Report No. 56716-BI BURUNDI Public Expenditure Review Improving the Efficiency of Public Investment October 1, 2010 PREM 2 Africa Region Document of the World Bank CURRENCY EQUIVALENTS (Exchange Rate Effective as of September 2010) Currency Unit = Burundi Franc US$1.00 = BIF 1,230 WEIGHTS AND MEASURES Metric System GOVERNMENT FISCAL YEAR January 1 – December 31 ABBREVIATIONS AND ACRONYMS ADC Alliance Démocratique pour le Changement AfDB African Development Bank BIF Burundian Francs BINUB United Nations Integrated Office in Burundi BRB Bank of the Republic of Burundi BRA Burundi Revenue Authority CAS Country Assistance Strategy CDC Central Development Committee CEM Country Economic Memorandum Commission Eléctorale Nationale Indépendante CENI (Independent National Electoral Commission) CET Common External Tariff CTE Comité Technique d’Exécution CNCA National Assistance Coordination Committee Conseil National de Défense de la Démocratie- Front de Défense de la CNDD-FDD Démocratie DeMPA Debt Management Performance Assessment DPO Development Policy Operation DRR Demobilization, Reinsertion and Reintegration Program EAC East African Community EDRRP Emergency Demobilization and Regional Reintegration Project EMSP Economic Management Support Project EU European Union FNL- PALIPEHUTU Front National de Libération-Parti pour la Libération du peuple Hutu FCE Fonds Commun de l’Education FDI Foreign Direct Investment FY Fiscal Year i GDP Gross Domestic Product GOVERNMENT Government of Burundi HIPC Highly Indebted Poor Countries IDA International Development Association IMF International Monetary Fund ICOR Incremental Capital Output Ratio JAVC Joint Agreement of Verification and Control Mechanism MDG Millennium Development Goals MDRI Multi-Donor Relief Initiative MDRP Multi-Country Demobilization and Reintegration Program MOF Ministry of Finance MOP Ministry of Planning MTEF Medium-Term Expenditure Framework NEPAD New Partnership for Africa’s Development NGO Non-Governmental Organization OTB Ordonnateur Trésorier du Burundi Projet d’Appui à la Reconstruction du Système Educatif (Education Support PARSEB Project) PBC Peace Building Commission PEFA Public Expenditure Financial Accountability PEMFAR Public Expenditure Management and Financial Accountability Review PER Public Expenditure Review PIM Public Investment Management PFM Public Finance Management PIP Public Investment Program Plan du Secteur de l’Education PSDEF (Sector Plan for the Development of Education and Training) PRS Poverty Reduction Strategy PRSP Poverty Reduction Strategy Paper PTBA Plan de Travail et Budget Annuel PTF Partenaires Techniques et Financiers SIGEFI (Integrated Financial Management Information System) TCTHE Technical Committee for Tracking HIPC Expenditures TFP Technical and Financial Partner Tableau des Opérations Financières de l’Etat TOFE (Table of Government’s Financial Operations) TSA Transitional Subsistence Allowance ii UK/DFID United Kingdom/Department for International Development UN United Nations UNOB United Nations Peacekeeping Operation for Burundi USA United States of America USD United States Dollar VAT Value-Added Tax Vice President: Obiageli K. Ezekwelisi (AFRVP) Country Director: John Murray McIntire (AFCE1) Country Manager: Mercy Tembon (AFMBI) Acting Sector Director: Jan Walliser (AFTPM) Sector Manager: Kathie Krumm (AFTP2) Task Team Leader: Eric Mabushi (AFTP2) iii TABLE OF CONTENTS ACKNOWLEDGEMENTS...................................................................................................................................... VI EXECUTIVE SUMMARY......................................................................................................................................VII CHAPTER 1 . THE RATIONALE FOR A PUBLIC EXPENDITURE REVIEW IN BURUNDI ...................... 1 A. INTRODUCTION .............................................................................................................................................. 1 B. OBJECTIVE AND JUSTIFICATION FOR THE PUBLIC INVESTMENT REVIEW .................................................. 2 Extension of the 2008 PEMFAR Exercise...................................................................................................... 2 Review of Investment Spending ..................................................................................................................... 2 Annual Review Exercises ............................................................................................................................... 3 Capacity Strengthening................................................................................................................................... 3 C. STRUCTURE AND SCOPE OF THE PUBLIC INVESTMENT REVIEW .................................................................. 3 CHAPTER 2 . CONTEXT AND EVOLUTION OF PUBLIC FINANCE MANAGEMENT IN BURUNDI .... 5 A. INTRODUCTION .............................................................................................................................................. 5 B. SOCIAL AND POLITICAL CONTEXT................................................................................................................ 5 C. MACROECONOMIC AND FISCAL CONTEXT AND RECENT DEVELOPMENTS ................................................. 6 Macroeconomic Context and Recent Developments ...................................................................................... 6 D. RECENT DEVELOPMENTS IN PUBLIC FINANCE MANAGEMENT ..................................................................... 9 E. CONCLUSION ................................................................................................................................................ 10 CHAPTER 3 . PUBLIC INVESTMENT ANALYSIS ........................................................................................... 11 A. EVOLUTION OF PUBLIC INVESTMENT SPENDING........................................................................................ 11 Historical Overview...................................................................................................................................... 11 Overall Execution Rates ............................................................................................................................... 12 HIPC Expenditures ....................................................................................................................................... 15 B. PLANNING AND BUDGETING PROCESS FOR PUBLIC INVESTMENTS ........................................................... 17 Legal Framework for Budget Preparation .................................................................................................... 17 Budget preparation process and institutional responsibilities ....................................................................... 18 Project Viability............................................................................................................................................ 19 The Public Investment Program ................................................................................................................... 20 C. IMPLEMENTATION AND MONITORING OF PUBLIC INVESTMENTS.............................................................. 22 Key Obstacles ............................................................................................................................................... 22 D. BURUNDI PUBLIC EXPENDITURE ISSUES ..................................................................................................... 25 CHAPTER 4 . PUBLIC INVESTMENT PROSPECTS ........................................................................................ 27 A. INTRODUCTION ............................................................................................................................................ 27 B. MEDIUM-TERM EXPENDITURE FRAMEWORK AND FUNDAMENTALS FOR IMPROVING THE MANAGEMENT AND EFFECTIVENESS OF PUBLIC INVESTMENT ..................................................................................................... 30 C. INCREASING THE LEVEL AND EFFICIENCY OF PUBLIC INVESTMENT AND THE PROSPECTS FOR POVERTY REDUCTION ............................................................................................................................................................ 31 Medium Term Expenditure Framework: Importance of Public Capital Spending ...................................... 32 Simulating an Improved Efficiency and the Level of Public Capital Spending ........................................... 33 iv D. CONCLUDING REMARKS .............................................................................................................................. 34 ANNEXES ................................................................................................................................................................. 35 Annex 1. Economic Trends, 2005-2009 ....................................................................................................... 35 Annex 2. Burundi Foreign-Financed Projects, 2007-2010 (US dollars ........................................................ 36 Annex3. Execution Rates and Breakdown of Public Expenditures in the National Budget, 2006-2009 ..... 37 Annex 4: Public investment in Burundi-Evolution of credits by source of financing .................................. 38 Annex 5: HIPC and FCE cases ..................................................................................................................... 39 List of Figures: Figure 2.1. GDP per Capita and GDP per Capita Growth, 1962-2008 ......................................................... Figure 3.1. Burundi Capital Expenditure: Domestic and Foreign Financed 2001-2009 (% of GDP)......... Figure 3.2. Burundi’s Investment Budget: Voted vs. Executed, 2005-2008 (billions BIF) ........................ Figure 3.3. Execution Rates of Investment Budget by Source of Finance, 2005-2008 (%)........................ Figure 3.4. Execution Rates by Sectors (%)................................................................................................ Figure 3.5. Monthly Execution Rates of HIPC Investment Projects, 2008-2009 ....................................... List of Tables: Table 3.1. Execution Rates and Breakdown of Expenditures in the Burundi’s Investment Budget ........................... Table 3.2. PIP Selection Criterion in Burundi ............................................................................................................ Table 3.3. Compliance of PFM with PEFA Criteria on Public Investment ................................................................ Table 3.4. Public Investment (% of GDP) .................................................................................................................. Table 4.1. Economic Distribution of Public Expenditures, 2005-2009....................................................................... Table 4.2. Comparison of Investment Volume and Efficiency in EAC, 2005-2009................................................... Table 4.3. Baseline Scenario—Medium-Term Expenditure Framework, 2007–2015 ................................................ Table 4.4. Impact of Improved Efficiency of Public Investment, 2009-2011............................................................. List of Boxes: Box 2.1. Macroeconomic context and recent developments ......................................................................................... Box 4.1. Definition and Use of the Incremental Capital Output Ratio ....................................................................... Box 4.2. Schematic Description of the Functioning of the Burundi MTEF................................................................ v ACKNOWLEDGEMENTS This Public Expenditure Review (PER) is a product of synchronized and tremendous efforts between World Bank Staff and many institutions in Burundi, including the government, the private sector, the civil society organizations, academia and bilateral and multilateral donors. The team would like to thank her Excellency Madame Clotilde Nizigama (Minister of Finance) and the government team for the excellent collaboration throughout the preparation of the PER report. The report was completed under the overall leadership of Jan Walliser (Sector Manager, AFTP 3) and Kathie Krumm (Sector Manager, AFTP2). The team received excellent guidance from Eric Bell (Lead Economist, AFTP3) and Jos Verbeek and Paolo Zacchia (Lead Economists, AFTP2) who provided invaluable inputs and quality control of the report. The team received also strong support and leadership from John Murray McIntire (Country Director) and Mercy Miyang Tembon (Country Manager, Burundi). The main author and task team leader of the report is Eric Mabushi (Resident Economist for Burundi, AFTP2). The PER core team includes Jean Pascal Nganou (Senior economist, AFTP2), Aurelien Beko (Poverty Economist (AFTP2), Ali Zafar (Consultant, AFTP2) and Gerard Muringa (Consultant, AFMBI). The Bank team worked closely with a national PER committee composed of: Donatien Bwabo, Jacques Gasuhuke, Jacques Ngendakumana, Pasteur Mpawenimana, Reverien Ndihokubwayo, Gratien Ninteretse, Pierre Niyubahwe, Bonaventure Sota, Leonce Sinzinkayo, Prosper Girukwishaka, Rwasamanga Ildephonse et Isidore Sindayikengera, Jean Claude Nsabimana, Ernest Nzitonda and MTEF team, Therence Ntabangana, Fabiola Nkundizanye, Jean Pierre Nguenang, and Alexis Bizimungu. The report benefited immensely from discussions with Renaud Seligmann (Sr.FMS specialist, AFTFM) and peer reviewers Martin Brownbridge, Chukwuma Obidegwu and Lev Freinkman. The team greatly appreciated the quality of information provided by multilateral and bilateral donors in Burundi. The team would also like to extend its gratitude to Judite Fernandes and Arlette Sourou (Program Assistants, AFTP2) and Mariama Daifour Ba (Program Assistant, AFTP3), who processed the document and Aurore simbananiye (Program Assistant, AFMBI) for logistic support. vi EXECUTIVE SUMMARY This report reviews the state of public expenditure in Burundi, with a focus on public investment, and recommends specific actions that need to be taken to stimulate growth and reduce poverty. Overall, the analysis indicates that poor performance in public expenditure and investment management are due to a number of factors. First, there is a lack of coordination among national institutions, and between those institutions and the donor community, which weakens the project selection process, implementation, and the integrity of procurement. Second, monitoring capacity is constrained by the lack of information about execution rates, and by inadequate audit and reporting mechanisms. Third, project performance is affected by disbursement delays and the volatility of external aid. Finally, in recent years, the significant increase in development spending has been mostly channeled toward the social sectors, with very limited public investment related to productive economic sectors such as infrastructure. Planning 1. The planning process for public investment is weak and does not comply with best practices for similar countries. The study finds a serious lack of compliance with international best practices related to public investment management (PIM), especially with regard to strategic planning and the preliminary project selection process in sectoral structures and ministries. Although the core ministries have clear and separate mandates in the project preparation and selection process, in practice public investment program (PIP) planning is weak. The roles of different stakeholders are not well defined; the budget process does not ensure that the best mix of demonstrably good projects is funded with the limited resources available; and the line ministries lack formal guidelines on the technical aspects of project appraisal. The integrity of the investment planning process is also undermined the lack of independent peer review. As a result, project selection is based more on the availability of financing than on feasibility. Many projects are not fully designed and have no credible staffing, procurement, or implementation plan. Recommendation 1. Establish a Central Development Committee (CDC) within the Ministry of Planning to ensure rigorous selection, track project quality and coherence with strategic planning tools, including medium-term expenditure frameworks (MTEFs). The CDC should be organized around three planning units in charge, respectively, of (i) sectoral planning; (ii) procurement (also see recommendation 5); and (iii) monitoring and evaluation (also see recommendation 6). 2. The Government needs to expand its fiscal space by finding a soft balance between current expenditures related to the post-conflict situation and investment expenditures critical for accelerated growth. Development spending, including public investment, has been increasing and now accounts for than 15 percent of GDP. However, high defense and security spending and the high wage bill remain key obstacles to reallocating public resources toward priority economic and social sectors. In addition, the unpredictability of aid flows has exacerbated the spending bias toward vii government consumption and away from investments in production-oriented sectors, as recently reported in the Country Economic Memorandum (CEM).1 Recommendation 2. Improve donor coordination and internal/external revenue mobilization to facilitate the allocation of aid toward strategic priorities. 3. Burundi’s public investment is the least efficient in the East African Community (EAC). Budget execution rates are low, although they are improving for domestically financed investments. A larger than average incremental capital output ratio (ICOR), at 5 percent, is one indicator; lower road and energy penetration rates, at 2 percent, are another. In general, the crowding in effect of public investment is less than it should be and less than elsewhere in the region. Improved efficiency and targeted specific productive investments could have a strong impact on growth. As noted in the CEM, the Government’s top priorities for accelerating growth and reducing poverty are energy, public works, and agriculture, for which needed investments will cost an estimated US$860 million over the 2010-15 timeframe. Simulations conducted for this PER and for the CEM show that reforms in three specific areas would significantly increase GDP growth and reduce the incidence of poverty: (i) improving execution rates for public capital expenditures; (ii) improving efficiency in the productive agriculture sector (food and export crops, livestock); and (iii) increasing public capital expenditure in the productive sectors and in infrastructure (maintenance of roads and small hydro plants around Bujumbura, construction). The horticulture and mining sectors also show high potential to increase growth, but this growth will rely primarily on an enabling policy environment for private investors and producers, and would not necessarily make a major claim on limited public investment resources. Recommendation 3. Based on these findings, ensure that the (identified) highest return and highest priority growth-oriented investments in the infrastructure and energy sectors and in agriculture support services are included in the next PIP, as well as in the planning framework for the second Poverty Reduction Strategy Paper (PRSPII) . 4. The current PIP is not comprehensive and not aligned with Government policies or strategic priorities, due to lack of coordination between the Government and external donors. Estimates of budget expenditures do not yet cover donor-funded activities, and there is lack of accurate information regarding expenditures due to poor reporting standards. As a result, basic indicators such as execution rates of donor-funded projects are rather poorly estimated. Donor-financed investments are neither properly tracked by line ministries nor fully included in the PIP, and do not appear in the Budget Law (most of them are in an annex to the Law). About half the investments in the Budget Law are funded by HIPC funds, and most others are essentially covered by external resources (through budget support funds). The remainder are covered by counterpart funds of donor-financed projects, which are classified in the Budget Law as investments but in reality represent projects’ current expenditures. 1 A Country Economic Memorandum for Burundi was prepared in 2010. viii Recommendation 4. Establish a Financial Unit (FU) corresponding with the CDC, in charge of centralizing financial and budgetary data on foreign-financed expenditures from the Finance Ministry, the line ministries, and the National Aid Coordination Committee (CNCA).. Enforce the current partnership (including sharing information campaigns and communication on donors ‘compliance to Paris Declaration). 