other Countries · Domestic political ownership of the pro-poor policy agenda, starting with political leadership by the executive, is crucial in fostering strong incentives for PRS implementation, and consequently in fostering resource allocations that are consistent with PRS implementation. · Pro-poor policies must be central to the domestic political agenda if a PRS is to have strong political ownership. And to remain relevant, a PRS framework must be flexible and responsive to the domestic policy environment. · To facilitate links between PRS objectives and budget allocations, high-tech budget reforms are not necessary; simple changes to budget systems can be introduced relatively easily for this purpose. The focus needs to be on presenting information on performance and expendi- tures in a way that is relevant for policy, both when formulating PRSPs and budgets and when reporting on their implementation. · Sector-level policy and review mechanisms are important for generat- ing deeper links between policies and resource allocation, and incen- tives to improve performance. · The sector, with associated strategies and an MTEF, is a key level for deepening the integration of the PRS and the budget, both technically and in terms of participation through sector working groups. · The integration of processes is just as important as technical consistency in planning and budgeting instruments. The high degree of technical ownership within Uganda's civil service, of the PEAP as well as sector strategies, is related to the breadth of participation and the reliance on sector working groups. · Integrating reporting processes with decision-making processes helps to reinforce incentives for good reporting. Results information needs to be provided in a form that decision makers can readily use. · Incentives for performance by a weak civil service can be strengthened in the short term by a strong political drive for performance, and by 150 Uganda: Building on Existing Systems for Monitoring Results having strong managers in key institutions, but they will remain vulnerable if the demand for performance slackens. Tackling broad public sector reform issues early on is therefore important. · External support, and demands for accountability, can help to improve reporting on policy implementation and support the establishment of domestic policy processes, in the context of which domestic incentives for performance can be built. Annex 1: Lessons from Higher-Income Reformers This analysis is based on a review of reform experiences in four countries (Australia, Chile, South Africa, and the Republic of Korea) that are internation- ally considered to be reformers in public financial management and effective service delivery.1 Clearly, given the stark differences that exist between higher- income reformers and the main case studies covered in this paper in terms of capacities, levels of income, and dependency on external assistance, the rele- vance of the findings from past reform efforts might be called into question. Nevertheless, some of the higher-income reformers were facing problems 10 or 20 years ago that are similar to those that a number of PRS countries are facing today. Therefore, the common issues and lessons emerging from the analysis of successful past reform efforts can shed some light on the factors that seem to have determined their success, and provide useful guidance for present reform efforts in Poverty Reduction Strategy (PRS) countries. Country Summaries Australia In Australia, planning and policy development are very much focused at the sector level. Reform efforts, as a consequence, were not based on a centrally determined master plan, but were experimental, pragmatic, and often opportunistic. Australia has been engaged in a continuous reform process since the mid-1980s, when a consensus emerged on the need to address fiscal imbalances and management weaknesses. The labor government that took office in 1983 saw improving the quality of public sector performance as a key electoral imperative. The starting point for reform efforts was changing the systems and processes at the center of government, the institutional framework that gener- ates many of the incentives that determine the effectiveness of the public sector. Budget reform has been underpinned by (i) political commitment to an 151 152 Annex 1: Lessons from Higher-Income Reformers aggregate constraint over the medium term; (ii) sector policies that recognize such aggregate constraints; (iii) a top-down approach to budgeting; and (iv) a set of forward estimates for the cost of government policies. A key to success was the political power invested in the Expenditure Review Committee (ERC), a sub-committee of cabinet. The ERC was responsible for determining the overall fiscal framework and for managing strategic policy making, including policy changes necessary to reflect fiscal realities as well as the shifting priori- ties of the government. As a result, policy decisions were given much greater legitimacy, as they were in a contest both as ideas and for funding. This demanded better analysis and underpinned the increased predictability of policy and funding. Providing the spine to institutionalize these changes was the system of forward estimates (FEs), which introduced a medium-term perspective to budgeting. FEs have been published since 1983. Publication has meant that the government has imposed on itself a requirement to disclose and justify the costs of policy decisions over the three-year FEs period. This process and system provided a framework for a more strategic approach to decision making, much greater predictability in funding for current policies, and for removing from the budgetary arena those decisions best made elsewhere (most notably management decisions about personnel and administrative resources). Political commitment to a disciplined budget framework, and clarity over the level at which decisions should be made, have been crucial for the institutionalization of these systems over time, and for the introduc- tion, in the early 1990s, of greater performance orientation. Chile Chile is well known for its sophisticated system of management control and results-based budgeting.2 Performance evaluation was introduced in the early 1990s, partly as a response to demands from the legislature to have better information about government programs to increase its ability to affect budget decisions. In the beginning, evaluations were simple desk-based exercises based on secondary sources. Later, more thorough impact evalua- tions were added, with more complex methodologies for data collection and analysis aimed at assessing the actual impact of government programs. In 2000, when the Lagos administration took office, the budget office merged the various existing monitoring and evaluation (M&E) mechanisms, estab- lished the management control system, and subjected ministerial targets to several methodological improvements to better assess their compliance with the president's policy priorities. At a later stage, a bidding fund for Annex 1: Lessons from Higher-Income Reformers 153 government priorities was introduced to improve resource allocation to new initiatives, and to promote increased performance. An interministerial committee is responsible for ensuring that the performance evaluation system is consistent with government priorities, and that conclusions from evaluations are shared and taken into account in policy making. Adequate technical support is provided, including in the selection of the external consultants carrying out the evaluations. The important role of the Ministry of Finance (MoF), and its powerful position within the government, has ensured the sustainability of the M&E system. However, its lower level of ownership and utilization by sector ministries means that there is an unexploited opportunity for sectors to use the gener- ated information in their strategic planning, policy development, and ongoing management and control. South Africa In South Africa, the major driving force for public service reform in general, and budget reform in particular, has been the government's efforts to deliver on its mandate to ensure adequate service delivery to the poor and vulnerable after the African National Congress (ANC) took power in 1994. An internal debate within the ANC was won by the side that stressed the need to ensure sound macroeconomic fundamentals before expanding service delivery more rapidly. Therefore, the accountability structures put in place in the initial stages of the budget reform process were designed to enforce fiscal discipline and expenditure control. It was only after the budget deficit targets of the growth, employment, and redistribution strategy were attained and the expenditure overruns of social service departments were curbed that atten- tion returned to the broader accountability for service delivery. The budget reform process in South Africa was largely driven by the executive arm of government, and notably the minister of finance and the presidency. Even though the National Treasury opted for an inclusive process that was mediated by various intergovernmental and interministerial committees, it remained the driving force, and its leadership was built and consolidated on the dual foundations of political clout of the minister of finance and the technical capacity of the team of civil servants working with him. Early success in reforming tax administration and raising additional revenues brought about increased support for the reform effort. The South African approach to budget reform has placed great emphasis on drawing political office bearers into the resource allocation and application process. This took the form of executive consideration and approval of critical 154 Annex 1: Lessons from Higher-Income Reformers reforms. The reformed budget process has involved national and provincial cabinets in overseeing and managing the entire process, thus creating a link between policy planning, budget drafting, and execution. This led to a greater degree of contestability and buy-in within the executive and cabinet, and to what Folscher and Cole (2004) call "political peer pressure." The decisions of the cabinet were widely publicized through the pre- budget statement3 and other budget-related documents. This combination of inclusiveness and transparency increased the pressure on political office bearers to adhere to the decisions that were made. As Folscher and Cole put it, "Together with appropriately timed public statements to signal closed-off budget decisions, a transparent budget process, the system of hard budget constraints, and the provision of good technical support to these forums on the financial implications of policies, this reduces the potential of accessing fund- ing `through the political back door' or through in-year budgeting games." Political buy-in is not only important in the formulation and allocation phase of the budget process. In many countries the allocation process and buy-in to it will be compromised by nonadherence to the approved budget and the skewing of budget priorities by over- and under-expenditure. For this reason the National Treasury involved both the cabinet and parliament in the expenditure monitoring and control aspect of the budget process. Actual expenditure reports were also referred to national and provincial cabinets, along with parliament and the nine provincial legislatures. This ensured collective policing of the "political bargain" that was struck in the formulation of the budget. One of the main challenges faced by South Africa lies in deepening the reforms that have been introduced so far. On one hand, after improving fiscal discipline, improved internal accountability, and overall expenditure control, the National Treasury has struggled to formulate and monitor appropriate indicators of efficiency and effectiveness of public spending, hampering a shift to performance orientation. On the other hand, despite the improved public availability of information, the accountability of the executive to parliament and to the public at large remains to be developed to any significant degree. The Republic of Korea Korea is seen as a successful example of the integration of planning and budgeting. As an instrument of central government planning, five-year economic plans have been adopted from 1962 to 1996. After the Asian financial crisis in 1997, Korea formulated a medium-term fiscal plan focusing on recovering its fiscal soundness, which rapidly deteriorated right after the crisis. With the beginning of a new administration in 2003, Korea further Annex 1: Lessons from Higher-Income Reformers 155 strengthened its efforts to integrate medium-term fiscal planning and annual budgeting by introducing a five-year national fiscal management plan (NFMP) and a medium-term expenditure framework (MTEF), along with top-down budgeting. The primary objective of these initiatives was to integrate planning of national policy priorities into budgeting and set up a new strategic decision-making mechanism. While the previous medium-term plans were only used to provide overall guidelines on operating public finance and the planning did not take place within the formal budget cycle, the current MTEF (NFMP) provides not only medium-term fiscal targets but also a specific budget resource allocation plan that line ministries should comply with in their annual budget requests. Another important point is the changes that took place in the roles played by the Ministry of Planning and Budget (MPB) and line ministries in fiscal strategy planning and annual budgeting. Through the top-down approach in budgeting, the MPB shifted its role from controlling individual programs at the micro level toward establishing medium-term fiscal strategies and coordinating sectoral and ministerial expenditure ceilings. As line ministries allocate resources autonomously among their programs within their expen- diture ceilings, they are changing from irresponsible budget requesters into self-disciplined budget compilation partners (Figure A1). Some of the main success factors in the Korean reform program include (i) strong support of political leadership; (ii) the introduction of a special Figure A1 Framework and Stages to Integrate Planning and Budgeting in Korea stage 1: medium-term plan (Jan­Apr) stage 2: cabinet meeting (Apr) · 5-year fiscal targeting · sectoral/ministerial ceilings · national goals & priorities · based on MPB proposal · sectoral/ministerial ceilings integrating tools key players · long-term vision · MPB & line ministries · NFMP (MTEF) · president & ruling party · top-down · national assembly budgeting · civil society (NGOs) stage 4: MPB's consultation and stage 3: line ministries' budget review (Jun­Aug) requests (May­Jun) · expenditure ceilings & budget · autonomous resource formulation manual allocation within ceilings · changes after cabinet meeting · budget formulation manual Source: Sang Dae Choi, "An Overview of Integrating Planning and Budgeting in South Korea." 156 Annex 1: Lessons from Higher-Income Reformers retreat cabinet meeting chaired by the president, where consensus is built on medium-term fiscal strategies and resource allocation among line ministries; (iii) a legislative framework that institutionalizes the integration of planning and budgeting mechanisms; and (iv) the concentration of the main functions in a single agency. Present challenges relate to the development of an adequate monitoring system, and to ensuring performance and accountability through results-oriented management. Emerging Themes Focus on Linking Budgets and Policies, not Budgets and Plans A striking feature of the higher-income reformers is the fact that none of them except Korea had a development strategy or plan whose implementa- tion was driving budget reforms. The main emphasis was not on how to turn long-term plans into annual budget allocations, as is often the case with PRS countries, but instead on how to ensure that budgets were better tools for the implementation of different government policies, ensuring aggregate fiscal discipline, prioritization of resource allocation, and efficiency and effectiveness in spending. In particular, budget reforms were often born out of a need to ensure fiscal discipline, whether in response to a regional crisis (Korea) or to create a solid basis for expansion of service delivery (South Africa). Reconciling fiscal discipline with policy objectives in turn created the need to develop a medium-term budgeting perspective, to allow for better prioritization and clearer contestability. There were no cases of stated conflict between the enti- ties responsible for managing the finances and those in charge of planning. In most cases there was a single agency in charge of both, with a single reform strategy. Importance of Core Policy Processes, and the Changing Role of Sectors All successful budget reforms considered here emphasized the importance of core policy-making processes. Most of these were in the form of cabinet- level structures, such as committees and review bodies that discussed fiscal frameworks and budget priorities, as well as getting involved in monitoring the execution of the budget and evaluating its performance. These struc- tures ensured the direct engagement and involvement of politicians (mostly ministers, but in some cases also parliamentarians) at key decision-making points, ensuring the necessary buy-in and political peer pressure for "sticking to the path." Annex 1: Lessons from Higher-Income Reformers 157 All cases also highlight the