5. Both procurement issues and weak implementation capacity have undermined public investment efficiency. Procurement issues are at the heart of low execution rates; specifically, (i) inflated and inconsistent estimates for purchases of certain goods and services; and (ii) corruption and fraudulent practices. In addition, procurement cells in line ministries lack the capacity to deal appropriately with procurement requirements. While the introduction of the new Procurement Code has been an especially meaningful reform, the transition to the new system (including training for procurement staff) has caused delays, leading most donors to continue to use own procedures. At the same time, the line ministries have weak technical capacity to supervise and monitor public works. Expenditure tracking has been significantly improved under the special HIPC regime, but there still are no implementation plans or specific requests for information on the performance of ongoing investment projects from a structured institution. . In addition, there is total absence of a monitoring and evaluation (M&E) system to track the performance of public investments. Recommendation 5. Ensure speedy implementation of new Procurement Law, by: (i) extending capacity building and training programs in all the remaining ministries; (ii) creating a unit within the CDC to oversee procurement in all ministries; (iii) putting quality control mechanisms in place in the procurement units of each ministry; and (iv) harmonizing Government and donor procedures over time. Recommendation 6. Create an M&E unit within the CDC to ensure that line ministries are monitoring and assessing the physical execution of projects and adhering to standards of information, oversight, audit, and reporting, with an initial focus on large projects and programs. Capacity building 6. Weak capacity is a cross-cutting issue, since it continues to hamper every aspect of public expenditure. A weak project selection process, implementation issues, the absence of strong project guidelines and M&E mechanisms; and limited procurement skills are major obstacles to an efficient public investment program. Recommendation 7. Articulate a phased capacity building program in public investment management, with an initial focus on CDC’s capacity to assess the adequacy of project appraisal and selection processes in the largest programs. One or two sectoral ministries could also be identified for upgrading of project-related skills, including procurement skills. Box 1. Burundi: Main Issues of Public Expenditure Management, and Progress in Public Investment Management ix I. Current Status Strengths • Existence of a strategic framework (including gradual implementation of MTEF). • Existence of a framework for the public investment program. Weaknesses • Fiscal space. About 50 percent of Burundi’s investment budget is financed from external sources, with the remaining 50 percent being financed essentially by HIPC funds. While public investment has been increasing and has now reached more than 15 percent of GDP, aid volatility and the high wage bill threaten the sustainability of the fiscal space and investment planning in the medium and long term. • Execution rates. The execution rates of the investment budget have been low compared to the recurrent budget. Although these rates have been improving recently for domestically financed investments, they are still negatively affected by: (i) lack of coordination; (ii) a weak project selection process; (iii) weak implementation; (iv) the absence of strong monitoring; and procurement issues. • Coordination. Lack of coordination between donors and the Government has had a negative effect on public investment; and donor-financed investments are not properly tracked in the PIP. • Project selection process. There is a lack of rigor in project selection; while the PIP has guidelines for selected feasible projects, in practice projects are selected for seemingly arbitrary reasons • Implementation. Technical capacities and M&E in the line ministries (including oversight, audit, and reporting) need to be institutionalized and strengthened. • Procurement. Procurement issues include corruption, inflated estimates for purchases, and slow implementation of the Procurement Code, resulting in most donor-financed investments still using donor procurement procedures. How to address weak points: II. How to address weak points • As shown above, the challenges facing Burundi’s public investment program are similar to those faced by other post-conflict and donor-dependent countries. The report draws on the reform experiences of these other countries in shaping feasible and phased recommendations (see details in Table 1). Planning • Establish a Central Development Committee (CDC) within the Ministry of Planning that should be organized around three planning units in charge, respectively, of (i) sectoral planning; (ii) procurement; and (iii) monitoring and evaluation • Improve donor coordination and internal/external revenue mobilization to facilitate the allocation of aid toward strategic priorities • Ensure that the (identified) highest return and highest priority growth-oriented investments in the infrastructure and energy sectors and in agriculture support services are included in the next PIP, as well as in the planning framework for the second Poverty Reduction Strategy Paper (PRSPII). • Establish a Financial Unit (FU) in the Ministry of Finance corresponding with the CDC, in charge of x centralizing financial and budgetary data on foreign-financed expenditures from the Finance Ministry, the line ministries, and the National Aid Coordination Committee (CNCA )and enforce the current partnership Capacity building • Articulate a phased capacity building program in PFM that start with two sectoral pilots. xi Table 1. Matrix of Policy Actions Responsible Expected Priority Measure implementing Timeline Performance intermediate measure results order agency (1) Continue the current partnership between the Government and development partners. (2) Ensure a speedier flow of project information Adoption of a between the Government and donors. Improve the budget (3) Establish a Financial Unit (FU) within the aid Ministry of I incorporating 2012 Finance Ministry to centralize financial and management Finance, 1 donor aid flows budgetary data on foreign-financed expenditures, regime external partners and collaborate with the CDC, line ministries, and CNCA. (4) Hold annual information-sharing workshops involving donors and central authorities as part of the budget preparation process. (5) Link the continuation of partners’ activities on the production of annual budgeted working plans and implementation reports. (6) Publish, on the Government’s website, a list of partners that comply with the Paris Declaration criteria. (1) Conduct more rigorous analysis of projects based on rates of return and Government priorities. (2) Ensure that line ministries prepare proper feasibility studies. (3) Ensure coherence of projects with line ministry Improve the Rigorous strategies and the Government’s overall growth Ministry of strategy (PRSP). II quality of the project 2012 1 Planning (4) Coordinate between donors and line ministries PIP selection to ensure that the PIP database includes all projects. (5) Create a Central Development Committee (CDC) within the Ministry of Planning to rigorously track project quality and ensure better coherence with MTEFs. Ensure speedy implementation of the new Procurement Law, by: (1) Providing capacity building and trainings in all Higher ministries. Improve the Ministry of execution rate 2010- (2) Creating small units in charge of procurement III procurement 1 Finance of donor 2012 in each ministry. system projects (3) Preparing a manual of quality control procedures. (4) Promoting harmonization between Government and donor procedures. Track execution rates by Strengthen capacity in public finance management Unified 2010- IV monitoring Treasury 2 to improve Treasury’s management of aid flows accounts 2012 accounts in and disbursements. commercial banks Monitoring of Database of 2010- Create M&E units in line ministries to monitor and V investment Line ministries projects and 2 2012 assess the physical execution of projects projects impacts i Chapter 1. The Rationale FOR a Public Expenditure Review in Burundi A. INTRODUCTION 1.1 This report was prepared in response to Government of Burundi’s request for World Bank assistance in reviewing specific aspects of public expenditure on an annual basis. The analysis was conducted in collaboration with the ministries of Planning and Finance, representatives from the line ministries, and technical and financial partners (TFPs) in the donor community. It aims to strengthen the effectiveness of the budget as a tool for implementing the Poverty Reduction Strategy (PRS), through concrete suggestions to improve the efficiency of public investment spending. 1.2 This report is the first of a series of three Public Expenditure Reviews to cover the period 2010-12. After a PER undertaken in 2001, the Government of Burundi and the Bank conducted a Public Expenditure Management and Financial Accountability Review (PEMFAR) that gave an overall picture of the strengths and weaknesses of public finance management (PFM) and made key recommendations. One recommendation was that PERs be conducted annually, as stipulated in the country’s Poverty Reduction Strategy Paper (PRSP, 2006). The World Bank responded by planning the series of PERs as part of Country Assistance Strategy (CAS, 2008). 1.3 In Burundi, a series of factors still undermine the efficiency of the whole PFM system in general, and public investment expenditure in particular. Despite considerable progress, and beyond general issues of PFM, public investment is still characterized by low execution rates; weak record and traceability of donor projects; weak methodology in selecting projects, and procurement issues. 1.4 The study is structured in four chapters. First, the study presents the justification for focusing the present review on public investment, and the scope of the work. Second, we describe the macroeconomic and fiscal contexts, assess progress made in by PFM system, and present the main elements of its medium term strategy. The third chapter provides a public investment analysis based on determinants of the execution rate, including planning and budgeting processes and implementation and monitoring issues. The fourth chapter presents scenarios for increasing the level and efficiency of growth- oriented public investment and their likely impacts on poverty reduction, using the newly introduced MTEF as a central tool for planning and prioritizing investment. The final section presents key findings and recommendations. 1.5 The recommendations are based on public investment management (PIM) methodology, which includes a set of eight criteria for successful implementation of public expenditure. The analysis assesses the level of compliance of the Burundi PIM process with this methodology, and suggests areas for improvement. 1 B. OBJECTIVE AND JUSTIFICATION FOR THE PUBLIC INVESTMENT REVIEW 1.6 With the Government’s recent shift of emphasis to economic growth as the main strategy to reduce poverty, the management of investment expenditures has become a high priority. It is the first of a multi-year exercise that will focus each year on a different investment theme. This review provides the framework to guide future reviews. Its aim is to strengthen institutional efficiency by: (i) improving knowledge and strengthening the capacity of the principal ministries involved in investment spending; (ii) strengthening reforms to improve the basis for public sector performance; (iii) strengthening the Bank’s assistance and programs; and (iv) improving donor coordination and harmonization. Extension of the 2008 PEMFAR Exercise 1.7 The Government has requested Bank assistance to review specific aspects of public expenditure on an annual basis, to strengthen the effectiveness of the budget as a tool of PRSP implementation. Such annual PERs were recommended both in the country’s Poverty Reduction Strategy Paper (PRSP, 2006) and the 2008 Public Expenditure Management and Financial Accountability Report (PEMFAR, 2008). The latter report provides a comprehensive overview of the challenges and a set of recommendations for public expenditure management, and also summarizes the key medium-term macro-fiscal challenges. Building on the PEMFAR, this PER tackles issues pertaining to the investment budget process. 