shift in the role played by sectors as a result of the implementation of budget reforms. Countries moved from a bottom-up to a top-down approach to budgeting, but provided clarity about what that meant (i.e., cabinet sets high-level policy goals, ministers define sectoral priorities, managers work out implementation details using given adminis- trative flexibility). The interaction and division of roles between central agencies of government (MoF, cabinet) and between them and the sectors was well thought out. In particular, the role of budget analysts within finance ministries was important in assisting sectors to comply with general rules and guidelines. Starting with the Basics, but Promoting Integrated Reforms Interestingly, in all four countries reforms started with basic measures, tackling capacity constraints and existing incentive systems, then proceeded gradually to introduce more sophisticated systems. As highlighted above, demand for macro-fiscal discipline was the starting point in Australia and South Africa. The importance of credible budget information and effective expenditure controls therefore received most attention, along with mech- anisms for encouraging more disciplined behavior from the main actors. Mechanisms for introducing a medium-term dimension came next (Australia and Korea), with a view to enhance the policy orientation of the budget, and turn it into a more useful tool for policy planning and implementation. Performance orientation came last, and in many countries (except Chile) is still lagging behind, sometimes even after 20 years of budget reform. Even in the more successful case of Chile, performance orientation developed gradually, starting from a very simple base. Despite their gradual implementation, reform agendas in higher-income reformers were integrated, in the sense that they did not cover just budget formulation, but also accounting, monitoring, and reporting. In some cases (notably Australia and Chile) they also involved changes in core public administration structures, dealing with the flexibility of managers, pay structures, performance contracts, etc. Reform, in this sense, was much more than identifying specific measures to implement, but identifying inter- linked sets of issues that were key in determining outcomes, knowing that different reforms would reinforce each other. Local Drive and Internal Accountability As can be expected in countries with limited donor intervention, budget reform processes in higher-income reformers were largely internally driven, 158 Annex 1: Lessons from Higher-Income Reformers and resulted from government efforts to deliver on electoral promises (Australia, Chile, and South Africa), while at the same time responding to the external environment (Korea and Australia). Ministries of finance were inevitably the main initiators and guardians of the reforms in all countries, although mostly through defining the rules and facilitating the engagement of other actors; cabinets were also heavily involved in all cases, but in some cases parliaments and civil society also played a part (Chile and South Africa). Accountability worked mostly in its internal, horizontal dimension within cabinets, ensuring that the collegiality of decisions made generated political peer pressure and reciprocal monitoring. The importance of keeping all key actors on board meant that everyone felt that they were "sharing the pain" of reform, or reaping the gains in an equally proportional manner. In some cases (Australia and Korea) spending agencies were given more autonomy and flexibility in return for their compliance. Implications This review of the experience of higher-income reformers reinforces a number of the themes suggested by the case studies of PRS countries. Three common strands in particular are worth pulling out. The first is the importance of understanding the different dimensions of domestic accountability and the different roles that they can play. The important role of cabinet-level structures and of horizontal accountabili- ties within the executive branch of government in driving the successful reforms is quite striking, adding weight to the concerns expressed in Chapter 3 about the way "domestic accountability" tends to be conceived in PRS contexts. A second point on which the two sets of findings agree is the need to take "ownership" seriously, not just as an influential variable but as something that needs to be investigated and understood in a context-specific way, and not assumed on the basis of general formulas. Of course, not being aid dependent makes a big difference. Nevertheless, the observation that most of the higher-income reformers made their budgets more responsive to policy concerns provides additional reasons for recommending a nondog- matic and flexible approach to what should count as a valid PRS. Third, the experiences reviewed not only further underscore the importance of thinking about incentives. They also provide additional support for the argument developed particularly in relation to Uganda, about the greater feasibility and effectiveness of simple reforms. Across Annex 1: Lessons from Higher-Income Reformers 159 the successful higher-income reformers and the PRS countries we seem to detect a tendency of simple reforms, with plenty of mutual reinforcement, to generate strong incentives, while technically demanding reforms tend to have weaker impacts on what actually gets done, especially when intro- duced independently in different spheres. This points a clear way forward for efforts to achieve a better link between PRS and budget reporting. Annex 2: Priority Sectors and Pro-Poor Spending in Selected Case Studies Burkina Faso Since 2000, there have been two interlinked definitions of priority expenditures: first, Poverty Reduction Strategy (PRS) priorities, which are broad and for which specific poverty reduction expendi- tures cannot be identified in the current budget classification; and second, highly indebted poor countries (HIPC) programs that are narrower in scope and are "monitorable," but through a mechanism that separates them from the regular national budget. The Priority Action Program was created in 2004 as a mechanism to allow more precise identification and tracking of PRS priorities. However, in the absence of changes in the budget classification, this is not linked to government public financial management systems, while related donor funding is not reported on. Malawi In Malawi, pro-poor expenditures (PPEs) are a concept introduced as part of HIPC negotiations. To ensure that HIPC resources would result in increases of PPEs, a subset of high-priority activities within the three pillars of the interim Malawi Poverty Reduction Strategy Paper were identified as priority PPEs, and specifically monitored. The 2001­02 budget was the first budget to include PPEs in its budget plan. Mozambique Mozambique's PRS, or PARPA, is founded on the idea that there are certain key sectors that should be favored in the allocation process, and protected from expenditure cuts in the case of liquidity shortages. The priority sectors have been identified in the PRSP, and are tracked as a percentage of total expenditure, with an agreed target of 65 percent. Mozambique's definition of PPE, like Tanzania's initial one, is based on broad sectoral categories, and is criticized for giving too little attention to allocations within priority sectors, to territorial distribution, and to the division between current and capital expenditure. Rwanda The term "pro-poor expenditure" has not so far been used in Rwanda. The focus has been on "priority programs" and "priority (continued) 161 162 Annex 2: Priority Sectors and Pro-Poor Spending in Selected Case Studies Annex 2 (continued) areas," which have been favored in allocations, and protected from cuts. In the 2002 PRSP six "priority areas" were introduced. Within these, specific priorities were to be identified and action plans to be developed. The PRSP provided a set of criteria for prioritization of expenditure for poverty reduction. However, there was no explicit application of the criteria to the extensive set of priority programs identified in Annex 6 of the PRSP. Tanzania Tanzania originally adopted an approach for allocating expenditures to priority PRS areas by including all spending on the seven broad priority sectors. A system of reporting on PPEs was also introduced. However, the disadvantage was that it treated all lines of expenditures within a priority sector as equally important. Since the 2004­05 budget, the priority sectors have been replaced by the Strategic Budget Allocation System (see Box 4.1), which aims to make a more comprehensive link between the sec- ond PRSP and budget allocations. Uganda The programs in the Poverty Action Fund, which was formed in 1998, represent the government of Uganda's PPEs, in response to the HIPC initiative. At the inception of the Fund, they were a selection of priority programs from the 1997 PEAP. In 2000 a defin- ition of PPEs was agreed for inclusion in the Poverty Action Fund, and the number of such programs has grown steadily. While alloca- tions to these programs are integrated within the medium-term expenditure framework, a separate Poverty Action Fund budget is presented in budget documentation. Releases to Fund programs, which are protected, were originally reported on in quarterly reports up until 2000 and since then they have been reported in half-yearly budget performance reports. Source: Country case studies and general budget support evaluation. Annex 3: Nature of the APR in the Case Study Countries Albania The annual progress report (APR) covers a rather short span of time, and a certain degree of formalism accompanies the entire process. After the approval of the Poverty Reduction Strategy Paper (PRSP) by the cabinet, there is little follow-up activity on the shortcomings evidenced in the APR. APRs have an emphasis on descriptive reporting instead of making an evaluation of policies and impacts. Meanwhile, a striking feature of the APR process is the number and variety of actors involved in its preparation; nonethe- less, their participation in the process is motivated by a specific interest, mainly to ensure as much funding as possible for their areas of interest. Burkina Faso The main pillars of PRSP monitoring are the implementation report of the Priority Action Programs (PAPs) of the previous year and the PAP annual review for the coming year. There's been improve- ment in the quality of the documents produced for the annual review with a list of indicators, a simulation of poverty dynamics and poverty-growth links, a programming of investments through PAP. Indicators are not broken down by region. However, no functional links exist between the PRSP annual review process and joint sector reviews, so much so that the useful analysis of joint sector reviews is not always reflected in PRSP review reports. Madagascar The budget and the PRSP reporting systems are not yet integrated, except for six pilot ministries. The current PRSP reporting proce- dural arrangements, as well as the lack of institutional role and political recognition of the entities in charge of PRSP coordination, weaken the quality of the PRSP reporting. The APRs are long and complex documents. The quality of the information limits their use for policy decision making. No analysis is provided on (i) reported developments of core indicators of the PRSP strategy; (ii) comparing planned and realized/achieved activities; or (iii) linking poverty results with sector programs. The line ministry (continued) 163 164 Annex 3: Nature of the APR in the Case Study Countries Annex 3 (continued) buy-in is limited in the PRSP reporting. This can be explained partly by the APR being one of many reports to be produced. The PRSP reports are of less importance than the reporting on Politiques Generales de l'Etat matrixes sent directly to the president's office by the ministers. Malawi The first PRSP ran from FY 02-03 until FY 04-05, and three APRs were produced, the last one in January 2006. APRs were largely activity reports from line ministries compiled in a single document. There is no systematic linking of resource allocations to results. Blueprints for an ambitious poverty monitoring and evaluation system exist, but the system remains fragmented and rudimentary in practice. Mali Two reports (2002 and 2003-04) of relative good quality were produced, while the 2005 APR is in preparation. The APR is explicitly based on annual sector reports; however, baselines are missing in many sectors and there is weak statistical capacity. There has been considerable change in the institutional setup with responsibility now shifted to the Ministry of Finance, but there is low capacity there, too. The process is owned by the technocrats, but the role of civil society and parliament is weak, and there is little public interest. Mozambique The existing Balanço do PES (Report on the Implementation of the Economic and Social Plan) has been adopted as the APR. In the last few years, the PAF policy matrix has been included as an annex to it, providing better incentives for sectors to engage. There is a gradual alignment between the PRSP and the APR, although missing "building blocks" limit the quality of the infor- mation. The joint review process put in place as part of general budget support arrangements is focusing the attention of various actors on PRSP reviews, and ongoing reforms can further strengthen such process. Rwanda The APR is a compilation of reports from joint sector reviews, and there is little overarching analysis. Weak monitoring and evaluation hampers more in-depth analysis. PRSP reporting is not sufficiently integrated into the budget cycle, and the APR lacks traction out- side the Ministry of Finance. Tanzania Early poverty monitoring focused on outcome level information, and was able to flag key issues. However, the limited routine data available undermined the quality of the APRs, which were fairly unsystematic. The poverty monitoring system is under review, and now there is greater focus on integration with routine reporting systems from ministries and local governments. In addition to the Annex 3: Nature of the APR in the Case Study Countries 165 Annex 3 (continued) APR and Human Development Report, other reports aimed at policy makers and the general public are intended to be produced. The strategy could potentially be undermined unless the incentives to collect routine data are addressed. Uganda Reporting on the first PRSP commenced with the poverty status report, which provides information on poverty outcomes and has been prepared every two years since 2000. The APR took the form of the pre-existing background to the budget, expanding its scope to report on Poverty Eradication Action Plan implementation. However, this was weakly linked to the budget. No APRs have been prepared for the second PRS, although they are planned as a separate instrument, and not as part of the background to the budget. Source: Country case studies. Notes Chapter 2 1. See, for example, Bedi and others (2006), which also includes the development of a diagnostic tool to strengthen national monitoring systems. 2. The PRS has become the centerpiece of a more wide-ranging approach to aid policy based on mutual accountability between donor governments and developing countries. This approach is reflected in the Monterrey Consensus of 2002 and in the Paris Declaration on Aid Effectiveness of 2005, which state that reducing poverty requires a renewed partnership linked to the scaling up of international assistance to achieve the Millennium Development Goals (MDGs). 3. Net official development assistance to grow national income (ODA/GNI) in 2004 was 4.7 percent in Albania, 12.2 percent in Mali, and 28.9 percent in Madagas- car. http://www.oecd.org/dataoecd/55/36/23730449.gif and http://www.oecd. org/dataoecd. 4. Public Expenditure and Financial Accountability indicators were reviewed for six countries but individual ratings are considered confidential, except for Tanzania. For aggregate information see chapter 3's section on "Incentives in the PRS and Budget." 5. For example the Background to the Budget document in Uganda. 6. Case study research is increasingly used to study situations like PRSP imple- mentation, where "interest has shifted to organizational rather than technical issues" (Benbasat and others 1987). The value of high-quality analytical narrative has even been recently embraced in the mainstream economic development literature, heralded by Rodrik (2003). Chapter 3 1. A principal-agent model is adopted here because of its wide use in the litera- ture on governance and public management. The model is highly accessible given its extensive use, including the recent IMF paper on agency in public expenditure man- agement systems (Leruth and Paul 2006). Easterly (2005) references principal-agent 167 168 Notes theory as a way of thinking about the problems PRSPs create for aid agencies and governments--with multiple PRSP objectives comprising multiple goals and multiple principles that potentially undermine each other. Accountability has also been defined as a process "by which public officials inform about and justify their plans of action, their behavior, and results and are sanctioned accordingly" (World Bank Institute 2005: 7) or as mechanisms, both within the public sector, and the multiple interfaces among the bureaucracy, elected political leaders, and civil society, of reward and sanctions for public actors for their performance (Gauri 2005). 