1.8 As described in the 2008 Burundi PEMFAR, there are many issues regarding public expenditures in Burundi. These include (i) a weak public expenditure management system; (ii) the budget structure and allocations; (iii) efficiency issues; (iv) fiscal deficits, and (v) dependency on external aid and the volatility of such assistance. Progress has been made in several areas over recent years, and these reforms need to be consolidated. Major recent achievements include: (i) the launching of a consistent process for setting a legal and regulatory framework for PFM (including adoption of a new Organic Law of the Public Finances); (ii) adoption of a Public Procurement Code and a law on the sale of public property; (iii) the progressive introduction of Medium Term Expenditure Frameworks in budget planning; and (iv) adoption of a new PFM strategy. 1.9 The PFM strategy includes six objectives: (i) adoption and implementation of a legislative and regulatory framework for PFM; (ii) effective mobilization of internal and external resources to finance expenditure, while respecting fiscal sustainability; (iii) transparent and efficient management of public resources; (iv) creation of an efficient control system consistent with EAC and international standards; (v) comprehensive, reliable, and regular public finance data reporting; and (vi) strengthening the institutional capacity of the Ministry of Finance and associated entities. Review of Investment Spending 1.10 Past PER and PFM analytical work has served to advance the dialogue on improving public finance management. Prior to the recent PEMFAR, the last PER was carried out in 2001. This was not a 2 core World Bank PER but a technical assistance (TA) PER, the conclusions of which informed the design of the Public Finance Management (PFM) component of the Bank’s Economic Management Support Project (EMSP) in 2004. The 2008 PEMFAR exercise moved the dialogue forward by providing an overall picture of the status of PFM reforms, consolidating several donor reports, and focusing especially on the implementation of related recommendations and capacity building needs. These recommendations focused more on current expenditure issues and PFM processes, and provided the basis for reform programs supported by IMF and the Bank’s development program operations (DPOs). 1.11 Following on past work, this PER exercise provides concrete suggestions for improving the efficiency of public investment spending. While the authorities recognize the importance of investment for economic growth and poverty reduction, due to limited fiscal space the country relies primarily on foreign assistance to support its growth and development strategy (PRSP), especially for public investment. Domestically financed capital expenditures represent a mere 4 percent of GDP, compared to more than 10 percent of GDP for foreign-financed capital spending in recent years (Figure 3.1). Because foreign assistance is very volatile, however, implementation of the growth and poverty reduction strategy remains difficult. Until recently, investment spending continued to be concentrated in the defense and security sectors, despite the current shift in development priorities towards growth promotion. The Government is determined to increase budgetary allocations toward capital spending and to improve the efficiency of that spending. Annual Review Exercises 1.12 The planned series of PERs aims to build the knowledge base on public investment management gradually and systematically, and to support a more consistent dialogue on PFM issues. 1.13 This analysis of the processes and institutions involved in the preparation and implementation of public investments will set the context for more detailed sectoral analysis of public investments in the following years. The follow-on PER for FY12 will analyze the preparation and implementation of public investments in the infrastructure sector (transport, energy, and telecom subsectors). The PER for FY13 will look at the preparation and implementation of public investments in two selected social sectors (health and education). These sectors have been identified by the Government as those that offer the best prospects for growth and poverty reduction. Capacity Strengthening 1.14 The PER exercise has been carried out in collaboration with the ministries of Planning and Finance. It examines the role of these core ministries, as well as the role of the line ministries in preparing their public investment programs and budget. The core ministries have identified a few technical staff to work closely with Bank staff on the analysis. Line ministries were also brought into the work whenever necessary. C. STRUCTURE AND SCOPE OF THE PUBLIC INVESTMENT REVIEW 1.15 The analytical section of the report consists of three chapters. Chapter 2 provides an overview of recent sociopolitical developments, the country’s macroeconomic and fiscal context, and recent 3 developments in PFM reforms. Chapter 3 discusses the institutional arrangements for implementing the PIP. It also analyzes the evolution of public investment spending in the recent years, the adequacy of the planning and budgeting methods, implementation issues, and the monitoring and evaluation system. Chapter 4 discusses how initiatives such as the MTEF might help to improve the management and effectiveness of public investments, and simulates the impacts of increased public investment spending on growth and poverty reduction. The final chapter summarizes key findings and recommendations. 1.16 Several other issues related to PFM are evoked but not addressed in this PER. These include issues related to specific sectors. Some of these issues/aspects are evoked in order to deepen the understanding of the analysis, but a full treatment of these issues was beyond the scope of the study. 4 Chapter 2. Context and Evolution of Public Finance Management in burundi A. INTRODUCTION 2.1 In the ten years since the Arusha Agreement, Burundi has experienced both great strides and major setbacks on the road to lasting peace. After 13 years of civil war, Burundi’s 39 politico- ethnic groups signed a comprehensive peace and reconciliation agreement in Arusha (Tanzania) in August 2000. This paved the way for the formation of a transitional government in 2002, a new constitution in February 2005, and presidential and parliamentary elections in August 2005. A year later, a cease-fire accord was reached with the last hold-out rebel faction, the Front National de Libération- Parti pour la Libération du peuple Hutu (FNL-PALIPEHUTU). However, the political and security situation deteriorated somewhat over 2007 and early 2008, leading to a long political deadlock in 2007 and armed skirmishes with government troops (with several deaths) in 2008. Political dialogue resumed in 2009 and led to democratic elections in 2010. 2.2 Since January 2004, the authorities have also been implementing an economic reform program supported by the World Bank and the IMF. As a result, the Government has made good progress in terms of macroeconomic stabilization, implementation of financial and structural reforms, promotion of investment and economic growth, and restoration of social services despite the difficult political, social, and economic environment. This progress led to some relative stability due to strict macroeconomic policies; a continuous recovery of credit to the economy and debt relief, and considerable improvements in the fiscal context. Nevertheless, the country remains vulnerable to debt unsustainability in the future, necessitating efforts to improve debt management. B. SOCIAL AND POLITICAL CONTEXT 2.3 Helped by the national Demobilization, Reinsertion, and Reintegration program (DRR, launched in 2004), the security situation has much improved but remains somewhat fragile. In June 2008, after pressure from the international community and regional powers (South Africa and Tanzania), discussion between the FNL-PALIPEHUTU and the Government led to a power-sharing agreement in December 2008. In January 2009, in addition to substantive progress made on the peace process,2 the disarmament, demobilization, and reintegration of ex-combatants effectively began. However, isolated cases of crimes and violence (including sexual violence against women and children, assassinations, and armed robbery) are still often reported. While the United Nations program will end in December 2010, the security issue will remain a critical priority for the new Government, which has asked the World Bank and other interested development partners to assist in preparing a new phase of DRR. Focusing on the 2 The two sides resolved two significant unresolved issues in the peace process: (i) removal of ethnically charged part of FNL’s name PALIPEHUTU, making the group eligible to register as a political party; and (ii) the freeing of FNL political prisoners and prisoners of war. Other important issues in the peace process, such as the reunification of FNL child soldiers with their families and the integration of some of the FNL forces and leadership into the Burundi armed forces and national institutions, were also addressed. 