2. For a more complex framework of accountability relationships, which also stresses the role of rewards and sanctions along with the importance of rules, infor- mation flows, and transparency, see World Bank 2006a: 125. 3. However, other reviews concluded that poverty reduction objectives received high-level endorsement across government, including through discussions at cabinet level (Nankani, Page, and Judge 2004). 4. In the case of Tanzania, CSO representatives noted that access to information from central ministries is often difficult to obtain and there is little interest from par- liament in the analytical work produced by CSOs. 5. The recently published Open Budget Index, compiled by civil society organi- zations in 59 countries around the world, suggests that 90 percent of the countries covered do not provide the accurate, timely, and comprehensive information at each stage of the budget cycle that is required to ensure governments' accountability to citizens (International Budget Project 2006). The Open Budget Index assesses the availability of key budget documents, the quantity of information they provide, and the timeliness of their dissemination to citizens in order to provide reliable information on each country's commitment to open budgeting. For the case study countries covered here it reveals that Tanzania and Malawi provide some informa- tion to citizens, while Albania and Uganda provide minimal information and Burkina Faso provides scant or no information to citizens. Other case study countries are not covered. 6. The criteria are (i) policy-based budgeting and the translation of public policies into specific budget expenditures; (ii) arrangements for predictability, control, and stewardship in the use of public funds; (iii) systems of accounting and recordkeeping to provide information for accountability purposes and external audit; and (iv) other mechanisms that ensure external scrutiny. 7. In spite of these broad improvements in public financial management, many countries need to further improve their budget systems, in particular in the area of budget execution compared to initial plans. HIPC indicators in 2004 for 25 countries reveal that only about one-third of the countries had integrated medium-term projections into their budgets (indicating that they used policy-based budgeting), had somewhat credible budgets in terms of aggregated expenditure outturns com- pared with the originally approved budget, and satisfied at least three of the five criteria for good expenditure execution systems. Only four countries satisfied as many as three of the four criteria for budget credibility (World Bank 2006a: 140; World Bank-IDA and IMF 2005). Notes 169 8. Defined as the ratio of the sum of the highest (A, B) over the sum of the lowest ratings (C, D). Individual PEFA ratings are considered confidential and may not be published except for Tanzania. 9. Note that budget credibility was limited to aggregate expenditure outturns in most countries. Chapter 4 1. This is further complicated in francophone countries where MTEFs were intro- duced when program budget reform was already under way. As the link between the two areas of reform was not always made clear, budget information and presentation in the annual budget, the program budget, and the MTEF can differ. 2. Unfortunately the program classification annexed to the budget is not the same as the one for the MTEF, which has introduced different classifications. 3. In addition, the literature on public financial management reform also stresses the importance of "getting the basics right" (Schick 1998; Holmes and Evans 2003) before pursuing ambitious reform agendas such as MTEFs and perfor- mance budgeting. 4. Ministries of Finance, Agriculture, Education, Health, and Office of the Presi- dent and Cabinet. 5. This should serve as a reminder that donor expectations about the speed of implementation of reforms in planning and budgeting systems are often overly optimistic, and do not take into account some of these lessons from recent history. Chapter 5 1. Guidelines on annual progress reports (APRs) issued by the World Bank and the IMF state that, while there is no required format for APRs, as part of the PRS Initia- tive, APRs are expected to have three basic elements: (i) an evaluation of performance and analysis of outturns relative to benchmarks, for monitoring and evaluation purposes; (ii) an overview of the coming year's policy intentions, particularly, but not exclusively, as reflected in the budget; and (iii) a report on how specific shortcomings identified in joint staff advisory notes have been addressed. In addition, the APR should indicate how annual reporting is being used to provide information on imple- mentation progress to key domestic and external stakeholders. 2. See guidelines for APRs at http://siteresources.worldbank.org/INTPRS1/ Resources/PDFs/jsan_apr_guidelines.pdf. 3. In addition to the case studies, this paragraph draws from Bedi and others (2006: 82, 83). 4. Also, background work for the recent public expenditure and investment review (PEIR) in Tanzania considered the creation of cabinet subcommittee structures. 170 Notes Chapter 6 1. In Tanzania, for example, regular biweekly meetings of public expenditure working groups were seen as becoming overly bureaucratic and crowding out discussions on substance. Similar observations were made about Mozambique. Chapter 9 1. A ministerial restructuring in early 2007 has created a new Ministry of National Economy to improve coordination of all activities related to development planning. Under the new arrangement the Secrétariat Technique de l'Ajustement and the Cellule Technique de Suivi are attached to the Ministry of National Economy. Chapter 10 1. A longer version of this case study has been published by GTZ (Krause 2007). 2. Before these sectors were decentralized, the local sector administrations were part of their respective national ministries. They would initiate the annual planning process by delivering their spending proposals upward in the ministerial hierarchy and at some point around or after the beginning of each fiscal year would receive their allocations from the ministry. These allocations were part of the ministerial chapter in the national budget. Chapter 11 1. A longer version of this case study has been published by GTZ (Fritz and Lang 2007). Chapter 12 1. At present, and in addition to Frelimo's electoral program, the five-year plan draws inspiration from a document called Agenda 2025, which was the result of a national consultation exercise undertaken in 2001­02 to identify the key strategic objectives that the Mozambican people aspired to achieve by 2025. The Millennium Development Goals are also playing an increasing role in defining some overarching priorities for government action. 2. Cabral and others (2005) found that 29 percent of the total resources for the health sector remained off budget at the programming stage, 60 percent at the execu- tion stage, and 44 percent at the accounting stage. 3. Also, significant progress has been made recently to bring donor project aid into the budget. See http://www.odamoz.org.mz. 4. To operationalize the government's commitment to reduce poverty levels, and to better link the PARPA with the annual budget, the government and donors agreed to monitor the percentage of total expenditure devoted to the priority sectors Notes 171 identified in the PARPA: education, health, infrastructure, agriculture and rural development, good governance, social action, and employment. This indicator has been criticized for a number of reasons. First, the basis on which the percentage was chosen is unclear. Second, given the large amount of donor resources spent in the priority sectors but not recorded in the budget, reporting is inevitably incomplete and therefore potentially misleading. Third, its simplicity provides a strong incentive for government officials to adjust the figures to attain the desired numerical value. Related to this, the 65 percent target focuses exclusively on inputs, rather than promoting a results orientation. Also, it tracks broad sectoral allocations rather than identifying more specific actions that are most likely to lead to positive development outcomes. As a consequence of these features, it hardly provides any insights on the content, success, or impact of policies and programs intended to reduce poverty. 5. This may gradually change in the future. In 2006, for the first time, all 128 districts were allocated a fund of about US$300,000 for investment expenditure, to be allocated through local planning processes. 6. Possibly with the exception of large municipalities such as Maputo and Beira. 7. For example, in February 2006, the Minister for Planning and Development issued a directive with the aim of harmonizing the timing of central and sectoral reviews. 8. From interviews with sector officials. Annex 1 1. The topical briefs were written by Malcolm Holmes (Australia), Azul del Villar (Chile), Albert van Zyl (South Africa), and Sang Dae Choi (Republic of Korea). The authors would like to thank Matthew Andrews for his assistance in summarizing the results of the four reviews. 2. See also Mackay (2006). 3. Called the Medium-term Budget Policy Statement. Bibliography Anderson, G., and J. Isaksen. 2002. "Best Practice in Capacity Building in Public Finance Management in Africa: Experiences of NORAD and Sida." Report R2002:16, Chr. Michelsen Institute, Bergen, Norway. Alonso, R., Lindsay Judge, and Jeni Klugman. 2006. "PRSPs and Budgets: A Synthesis of Five Case Studies." In Budget Support as More Effective Aid? ed. Stefan Koeberle, Jan Walliser, and Zoran Stavreski. Washington, DC: World Bank. Bedi, T., A. Coudouel, M. Cox, M. Goldstein, and N. Thornton. 2006. Beyond the Num- bers: Understanding the Institutions for Monitoring Poverty Reduction Strategies. Washington, DC: World Bank. Bennet, T. 2004. "Good Practices in Public Financial Management Capacity Develop- ment." Background paper for Capacity in Public Financial Management course. Washington, DC. Benbasat, I., D. K. Goldstein, and M. Mead. 1987. "The Case Research Strategy in Studies of Information Systems." MIS Quarterly 11 (3): 369­86. Booth, D., K. Christiansen, and P. de Renzio. 2006. "Reconciling Alignment and Per- formance in Budget Support Programs: What Next?" In Budget Support as More Effective Aid? ed. Stefan Koeberle, Jan Walliser, and Zoran Stavreski. Washington, DC: World Bank. Booth, David, and Xavier Nsabagasani. 2005. "Poverty Monitoring Systems: An Analysis of Institutional Arrangements in Uganda." Working paper 246, Over- seas Development Institute, London. Cabral, L., A. Cumbi, L. Vinyals, and S. Dista. 2005. "Estudo sobre os Off-Budgets no Sector Saude: Relatório Final." Ministry of Finance, Maputo. Drazen, A., and P. Isgard. 2004. "Can Public Discussion Enhance Program `Ownership'?" IMF Working Paper No. 163, International Monetary Fund, Washington, DC. Easterly, William. 2005. "How to Assess the Needs for Aid? The Answer: Don't Ask." Paper Prepared for Third AFD/EUDN Conference, Paris, December 14. 173 174 Bibliography Eberlei, W. 2001. "Institutional Participation in Processes beyond the PRSP." GTZ, Eschborn. Entwistle, Janet, and Filippo Cavassini. 2005. "An Operational Approach for Assessing Country Ownership of Poverty Reduction Strategies, Volume I: Analysis and Impli- cations." World Bank, Washington, DC. Folscher, Alta, and Neil Cole. 2004. "South Africa: Transition to Democracy Offers Opportunity to Whole System Reform." www.africa-sbo.org/cabri/uploads/ pdf/Chapter_7.pdf. Fritz, V., and F. Lang. 2007. Strengthening Domestic Accountability in Mali. GTZ Fiscal Study No. 11. Eschborn: GTZ. Gauri, V. 2005. "Development, Human Rights, and Security: Framing the Challenges for the World Bank." Unpublished, World Bank, Washington, DC. GTZ (Deutsche Gesellschaft für Technische Zusammenarbeit). 2004. National Moni- toring of Strategies for Sustainable Poverty Reduction PRSPs. Main report Vol. 1. Eschborn: GTZ. ------. 2005. Making Poverty Reduction Strategies Work. Good Practices, Issues, and Stakeholder Views. Eschborn: GTZ. ------. 2006. Good Financial Governance­­Good Governance in Public Finances. Fiscal Studies No. 3. Eschborn: GTZ. Heidenhof, Guenter, Stefanie Teggeman, and Cia Sjetnan. 2007. "The Leadership Approach to Achieving Change in the Public Sector: The Case of Madagascar." World Bank Institute, Washington, DC. Holmes, Malcolm, and Alison Evans. 2003. "AReview of Experience in Implementing Medium-term Expenditure Frameworks in a PRSP Context: A Synthesis of Eight Country Studies." Overseas Development Institute, London. International Budget Project. 2006. http://www.internationalbudget.org/. IMF (International Monetary Fund). 2001. "Code of Good Practices on Fiscal Trans- parency." Fiscal Affairs Department, Fiscal Transparency Unit, IMF, Washington, DC. http://www.imf.org/external/np/fad/trans/code.htm. IMF and World Bank. 1999. "Poverty Reduction Strategy Papers: Operational Issues." SM/99/290, 12/10/99 and IDA/SecM1999­99/789. Washington, DC. ------. 2005. "2005 Poverty Reduction Strategies Review: Balancing Accountabilities and Scaling Up Results." Washington, DC. Isaksen, Jan, Anette Staaland, and Bernhard Weimer. 2005. "Poverty in Mozambique: Discourse, Analysis, and Monitoring." Christen Michelsen Institute, Bergen. Krause, P. 2007. Strengthening Domestic Accountability in Malawi. GTZ Fiscal Study No. 10. Eschborn: GTZ. Lawson, A., P. de Renzio, and M. Umarji. 2006. Assessment of Public Financial Manage- ment in Mozambique 2004/05. Final Report. Maputo: SAL Consulting and Overseas Development Institute. Bibliography 175 Le Houerou, P., and R. Taliercio. 2002. "Medium-term Expenditure Frameworks: From Concept to Practice. Preliminary Lessons from Africa." Africa Region Working Paper Series No. 28, World Bank, Washington, DC. Leruth, L., and E. Paul. 2006. "A Principal-Agent Theory Approach to Public Expenditure Management Systems in Developing Countries." Working Paper No. 204, International Monetary Fund, Washington, DC. Levy, B., and S. Kpundeh, eds. 2004. "Building State Capacity in Africa, New Approaches, Emerging Lessons." World Bank, Washington, DC. Mackay, Keith. 2006. "Institutionalization of Monitoring and Evaluation Systems to Improve Public Sector Management." ECD Working Paper Series No. 15, IEG, World Bank, Washington, DC. Nankani, G., J. Page, and L. Judge. 2004. Human Rights and Poverty Reduction Strategies: Moving toward Convergence? Washington, DC: World Bank. Norton, A. 2002. "What's behind the Budget? Politics, Rights, and Accountability in the Budget Process." Overseas Development Institute, London. Rakner, L., L. Mukubvu, N. Ngwira, K. Smiddy, and A. Schneider. 2004. "The Budget as Theater­­The Formal and Informal Institutional Makings of the Budget Process in Malawi." Draft. Chr. Michelsen Institute, Bergen, Norway. Rodrik, Dani. 2003. In Search of Prosperity: Analytic Narratives on Economic Growth. Princeton: Princeton University Press. Schick, Allen. 1998. "Why Most Developing Countries Should Not Try New Zealand's Reforms." World Bank Research Observer 13 (1): 123­31. Sida (Swedish International Development Cooperation Agency). 2006. "Budgeting for Poverty Reduction: Evaluation of Poverty Reduction Strategies in Latin America." Article No. SIDA28769, Stockholm. Stevens, M. 2004. "Institutional and Incentive Issues." In Public Financial Management Reform in Poor Countries. Washington, DC: World Bank. Suleman, J. 2005. "Toward Integrated Planning and Budgeting: The Mozambican PRSP Experience." Presentation to Tunis Workshop on Pro-Poor Growth, November. The International Budget Project. 2006. "Open Budget Initiative 2006: More Public Information Needed to Hold Governments to Account." Center on Budget and Policy Priorities, Washington, DC. Tikare, S., D. Youssef, P. Donnelly-Roark, and P. Shah. 2002. "Participation." In A Sourcebook for Poverty Reduction Strategies. Chapter 7. Washington, DC: World Bank. VENRO (Verband Entwicklungspolitik Deutscher Nichtregierungs Organisationen). 2005. "Fighting Poverty without Empowering the Poor?" Germany. World Bank. 1998. Public Expenditure Management Handbook. Washington, DC: World Bank. ------. 2006a. Global Monitoring Report. Washington, DC: World Bank. 176 Bibliography ------. 2006b. "Poverty Reduction Strategies 2005 Review: Lessons and Resources." Note, Washington, DC. World Bank Institute. 2005. "Social Accountability in the Public Sector: A Conceptual Discussion and Learning Module." World Bank, Washington, DC. World Bank IEG (Internal Evaluation Group). 2004. "OED Review of the Poverty Reduction Strategy (PRS) Process." World Bank Internal Evaluation Group, Washington, DC. World Bank and IMF (International Monetary Fund). 2002. Poverty Reduction Strategy Papers­­Progress in Implementation. Washington, DC: World Bank and IMF. ------. 