5 dismantling of the irregular forces,3 this new program is supported by the World Bank through an IDA grant and a single-purpose, multi-donor trust fund administered by the World Bank. 2.4 Presidential and legislative elections in 2010 resulted in a large victory for the ruling party. The electoral season began with local council elections in May 2010, in which the ruling party, Conseil National de Défense de la Démocratie- Front de Défense de la Démocratie (CNDD-FDD), carried 64 percent of the vote. After the election, the consortium of opposition parties, Democratic Alliance for Change (ADC), complained of a number of irregularities and announced a boycott of the presidential elections, leaving the incumbent president as the sole candidate. The National Independent Electoral Commission (CENI) and the European Union (EU) observers validated the results, recognizing the existence of some minor irregularities to be corrected with the presidential elections. They were held in June, and the incumbent and sole candidate, Pierre Nkurunziza, was elected for a five-year term with about 92.6 percent of the vote. 2.5 Burundi has had a new democratically elected Government since August 2010. The president has set up a Government composed of more than 40 percent of women, sending a strong signal of gender equity, far beyond the Arusha Peace Agreement provision of 30 percent. The Government’s first major activity will be to launch the PRSP 2 preparation process. The Government announced during the first Burundi Consultative Group meeting in October 2009 in Paris that the new PRSP will be growth- oriented. Thus, identifying investments for growth and channeling the right spending through the budget has become a critical axis of economic policy. In this context, the PER’s findings and recommendations are highly relevant. C. MACROECONOMIC AND FISCAL CONTEXT AND RECENT DEVELOPMENTS Macroeconomic Context and Recent Developments 2.6 After about 25 years of protracted decline in GDP per capita (Figure 1), Burundi’s economy is beginning to experience positive growth with the gradual return of stability. Burundi’s per capita income in 2007 was about one-ninth of the Sub-Saharan Africa (SSA) average, roughly on par with its 1975 income per capita level. It fell by almost 45 percent during the war, from a period high of US$156 in 1993 to a low of US$86 in 2003. Since that period, growth has rebounded somewhat, with GDP per capita of US$144 in 2008, but has not yet returned to its pre-war levels. This volatility can be explained by a number of factors, including external shocks (fuel and food prices) and the recent global economic slowdown, which dampened real GDP growth from about 4.5 percent in 2008 to 3.5 percent in 2009. 2.7 Nonetheless, since 2000, the country has progressively recovered from a decade of negative growth (average -1.8 percent) to a positive growth rate of 3 to 5 percent in 2006-08. However, this is far below the 8 percent rate needed over the next three years for the country to reach its pre-war level of GDP per capita by 2013. 3 The following activities are being financed under the new phase: (i) support for demobilization, reinsertion, and reintegration of up to 5,000 FNL combatants; (ii) delivery of reintegration support to 1,100 ex-combatants not yet served by the previous program, (iii) support for the DRR of up to 1,564 members of the FNL dissident group; and (iv) additional specialized support for especially vulnerable groups (children, handicapped, female combatants). 6 2.8 Economic growth has been Figure 2.1. GDP per Capita and GDP per Capita Growth, driven mainly by the agriculture 1962-2008 sector. Despite that domination, a nascent recovery in Burundi’s other sectors is beginning to contribute to growth4. Economic growth remains volatile due to dependence on the widely fluctuating agricultural sector, which constitutes about 46 percent of GDP. The volatility of Burundi’s agricultural performance is largely due to the international coffee market and variations in the coffee harvest (e.g., the 2005 drought limited real GDP growth to only 0.9 percent). Recent growth in the service sector, which accounts for 37 percent of GDP, is largely due to increases in public services and, to a lesser extent, in the transport and trade sub-sectors. Burundi’s industrial sector (about 17 percent of GDP) is dominated by agricultural processing (Box 2.1). 2.9 Prudent monetary and fiscal policies during 2006–08, led to a contained fiscal deficit despite the effects of the global financial crisis) from 10 percent in 2007. But the country remains vulnerable to debt unsustainability, needs to diversify its exports and to improve its business environment (Box 2.1). 4 The agriculture sector contributes for about 46 percent of GDP and provides about 90 percent of the population with income and employment. The service sector accounts for 37 percent of GDP, largely from the recent increase in public services and, to some extent, in the transport and trade sub-sectors. The industrial sector contributes about 17 percent of GDP. Agricultural processing dominates the country’s small industrial sector, and includes the production of beverages, cigarettes, consumer goods, textiles, and construction materials for the domestic, and to a limited extent, the regional market. In the 1990s, 7 Box 2.1. Macroeconomic context and recent developments • Recent developments in the domestic and international economies have created upward pressure on prices, but the authorities are attempting to maintain price stability through strict macroeconomic policies. From 8.3 percent in 2007, inflation rose to 24.4 percent (period average) in 2008, and declined to 10.7 percent in 2009. With the continuation of the reforms, the inflation rate should be maintained in the single digits in 2010 and beyond. • Although there were some slippages in 2009, prudent monetary and fiscal policies during 2006–08, led to a contained fiscal deficit. First, prudent monetary policy has meant that credit to the economy has continued to recover, rising by about 19 percent on average in 2008-2009 (despite the effects of the global financial crisis) from 10 percent in 2007. Second, stronger expenditure management, accompanied with increased domestic revenue and foreign aid inflows, helped to maintain the fiscal deficit (commitment basis before HIPC grants) to an average of 3.5 percent of GDP in 2007-08, from an average 4.5 percent of GDP in 2005- 06. Furthermore, despite the effects of the global economic and financial crisis, domestic revenue broadly stabilized in 2009 at 18.6 percent of GDP (from 18.5 percent in 2008), reflecting improved collection of income and indirect taxes. This level remains above the average of 17.3 percent of GDP for post-conflict countries. At the same time, total government expenditure declined by 6.7 percentage points of GDP between 2008 and 2009, due to the reduction in externally financed spending.5 As a result, overall fiscal outturn (commitment basis after grants, excluding HIPC grants) in 2009 worsened slightly to 5.2 percent of GDP, from 3.5 percent in 2007- 08. • Despite recent debt relief, the country remains vulnerable to debt unsustainability, necessitating efforts to improve debt management and diversify the economic base. Burundi reached its HIPC Completion Point (HIPC CP) in January 2009, thus reducing its public and publicly guaranteed (PPG) debt by more than 90 percent in net present value terms, and scheduled debt service by some US$30-40 million per year over the next 30 years. The debt forgiveness will provide much relief to the country’s balance of payments. Having reached the HIPC CP, Burundi was also granted further debt relief under the Multilateral Debt Relief Initiative (MDRI). • However, as shown in the most recent joint external Low-Income Country (LIC) Debt Sustainability Analysis (DSA)6, Burundi’s external public debt burden indicators are expected to remain high over the medium term recent debt relief (HIPC relief and MDRI assistance). The DSA strongly recommended: (i) accelerating structural reforms to increase and diversify the export base; and (ii) contracting new external/domestic debt in amounts and on terms that are mostly conducive to debt sustainability. A new Dept Management Performance Assessment (DeMPA) report set basic recommendations that would lay the groundwork for the preparation of a medium-term debt management strategy in a near future. • Lack of diversity in Burundi’s exports, and reliance on imported oil and capital goods, have led to consistent current account deficits. From 12.3 in 2008, the current account deficit worsened in 2009 (to 15.9 percent of GDP) mainly due to a 40 percent reduction in donor and a further worsening of net services. Burundi’s trade deficit reflects the country’s narrow range of exports that largely depend on the size and quality of the coffee and tea harvests and on fluctuant international price movements for both products. However, Burundi’s export earnings increased by more than 10 percent in two consecutive years (15 percent and 22 percent in 2008 and 2009, respectively). This was due to (i) the marginally higher prices of coffee in 2009, and by the record price of tea experienced during the same year; (ii) decreases in international food and fuel prices that contributed to a reduction of total imports from 29.4 percent of GDP in 2008 to 19.6 percent in 2009 (food and oil imports comprise one-fourth of total imports). • Foreign direct investment (FDI) remains low in Burundi, but is beginning to improve. In 2008, FDI increased from US$0.5 million in 2007 to US$13.6 million, as a result of renewed involvement of foreign firms in the banking and telecommunication sectors. However, it is estimated to have slightly declined to US$10.2 million in 2009, mainly due to the global economic downturn • Partly as a result of a surge in donor disbursement(HIPC and MDRI relief, and IMF’s SDR allocation, in particular for 2009,came in addition to traditional bilateral and multilateral aid), gross official reserves (in months of imports) have continuously increased during 2005–09. Gross international reserves are estimated to have increased to US$323.0 million in 2009 from US$266.6 million at end-2008, representing a steady 5.1 months of import cover for the two years 5 Several capital expenditures in the 2009 initial budget, which presented lower execution rates and that could not be executed within the remaining months of the fiscal year, were simply dropped during the budget revision. The budget support of many donors planned for the 2009 budget actually arrived during the following year. 6 The DSA was– prepared jointly between the Bank and the IMF in July 2010 in the context of the fourth review of the IMF’s ECF program. 8 D. RECENT DEVELOPMENTS IN PUBLIC FINANCE MANAGEMENT 2.10 Since 2005, the authorities have been implementing an ambitious set of reforms aimed at improving PFM. These reforms have included: (i) modernizing budget operations; (ii) ensuring the alignment of public expenditure with Government programs, particularly the PRSP; and (iii) improving transparency and accountability. To support this agenda, in May 2009 the Council of Ministers adopted a reinforced strategy and action plan as a holistic framework for PFM reforms. 2.11 The recently adopted PFM strategy, which builds on the findings of the 2008 PEMFAR, has begun to lead to improvements in PFM. Within a short period of time, important fiscal reforms have been introduced. These include (i) modernization of the tax system through adoption of the EAC’s Common External Tariff (CET) and the value-added tax (VAT) in July 2009; (ii) promulgation of the Law on the Burundi Revenue Authority (BRA), also in July 2009, which has been operational since the first half of 2010; (iii) the partial consolidation of cash management by closing off-budget accounts; and (vi) continuing with gradual implementation of the MTEFs. 