2004. Poverty Reduction Strategy Papers--Progress in Implementation. Washington, DC: World Bank and IMF. World Bank/International Development Association (IDA) and IMF. 2005. "Update on the Assessments and Implementation of Action Plans to Strengthen Capacity of HIPCs to Track Poverty Reducing Spending." Washington, DC. Index Boxes, figures, notes, and tables are indicated by b, f, n, and t, respectively. A SWAp mechanisms in, 67 accountability. See domestic accountability annual progress reports (APRs), 1, 5, in PRS and budget processes 163­165. See also under specific agenda. See future agenda countries, e.g. Burkina Faso AIDS/HIV, 87, 95, 141 decision-making processes, targeting Albania, 1, 14, 65­74 reporting to, 58 APRs, 163 guidelines, 169n1­2 CSOs in, 70, 72­73 integration of budget and PRS reporting decentralization in, 39b systems and, 43­44 domestic accountability in PRS and bud- introduction and uses of, 12­13 get processes, 20, 22b, 23, 26, 69­70, asymmetrical owners of PRS and budget, 168n5 22­24, 23f donor agencies in, 66, 70, 73, 167n3 Australia, 1, 14, 151­152. See also EU and, 66, 68, 71 higher-income reforming countries, IPS (Integrated Planning System), 37, lessons from domestic 39b, 65, 70­71 accountability in PRS and budget lessons learned, 71­74 processes, 158 linking PRS and budget in, 4, 37, 39b, drivers of reform process in, 158 67­68 gradual nature of reform process in, 157 MDG in, 66, 69, 70 linking PRSs and budgets, 35, 36 MTBP (medium-term budget program), 68­72 B NATO and, 66, 68 budgets. See also integrating PRSs and NSDI (National Strategy for budgets Development and Integration), 67, challenge function in, 22b 70­71 performance information integrated NSSED (National Strategy for Social and with, 58, 60t Economic Development), 65, 66­73 policy links from budget up rather than Open Budget Index, 168n5 from PRS down, 56­57 other countries, lessons for, 73­74 program budgeting, 33­34 ownership in, 69­70 PRSPs, budget-friendly formulation reporting systems, integrating, of, 61 43, 44, 68­69 strategic phase, introducing, 57 177 178 Index Burkina Faso, 1, 14, 75­84 D APRs, 79, 165 decentralization CSOs in, 81 in Albania, 39b decentralization process in, 76, 77, 82 in Burkina Faso, 76, 77, 82 domestic accountability in PRS and in Malawi, 97, 170n2 budget processes, 2, 20, 23, 24, 25, in Mali, 106, 109 79­81, 168n5 in Mozambique, 117 donor agencies, role of, 75, 76, 78, 83­84 in Rwanda, 127 HIPC funds, 77, 79 in Tanzania, 134 lessons learned, 81­84 reporting systems, integrating, 46 linking PRS and budget in, 76­78 decision-making processes, targeting MTEF in, 75, 77, 78, 81 reporting to, 57­58, 59t Open Budget Index, 168n5 domestic accountability in PRS and budget other countries, lessons for, 83­84 processes, 2­3, 15, 17­27. See also ownership, 79­81 under specific countries, e.g. priority sectors and pro-poor spending, Madagascar 161 capacity and, 18 PRS/PRSP revision and implementation, drivers of accountability, 17­19, 18f 11, 76, 77, 80 gaps in accountability, closing, reporting systems, integrating, 26­27, 27f 43, 78­79 in higher-income reforming countries, 158 incentives, 24­26, 54 C as driver of accountability, cabinet committee structures, building or 17, 18f reforming, 56 in budget process, 25­26 capacity issues in PRS process, 24­25 domestic accountability in PRS and ownership, 19­24, 54 budget processes, 18 as driver of accountability, 17, 18f higher-income reforming countries, asymmetrical owners of PRS and lessons from, 157 budget, 22­24, 23f in Malawi, 95, 96, 101­102 of budget, 20­22, 21t, 22b linking PRSs and budgets, 34, 36 of PRS, 19­20 challenge function in budget process, 22b principal-agent model, 167­168n1 Chile, 1, 14, 152­153. See also higher-income donor agencies, role of reforming countries, lessons from in Albania, 66, 70, 73, 167n3 drivers of reform process in, 158 in Burkina Faso, 75, 76, 78, 83­84 gradual nature of reform process in Malawi, 102, 103 in, 157 in Mali, 106, 110, 167n3 linking PRS and budget, 35, 36 in Mozambique, 114, 119 civil society organizations (CSOs) in Rwanda, 124, 125, 128­129 in Albania, 70, 72­73 in Tanzania, 137, 138­139 in Burkina Faso, 81 lessons learned, 139­140 in Rwanda, 128 Madagascar, 167n3 in Tanzania, 168n4 other countries, lessons for, 140 PRS ownership by, 20 Uganda, 142 Côle, Neil, 154 Côte d'lvoire, 105 country institutional and policy E assessment indicators (CPIAs), European Union (EU) and Albania, 2, 25 66, 68, 71 CSOs. See civil society organizations existing processes, building on, 55 Index 179 F as driver of accountability, 17, 18f Folscher, Alta, 154 in budget process, 25­26 fragmented processes and institutions, in PRS process, 24­25 problem of, 54­55 future agenda for fostering, 57­58 future agenda, 6­7, 15, 53­62 reporting systems, integrating, 47­51 decision-making processes, targeting building blocks and, 46 reporting to, 57­58, 59t supplying information, 47­50, 48f, existing processes, building on, 55 49b, 50b gradual approach to reform, 55­56 using information to improve incentives, fostering, 57­58 performance, 50­51, 51b key findings for, 53­55 integrating PRSs and budgets, 1­7, 11­15 ownership and support, building, domestic accountability (See domestic 56­57 accountability in PRS and budget performance information integrated with processes) budget documents, 58, 60t future agenda for (See future agenda) policy links from budget up rather than links between PRSs and budgets from PRS down, 56­57 (See linking PRSs and budgets) practical lessons and reform points, 55­61 reporting systems (See reporting systems, simplicity, advantages of, 53, 58­61 integrating) K G Korea, Republic of, 1, 14, 154­156, 155f. Ghana, 11 See also higher-income reforming gradual approach to reform, 55­56, 157 countries, lessons from domestic accountability in PRS and H budget processes, 158 heavily indebted poor countries (HIPC) drivers of reform process in, 158 initiative/indicators linking PRS and budget, 35, 36, 156 Burkina Faso and, 77, 79 MTEF, 155 domestic accountability and, 2, 25, 168n7 Mozambique and, 114 L reporting systems, integrating, 5 linking PRSs and budgets, 3­4, 15, 29­40. higher-income reforming countries, lessons See also under specific countries, from, 1, 14, 151­159. See alsoAustralia; e.g. Albania Chile; Korea, Republic of; South Africa capacity issues, 34, 36 capacity issues, 157 in higher-income reforming countries, core policy-making processes, 156 35, 36, 156 domestic accountability in PRS and methods for, 29­34 budget processes, 158 MTEFs, 3, 29, 30­33, 32t, 33b, 34 drivers of reform process, 157­158 ownership, 34­35, 37 gradual approach to reform, 157 priority sectors or programs, 30, 31b linking PRSs and budgets, 35, 36, 156 pro-poor spending, 30 sectors, role of, 157 program budgeting, 33­34 HIPC. See heavily indebted poor countries PRSPs, role of, 34, 37 (HIPC) initiative/indicators reforms and more effective links, HIV/AIDS, 87, 95, 141 34­37, 38b, 39b technical fixes, inadequacy of, 55 I incentives M domestic accountability in PRS and Madagascar, 1, 14, 85­93 budget processes, 24­26, 54 APRs, 88, 90, 163­164 180 Index Madagascar (continued) PRS/PRSP revision in, 108 domestic accountability in PRS and reporting systems, integrating, 44, 50, 108 budget processes, 20, 21, 22b, 90­91 Mauritania, 11 donor agencies, role of, 167n3 MDGs (Millennium Development Goals), LAMP (leadership action and 2, 24­25, 66, 69, 70, 167n2, 170n1 management program), 37, 38b medium-term expenditure frameworks lessons learned, 91­93 (MTEFs). See also under specific linking PRS and budget in, 3, 4, 34, 37, countries, e.g. Albania 38b, 87­88 failure to execute budget as planned in, MAP (Madagascar Action Plan), 85, 26 87­92 linking PRSs and budgets, 3, 29, 30­33, MTEF in, 87 32t, 33b, 34, 169n1­2 other countries, lessons for, 92­93 technical reforms, inadequacy of, 55 ownership, 90­91 Millennium Development Goals (MDGs), 2, PRS/PRSP revision in, 86 24­25, 66, 69, 70, 167n2, 170n1 reporting systems, integrating, 88­90 Monterrey Consensus, 167n2 Malawi, 1, 14, 95­103 Mozambique, 1, 14, 113­122 APRs, 99, 100, 164 APRs, 164 capacity constraints in, 95, 96, 101­102 decentralization in, 117 decentralization in, 97, 170n2 domestic accountability in PRS and domestic accountability in PRS and budget processes, 20, 21t, 22b, 23 budget processes, 20, 21, 22b, 26, donor agencies in, 114, 119 100­102, 168n5 health sector, fragmentation and duplica- donors, role of, 102, 103 tion in, 49b Integrated Financial Management lessons learned, 121­122 Information System, 35 linking PRS and budget in, 4, 30, 32t, 37, lessons learned, 102­103 115­117, 116f linking PRS and budget in, 3, 35, 96­99 MDG in, 170n1 MGDS (Malawi Growth and MTEF in, 30, 32t Development Strategy), 98­99, 101, other countries, lessons for, 122 102 priority sectors and pro-poor spending MTEF in, 103 in, 161, 170­171n4 Open Budget Index, 168n5 PRS revision, 11 other countries, lessons for, 103 regular meetings as bureaucratic in, 170n1 ownership, 100­102 reporting systems, integrating, 5, 43, 44, priority sectors and pro-poor spending 47, 49b, 117­121 in, 97­98, 161 SWAp mechanisms in, 115, 118 PRSP/PRS revision in, 95 MTEFs. See medium-term expenditure reporting systems, integrating, 43, 44, 46, frameworks 51, 99­100 Mali, 1, 14, 105­111 N APRs, 165 NATO (North American Treaty Organiza- decentralization in, 106, 109 tion) and Albania, 66, 68 domestic accountability in PRS and Nicaragua, 11 budget processes, 22b, 108­109 North American Treaty Organization donor agencies, role of, 106, 110, 167n3 (NATO) and Albania, 66, 68 lessons learned, 109­111 linking PRS and budget in, 3, 4, 34, 106­107 O MTEF in, 105­107, 111 Open Budget Index, 168n5 other countries, lessons for, 111 ownership. See also under specific countries, ownership, 108­109 e.g. Madagascar Index 181 domestic accountability in PRS and in Mozambique, 161, 170­171n4 budget processes, 19­24, 54 in Rwanda, 161­162 as driver of accountability, 17, 18f in Tanzania, 162 asymmetrical owners of PRS and linking PRSs and budgets, 30 budget, 22­24, 23f technical reforms, inadequacy of, 55 of budget, 20­22, 21t, 22b program budgeting, 33­34, 55 of PRS, 19­20 PRSPs. See Poverty Reduction Strategy future agenda for building, 56­57 Papers linking PRSs and budgets, 34­35, 37 PRSs. See poverty reduction strategies Public Expenditure and Financial Account- ability (PEFA) program, 21t, 26, P 30, 32t Paris Declaration on Aid Effectiveness, 167n2 PEFA (Public Expenditure and Financial Accountability) program, 21t, 26, R 30, 32t Renzio, Paolo de, 9 performance information integrated with reporting systems, integrating, 4­5, 15, budget documents, 58, 60t 41­52. See also under specific policy processes countries, e.g. Burkina Faso core policy-making processes in higher- APRs, 43­44 income reforming countries, 156 building blocks for, 46­47 links from budget up rather than from decision-making processes, targeting PRS down, 56­57 reporting to, 57­58, 59t sector-level, 57 existing links between systems, 44­45 targeting reporting to decision-making incentives, internal and external, 47­51 processes, 57­58, 59t building blocks and, 46 poor, spending in favor of. See pro-poor supplying information, 47­50, 48f, spending 49b, 50b poverty reduction strategies (PRSs). See also using information to improve integrating PRSs and budgets performance, 50­51, 51b introduction and uses of, 11­13 nature of reporting systems, 41­45 ownership of, 19­20 budgets, 42 policy links from budget up rather than PRSs, 43­44 from PRS down, 56­57 regular administrative reporting from review of approach (2005), 11­12 implementing institutions, 46 Poverty Reduction Strategy Papers (PRSPs) sectoral monitoring and reporting, 46 budget-friendly formulation of, 61 surveys and statistics, 46 fragmented processes and institutions, Rwanda, 1, 14, 123­131 problem of, 55 AAP (annual action plan), 49­51, 50b, future agenda, role in, 6, 7 125, 127, 129­130 linking PRSs and budgets, 34, 37 APRs, 127, 128, 129, 130, 164 Parliamentary approval, general lack CSOs in, 128 of, 20 decentralization in, 127 performance indicators drawn from, 25 domestic accountability in PRS and reporting systems, integrating, 46, 50 budget processes, 2, 24, 128­129 principal-agent model, 167­168n1 donor agencies in, 124, 125, 128­129 Priority Action Programs, 25 lessons learned, 129­131 priority sectors or programs, 30, 31b, 55, linking PRS and budget in, 124­127, 161­162 125t pro-poor spending, 161­162 MTEF, 126, 129, 130 in Burkina Faso, 161 other countries, lessons for, 130­131 in Malawi, 97­98, 161 ownership, 124, 128­129 182 Index Rwanda (continued) ownership, 137­139 priority sectors and pro-poor spending, priority sectors and pro-poor spending, 161­162 162 reporting systems, integrating, 5, 43, 44, PRS revision (Mkukuta), 11, 135­139 47­51, 50b, 127­128 regular meetings as bureaucratic in, Vision 2020, 124 170n1 reporting systems, integrating, 43, 44, 47, S 133, 136­137, 169n4 SBAS (Strategic Budget Allocation sector-wide approach (SWAp) mechanisms System), 30, 31b, 34, 135, 136, 137 in Albania, 67 technical reforms, inadequacy of, 55 in Mozambique, 115, 118 in Uganda, 143­144 reporting systems, integrating, 5, 46, U 48­50, 49b Uganda, 1, 14, 141­150 study countries benefiting from, 14 APRs, 165 sectors domestic accountability in PRS and developing sector-level policy processes, budget processes, 2, 20, 21t, 22b, 23, importance of, 57 24, 26, 147, 168n5 higher-income reforming countries, donor agencies, role of, 142 lessons from, 157 lessons learned, 147­150 Senegal, 11 linking PRS and budget in, 3, 4, 30, 32t, simplicity, advantages of, 53, 58­61 33b, 34, 142­145, 143f South Africa, 1, 14, 153­154. See also higher- MTEF in, 30, 32t, 33b, 34, 143, income reforming countries, lessons 144, 145 from drivers of reform process Open Budget Index, 168n5 in, 158 other countries, lessons for, 149­150 linking PRS and budget, 35, 36, 156 PEAP (Poverty Eradication Action Plan), strategic phase in budget, introducing, 57 33b, 43, 51b, 141­149 SWAp. See sector-wide approach (SWAp) planning and budgeting instruments in, mechanisms 32, 33b, 34 PRS revision, 11, 142­143 T reporting systems, integrating, 42b, 43, Tanzania, 1, 14, 133­140 44, 45f, 47­49, 50, 51b, 145­146 APRs, 137, 164­165 simpler approach to reform in, 53 CSOs, 168n4 SWAp mechanisms in, 143­144 decentralization in, 134 domestic accountability in PRS and V budget processes, 2, 20, 22­23, 22b, Vietnam, 11 24, 26, 137­139, 168n4­5, 169n8 donor agencies in, 137, 138­139 linking PRS and budget in, 3, 4, 30, 31b, W 34, 35, 134­136 Washington Consensus, 114 MTEF, 134, 136, 139 Wilhelm, Vera A., 9 Open Budget Index, 168n5 Williamson, Tim, 9 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. · 6 trees The Office of the Publisher has chosen to · 4 million BTUs of total print Minding the Gaps: Integrating Poverty energy Reduction Strategies and Budgets for · 488 lbs. of CO2 equiva- Domestic Accountability on recycled paper lent greenhouse gases with 30 percent postconsumer fiber in accor- · 2,027 gallons of dance with the recommended standards for wastewater paper usage set by the Green Press Initiative, · 260 lbs. of solid waste a nonprofit program supporting publishers in using fiber that is not sourced from endan- gered forests. For more information, visit www.greenpressinitiative.org. Minding the Gaps By integrating their poverty reduction strategies (PRSs), national budgets, and the corresponding re- porting processes, low-income countries can strengthen domestic accountability and the implementa- tion of pro-poor policies. Minding the Gaps, based on nine low-income-country case studies and a review of relevant experience in four higher-income countries, offers practical insights for donors and national governments on how to strengthen the links between PRSs and budgets. PRS countries' efforts to integrate policy with budgeting processes have often had limited effect. Their policy making, planning, and budgeting are often embedded in fragmented processes and institutions. Going beyond mainly technical fixes that have been commonly used to address this fragmentation, this study frames domestic accountability in terms of ownership and incentive structures. Experience counsels the use of a simple approach that is not too ambitious. This approach should be centrally led and make use of existing systems while gradually improving them. It should build support from within and foster incentives for integration, for example by better linking PRS and budget report- ing to actual decision-making processes. Also, simple budget reforms can significantly improve the budget's responsiveness to policies. Structuring a poverty reduction strategy paper in a more budget- friendly manner can facilitate the interface with the budget by involving sector agencies more closely in elaborating policy priorities and establishing resource implications. It can also expand ownership and boost incentives for integration of a greater number of stakeholders, thereby strengthening domestic accountability. "This timely publication offers practical advice for donors and governments on translating poverty reduction strategies into budget policies. Drawing on country experiences, it shows how development outcomes in low-income countries can be improved through processes that explicitly consider poverty impacts when governments decide budget allocations." Allen Schick Distinguished University Professor, University of Maryland "All too often, the process of producing a country's budget is treated as an unimportant technical by-product of the larger political process, generally unrelated to the promotion of economic growth. Minding the Gaps: Integrating Poverty Reduc- tion Strategies and Budgets for Domestic Accountability addresses this misconception, concentrating on the importance of the relationship between poverty reduction strategies in developing countries and a budget process that is account- able, transparent, and comprehensive yet simple and understandable. The emphasis Minding the Gaps devotes to the application of basic budget principles is particularly appropriate for budget and policy officials before they begin to pursue more complex performance budgeting systems." Barry Anderson Head of the Budgeting and Public Expenditures Division, OECD ISBN 978-0-8213-7205-0 THE WORLD BANK SKU 17205