2.12 Concerning budgetary planning specifically, Burundi has made progress in improving the planning of public finances over the past several years. Major achievements include: (i) introduction of an annual review of the PRSP and sector strategies, for the purpose of strategic planning; (ii) adoption of a new functional budget classification, to improving traceability; (iii) establishment of a central MTEF as well as 8 pilot sector MTEFs in the ministries of Public Health; Primary and Secondary Education; Public Works; Energy and Mines; Water, Environment, Urban and Regional Planning; Agriculture and Livestock; Public Security; National Defense and Veterans Affairs). At the same time, Burundi continues to struggle to maintain a high level of quality of public expenditure. Indeed, for effective implementation of the Poverty Reduction Strategy Paper (PRSP), Burundi needs to improve the efficiency of expenditures through better targeting of budgetary resources, both across sectors at the central level, and within each sector, according to the priorities of ministerial action plans. The central and sector MTEF tools should help to improve planning and targeting as they are implemented in other ministries and their use becomes a regular part of the budget preparation process.7 2.13 In addition, the overall structure of public expenditures has been significantly improved, to ensure alignment with the Government program and the MDGs. Actual public spending in priority economic and social sectors increased sharply, in terms of both share of total public expenditures and share of GDP. Although security expenditures declined significantly in terms of share of total public expenditures, they increased in terms of share of GDP until 2006 (8 percent), but eventually declined in 2007 and 2008 (to 7.8 and 7 percent, respectively). Increased public spending in health and education led to dramatic improvements in primary school enrollment, higher rates of utilization of existing health facilities, and slightly reduced maternal and child mortality rates. In 2008, spending in the social sectors represented an estimated 28.8 percent of total government expenditure (excluding debt service), and an estimated 33.7 percent in 2009. 7 The MTEF process in Burundi has been given much attention. In December 2009, the second vice-president called for a high-level workshop that personally involved ministers in charge of social and economic sectors, along with their senior staff. One of the workshop’s recommendations was that the MTEF tool should be applied to the whole planning budget process. 9 2.14 Efforts have also been made to improve the transparency of public finance management. Owing to the introduction of a computerized financial management information system (SIGEFI), the Government is able to produce regular and timely budget execution reports. The new Procurement Code became operational at the end of 2009. Furthermore, agreement on a new methodology to identify pro- poor expenditures (see Annex 5, Box 3) will help better monitor the poverty impact of public spending. Further, most existing extra-budgetary accounts have been closed or integrated into budget documents (so far, more than 250 accounts have been closed, and efforts are made to close the few remaining ones). A new Organic Public Finance Law and a new Procurement Code have modernized the budget framework and public procurement procedures. As a result, the Cour des Comptes (Audit Court) has been steadily strengthened in its capacities to carry out audits of government accounts. Two major corruption scandals in 2006 and 2007 showed that developing a culture of transparency in public sector management will require additional efforts on the part of the country’s leadership. In both cases, strong remedial measures undertaken by the authorities initially reassured the donor community on the Government’s commitment to fight corruption. To maintain credibility, however, the Government will also need to accelerate prosecutions in relation to both cases and recover funds owed to the State by petroleum companies. E. CONCLUSION 2.15 Despite a number of sensitive issues related to the post-conflict situation, political dialogue has progressed well in Burundi, allowing for progress of the Government’s economic reform agenda. Political stabilization led progressively to democratic elections in 2010; and a strong degree of macroeconomic stability has been achieved, despite the global economic crisis, because of the disciplined implementation of the World Bank/IMF-supported economic reform program. Real GDP growth has been respectable, averaging more than 4 percent per year since the war ended, while inflation has remained mostly in single digits. Post-conflict recovery has resulted in stronger fiscal and current account positions, and the PFM agenda has progressed well, although gains still need to be consolidated. The next chapter analyzes the public investment budget. 10 Chapter 3. Public Investment Analysis F. EVOLUTION OF PUBLIC INVESTMENT SPENDING Historical Overview 3.1 Public investment imploded during the period of war, but is now growing steadily to support reconstruction and development efforts. In the 1990s, public investment as a proportion of GDP has collapsed due to the long period of war in Burundi. But after the war, there was a collapse followed by a recovery. In comparison to many other countries, where the capital budget has averaged 5 percent of GDP in recent years, Burundi’s public investment budget has averaged above 10 percent of GDP and is on a projected upward path, partly due to donor initiatives under HIPC and the MDRI. This level of investment will have to be sustained in order to finance the infrastructure needed to help Burundi fully recover from the war. Fiscal policy remains broadly supportive of economic recovery and capital budget expansion. Capital expenditure is projected to reach more than 15 percent of GDP in 2010, compared to 5 percent a decade earlier (chapter 4). 3.2 About half of Burundi’s capital expenditure8 is financed by external resources. Between 2001 and 2009, external financing of capital expenditure increased rapidly (Figure 3.1). The HIPC initiative also gave the Government the fiscal space needed to increase capital expenditure in a systematic way by providing resources to substitute for a decline in domestic financing in the aftermath of the war. Overall, revenue (excluding grants) has reached close to 18.5 percent of GDP and is still projected upward. However, the domestic financing contribution to capital expenditure (excluding HIPC) is projected to stay below 5 percent of the total in the short and medium term (Annex 4). Grants and concessional money have continued to finance the largest part of capital expenditure, although there has been some volatility in inflows due to donor countries’ politics and processes. Figure 3.1. Burundi Capital Expenditure: Domestic and Foreign Financed 2001-2009 (% of GDP) Chart 1: Burundi Capital Expenditure: Domestic and Foreign Financed 2001-2009 (% of GDP) D E Source: Ministry of Finance and Ministry of Planning. 8 Burundi’s capital expenditure is composed of Burundi’s investment budget (e.g., investment expenditure appearing in the Law of Finance also called “budget extraordinaire d’investissement”) and of externally financed expenditures (e.g., development projects financed by donors whose expenditures do not appear in the Law of Finance). 11 Overall Execution Rates 3.3 The execution rates of Burundi’s investment budget have been historically low, but have been increasing in recent years. The rate of execution (payment based) of projects financed by internal resources (excluding HIPC) was about 75 percent between 2006 and 2009. Given the relatively low share of domestic expenditure in the investment budget, it has been relatively easier for governments to channel the appropriate resources. For externally financed investment projects, the execution rate has averaged about 60 percent (Table 3.1), with significant volatility, due partly to gaps between commitments and disbursements. Thus the large gaps between the voted and executed budget observed in Figure 3.2 are reflective of greater foreign aid allocations (mainly budget support funds) without a commensurate increase in actual disbursements. While the execution rates are increasing overall, there remains significant volatility, especially with regard to foreign flows. Hence, improved execution rates regarding domestic resources are explained by the development of greater capacity during the post-war transition. Greater execution rates in priority sectors (Figure 3.4) reflect the Government’s commitment to use additional finances/HIPC and MDRI funds to promote PRSP social sectors, which have traditionally been underfinanced. Figure 3.2. Burundi’s Investment Budget: Voted vs. Executed, 2005-2008 (billions BIF) V A Source: Ministry of Finance and Ministry of Planning. 12 Table 3.1. Execution Rates and Breakdown of Expenditures in the Burundi’s Investment Budget 2006 2007 2008 2009 Average Breakdown Breakdown Breakdown Breakdown Breakdown Invest. Budget Invest. Budget Invest. Budget Invest. Budget Invest. Budget (budget) (budget) (budget) (budget) (budget) E/C 67.1 86.6 62.3 63.9 98.8 59.9 95.4 103.3 56.7 92.1 97.4 56.6 79.6 96.5 58.9 Ordinary P/C 61.9 80.3 66.3 57.3 95.9 62.6 88.6 101.2 56.1 90.8 88.1 58.5 74.7 91.4 60.9 P/E 92.3 92.8 77.2 89.8 97.1 64.0 92.8 98.0 56.9 98.5 90.4 57.1 93.3 94.6 63.8 Pro-poor E/C 47.7 86.9 26.8 74.0 88.6 31.6 93.0 109.4 33.1 65.8 91.0 34.4 70.1 94.0 31.5 (mainly P/C 45.0 40.1 28.6 60.2 81.5 29.6 82.2 104.0 34.7 61.7 87.4 33.3 62.3 78.2 31.5 budget support) P/E 94.4 46.1 16.6 81.3 91.9 28.6 88.4 95.0 34.1 93.7 96.1 34.5 89.4 82.3 28.4 E/C 43.4 38.3 10.9 82.5 86.0 8.5 90.6 93.8 10.2 71.7 85.7 9.0 72.1 75.9 9.7 HIPC P/C 42.4 36.8 5.1 73.4 78.4 7.7 84.0 88.5 9.2 65.2 81.2 8.2 66.2 71.2 7.6 P/E 97.6 96.0 6.2 89.0 91.2 7.4 92.7 94.4 9.0 90.8 94.8 8.4 92.5 94.1 7.8 E/C 45.9 81.4 100 78.2 94.5 100 91.9 104.3 100 71.2 94.1 100 71.8 93.6 100 Total P/C 44.1 64.8 100 67.6 89.9 100 83.5 100.8 100 66.4 87.2 100 65.4 85.7 100 P/E 96.0 79.6 100 86.4 95.1 100 90.9 96.7 100 93.2 92.7 100 91.6 91.0 100 Source: Bank staff estimations based on data from the Ministry of Finance 13 Figure 3.3. Execution Rates of Investment Budget by Source of Finance, 2005-2008 (%) D E Source: Ministry of Finance and Ministry of Planning. 3.4 The sector allocations of Burundi’s capital expenditure (overall investment budget) reflect the financing priorities of each of the donors. They are hardly reflective of a carefully calibrated growth model. During the postwar recovery, different donors have financed different projects based on their comparative advantages and areas of interest. Given the Government’s lack of control of aid- financed investment activities and the needs of post-conflict reconstruction, money goes to projects in the social sectors where donor interest is the strongest. Scandinavian donors have been very interested in projects oriented towards institution building and preventing conflict recurrence. Other donors have focused on demobilization and on human capital accumulation. The lack of a strong constituency for large infrastructure projects has meant underinvestment in that sector, which is executed at a lower average rate of 55 percent from 2006 to 2009 (Figure 3.4). 3.5 There has been a significant shift in the balance between recurrent and capital expenditure in recent years. This shift is largely explained by a rise in capital expenditure going to priority sectors. Due to HIPC and overall availability of external projects’ financing, the trend towards increasing capital expenditures has moved at a faster pace than that of (corresponding) recurrent expenditures in these priority sectors, although execution rates have been slower. Furthermore, the share of capital expenditures in total expenditures is higher in priority sectors (especially health and education) than in other sectors. This imbalance can be partly explained by lagging growth in operations and maintenance expenditures (recurrent expenditures), which does not match growth in capital expenditures in the same priority sectors. Also, the tendency to classify all donor-financed expenditures as capital expenditures, which are in fact made up of significant recurrent expenditures, seems to contribute to the imbalance. In this regard, the Government needs to be concerned with the recurrent cost implications of future capital expenditure budgets. In fact, the misclassification of capital expenditures has led to including non-investment related expenditures financed by donors in the social sectors for more than 5 percent of total public investment. 14 Figure 3.4. Execution Rates by Sectors (%) 100 90 80 70 60 50 40 30 20 10 0 Execution rates Source: Ministry of Finance and Ministry of Planning. HIPC Expenditures 3.6 To understand the impact of public expenditure at a disaggregated level, it is vital to consider HIPC expenditures. The sectors benefitting from HIPC-funded expenditures have mainly been those delivering social services, as per international practice. The large preponderance of HIPC resources has gone for (capital) investment budgeting, although there has been some consideration of recurrent salaries in order to deal with operations and maintenance. Overall, the Health and HIV/AIDS and the Education and Research sectors absorb most of the resource allocation; respectively, 46.3 and 16.2 percent, in 2009. Economic Infrastructure9 makes up 17.2 percent of resource allocation under the HIPC program. The Agriculture and Justice sectors absorbed, respectively, 11.2 and 4.9 percent of HIPC-funded expenditures in 2009. In summary, education and health remain the largest beneficiaries of HIPC resources, but there has been a clear move towards sectors with more immediate and direct impact on economic growth (e.g., Water, Energy, Mining, and Public Works) as depicted in Annex 3. 3.7 The execution rates of the national budget (Law of Finance) have been steadily improving since 2005, although there have been recent slippages.10 The average execution rate (commitment basis) was 96.6 percent from 2007-09, reflecting the Government’s ability to secure cash payment for at least 95 percent of the voted credits that are committed. In cash terms, a sharp decline occurred in 2009, reaching a rate of 87.2 percent, which was preceded by strong acceleration in execution rates from an average 63.2 percent in 2005-06 to 89.9 percent and 100.8 percent in the subsequent two fiscal years. The similarity in the evolution of both commitment-based and cash-based execution rates in the same period is evidence of the positive impact of the ongoing PFM reform in terms of eliminating payment arrears. There has been a strengthening of the quality of public expenditures through increasingly efficient project appraisal and budget preparation and an improvement in cash management policy so that prioritized commitment plans are aligned with prioritized cash plans. 9 Economic infrastructure includes water, energy, mining, and public works, with HIPC funds accruing to the ministry in charge. 10 One has to distinguish between the cash-based and commitment-based execution rate. The first is equal to the ratio of overall annual cash payment to overall annual voted credits; i.e., the basis is actual payments. The second is equal to the ratio of overall annual commitment to overall annual voted credits. This is the formula used by the Ministry of Finance, and the rate is the execution rate whose base is equal to commitment. 15 3.8 However, the execution rates of HIPC-funded investment expenditures, measured in cash terms, have only partially recovered from their low levels of 2006. Starting at barely 42.4 percent that year, the execution rate almost doubled in 2008, reaching 84.0 percent, before falling by about 19 percentage points in 2009. Such a pattern raises three issues: (i) the quality of the analysis (in terms of the poverty reduction objective) of HIPC-funded investment projects during budget preparation (sector and central levels); (ii) the capacity of the procurement process to allocate public works to reliable service providers; (iii) and the capacity of the Government to supervise execution of projects in the field. Investment projects under HIPC-voted credits were committed at 82.5 percent and 90.6 percent in 2007 and 2008, respectively. Starting at 43.4 percent in 2006, this evolution suggests some improvement in the quality of HIPC-funded investment projects. So, prior to implementation of the new procurement law, it is likely that procurement organs in line ministries were increasingly able to select reliable service providers, thereby increasing the amounts of voted credits that are committed. As HIPC improves its execution, it will continue to liberate more fiscal space and allow the Government to shift its domestic resources, given the fungibility of the budget. 3.9 Recent outcomes show that public investments funded by HIPC resources were executed better than public investments that were not. Public investment expenditures financed by domestic resources other than HIPC resources showed an execution rate lower (82.2 percent) than that of HIPC- funded public investment (84.0 percent) in 2008. In 2009 where the new Procurement Code was applied, the execution rate declined dramatically for both HIPC public investment (65.2 percent) and non-HIPC public investment (61.7 percent). The results indicate that HIPC-funded public investment resisted these changes better than non-HIPC-funded investment. Specifically, HIPC-funded public investment has benefited from: (i) an independent oversight committee that monitors execution performance (through desk reviews of public accounting documents and application of internal audits recommendations), and a technical committee that screens project proposals based on eligibility criteria; (ii) regular external audits of how HIPC funds are utilized on the ground; and (iii) the provision of HIPC resources to finance the monitoring and supervision of public works execution by allowing line ministries technical staff to do field visits on sites. Those specific instruments do not exist for non-HIPC public investment. 3.10 The monthly execution rates of HIPC-funded investment, however, have been very erratic and indicate overall stressful cash management conditions for the Government. The pattern differs markedly between the baseline situation in 2008, when the old Procurement Code was used, and the situation in 2009, with the application of the new Procurement Code (Figure 3.3a). However, uncertainty seems to be more appealing in 2008 than in 2009.11 In addition, the gaps between commitment and cash execution rates were often bigger in 2008 than in 2009. For the latter year, the rhythm of execution of all HIPC-funded investment was very weak (less that 50 percent) until August, just before the revision of the budget had occurred and capacity building activities for the new Procurement Code have been initiated. Thereafter, execution rates increased significantly, indicating some relief in budget management conditions or constraints, which allowed for acceleration in the commitment of investment projects under HIPC-voted credits. In addition, smaller gaps in 2009 between the execution rates indicate an improvement and closer alignment between available cash and the management of activities in budget. 11 Up until 2008, the execution of the budget (in either cash or commitment basis) tended to be especially intense at the beginning of the fiscal year (January and February), prior to the budget revision, if any (July and August), and at the closing of the fiscal year (November and December). 16 Figure 3.5. Monthly Execution Rates of HIPC Investment Projects, 2008-2009 2009 2008 J F M A M J J A S O N D J F M A M J J A S O N D B E B E B P B P J F M A M J J A S O N D J F M A M J J A S O N D Source: Bank staff based on “Lois des Finances”. Source: Bank estimates based on data from the Ministry of Finance. G. PLANNING AND BUDGETING PROCESS FOR PUBLIC INVESTMENTS Legal Framework for Budget Preparation 3.11 The preparation process for the investment budget follows the basic spirit and rules enacted in the legislation. The procedure for the investment budget has been governed since 2008 by Decree n° 100/100, enacted on June 3, which stipulates the framework for budget preparation12 described earlier in this chapter. The document prescribes the institutional framework and the functional link between the central and line ministries in relation to the role and responsibility of each during the process. Since mid- 2000, there has been some progress in consolidating the budget and in grouping recurrent expenditures and investment projects together by ministry in the documents submitted to Parliament. 3.12 Due to reforms enacted after 2006, the budget preparation process has followed a more disciplined calendar and is in coherence with international norms. Budget preparation starts in the early fall and ends with the approval of the budget by the National Assembly in December. The line ministries are responsible for preparation of the sector investment budgets, with assistance from the Ministry of Planning. The institutional coordination between the key players determines the strategic budgetary outcomes. The macroeconomic framework anchors the plan in the country’s three-year budgetary cycle. There is a consistent classification of expenditure by nature—administrative, functional, and economic. While explicit ceilings are not indicated in the budget circular, the line ministries are aware of their respective envelopes. The budget circular requires that each ministry present its sector strategy along with proposals for the PIP and budget investment program (BIP), including monitoring indicators, to show how the proposed projects would contribute to meeting sector strategy objectives. 12 The decree was supplemented by two pieces of legislation detailing the creation, organization, and operation of the budget preparation units (Ministerial Ordinance n°540/486 of April 13, 2009), and a third describing the modalities for preparing budget estimates. 17 3.13 A significant part of donor-financed investment is not included in the PIP or in the bills submitted to Parliament, and this has adverse consequences for budgetary management. Many donor-funded projects do not even appear in the Treasury accounts, although after execution, they do appear in the Tableau des Opérations Financières de l’Etat (TOFE) statistics. While in theory, Burundi operates with a consolidated budget, in practice the investment budget does not cover foreign-financed investment operations. Projects financed by external resources are not discussed or included in the BIP since the financing is already established. In fact, donors used to work directly with line ministries without consultation with the Ministry of Planning. These external financing arrangements appear in an annex to the Finance Law and shown in the general tables defined in the Finance Law. Furthermore, emergency assistance financed by NGOs and investments by public enterprises are not included in the PIP. To complicate matters, the 2008 budget includes projects that are not listed in the PIP. The absence of foreign-financed operations in the public accounts and expenditure chains complicates the process of monitoring public expenditures. Budget preparation process and institutional responsibilities 3.14 The responsibility for capital expenditure in Burundi is shared across several ministries and agencies, together with international multilateral and bilateral organizations. The public investment budget process follows the traditional budgeting approach commonly used in Sub-Saharan Africa, based on the public investment program. Since the financing comes from a combination of government resources and donor grants and loans from bilateral and multilateral agencies, there is a multiplicity of actors involved in the process. Under the current system, Burundi operates a dual budgeting system with a recurrent budget, managed principally by the Ministry of Finance, and a specific public investment budget prepared by the Ministry of Planning. The Ministry of Finance is responsible for the overall budgetary envelope and for the conformity of expenditures with budgetary ceilings. The existing legislation prescribes that the Ministry of Finance track the expenditure flow along the chain and maintain overall compliance with budgetary procedures. Since 2008, as a part of measures aimed at improving budget presentation, the investment and recurrent budget were consolidated. 3.15 The Ministry of Planning occupies a central role in the institutional landscape, although its mandate remains diffuse. The ministry is the focal point for all capital spending in the country. Its functions include public investment planning, articulation of a longer-term development strategy, and macroeconomic forecasting. Also, in theory, it is responsible for managing and monitoring external finance and acting as an external aid management entity. A presidential decree (January 2009) established the Programming Department (Département de la programmation) as the institutional body responsible for centralizing information and elaborating the Public Investment program (PIP).13 Thus, the planning ministry is in charge of preparing the macroeconomic framework document, as well as the Priority investments. It prepares the multi-year PIP with strong input from the line ministries, which send their lists of projects to it. 13 In 2007, the Government set up the National Aid Coordination Committee (CNCA). The Permanent Secretariat of CNCA first fulfilled the first role of centralizing information as an institutional framework for coordination of aid. 18 3.16 According to international norms and the PIM methodology, the Ministry of Planning is responsible for a strategic vetting of projects and ensuring a rigorous project selection process. In keeping with international best practice, the ministry is responsible for ensuring the technical quality of the projects and securing adequate funding. Overall, the Ministry of Planning is the lead agency for investment programming. 3.17 The line ministries are responsible for ensuring that the projects that are sent to the central ministries are aligned with sectoral plans and are technically viable. In theory, the line ministries are also responsible for physical monitoring of the projects. 3.18 The biggest players in the capital budget are the foreign donors, who finance a large chunk of expenditure. The World Bank, the EU, and AfDB are the main multilaterals, while the Netherlands, Norway, the United States, France, Belgium, and the United Kingdom (UK/DFID) are the main bilateral actors, financing a large part of the budget. In collaboration with the Government, the partners have financed HIPC debt relief. While traditional donor project assistance has been outside the budget, the virtue of the HIPC process is that it has worked closely within the budget. Although each donor has its own procedures and protocols, their general mandate is to oversee the coherence and viability of their projects and assess the development impact. 3.19 A range of other agencies also play a role in the preparation and execution of the investment budget. The Treasury is responsible for the preparation of a coherent and consistent cash management plan to the attention of the OTB (Ordonnateur Trésorier du Burundi), who issues payment visas to the central bank. The latter executes the payments instructions (A state cashier based at the central bank makes the payments). It also monitors the country’s monetary policy and ensures disbursements/payments from aid agencies via the commercial banks. The Ministry of Foreign Relations is responsible for signing of protocols, (donor agreements) and the courts confirm their conformity to Burundian law. .The Ministry of Commerce and Industry, which monitors all domestic trade, also supervises the imports of capital goods for infrastructure projects. Project Viability 3.20 One of the main methodologies used to assess project viability is the economic rate of return. The convention is to use a rate of about 10 percent to calculate the net present value of a stream of future assets. However, economic rates of return differ considerably across sectors and countries, and social sector projects are known have lower estimated rates of return due to a longer project cycle and greater difficulty in quantifying the results. The methodology of economic rate of return uses cost-benefit principles to measure the viability of projects and establish a numerical benchmark for selection. In the development economics literature, the economic rates of return are useful as a first approximation, but they are frequently ignored because of the multiple factors influencing project selection. Energy and infrastructure projects are most easily measurable using rates of return, but as one ventures into other areas, rates of return have a diminished significance. Table 3.2. PIP Selection Criterion in Burundi PIP CRITERIA COMPLIANCE WITH PIM METHODOLOGY Well-publicized strategic guidance for public Consultation between Ministry of Planning and line investment decisions at the central and ministerial ministries precedes PIP selection process, although in 19 levels practice, there is a lack of detailed discussion; the PIP does not include projects by humanitarian or non- government organizations (NGOs). The project must appear in the sector strategy Established process for screening project proposals for document based on the PRSP, must have the right basic consistency with Government policy and technical studies (with demonstrated rates of return), strategic guidance and have sufficient financing to be included in first year of PIP; in practice, PIP is arbitrary and incomplete. Feasibility studies are not used often due to lack of A formal appraisal process to evaluate proposed technical capacity, and economic rates of return are not public investment proposals for costs and benefits always used, especially for projects in the social sectors. Project appraisals are constrained by lack of internal Project appraisals formally undertaken by the financing and capacity, and lack of detailed knowledge sponsoring department or by an external agency by donor agencies of the realities on the ground. More than three-fourths of PIP is donor-financed; The proportion of the public investment program that absence of counterpart funds renders government is donor financed, and quality of appraisal monitoring difficult; donors have their own appraisal methods. Government conducts partial evaluation of nationally Screening of appraisals by an external agency for financed and HIPC projects, while donors conduct quality and objectivity of appraisal appraisals of donor-financed projects. In theory, project selection is undertaken prior to the Undertaking of final project selection as part of the multi-year budget preparation process; in practice, it is budget process arbitrary. The classification scheme is based on feasibility study, Controlling the gates to the budgeted public investment availability of financing, and links to sectoral strategy; program there is an overall lack of rigor in the process. The Public Investment Program 3.21 The PIP provides the central strategic basis for the country’s investment budget. It is envisaged as a medium-term program of public investments aimed at ensuring that the allocation of public resources is in line with the country’s development objectives. The PIP plays an integrative role in rebuilding and rehabilitation, as well as in reconciliation and consolidation of peace. For 2009, the PIP accorded priority status to projects with three criteria: a financing source, a feasibility study, and a link to sectoral development strategy. On the basis of these criteria, the PIP hierarchy of projects is: (i) projects in progress with a technical feasibility study and a source of financing; (ii) projects approved with at least financing under negotiation; (iii) projects with studies under preparation; (iv) ideas for projects not yet documented; and (v) projects rejected. As opposed to the past, there are no specific requirements for economic rates of return/cost-benefit analysis, environmental impact, and recurrent costs to capital budgets. There is also no specific requirement for information on the performance of ongoing projects. 3.22 Project selection is based more on the availability of financing than on a technical feasibility study or Government priorities, and thus can be considered supply driven. There is no ranking of projects according to rates of return and no serious arbitrage in which projects are rejected. The weaknesses of the overall budget system, the poor analysis and planning at budget preparation, the weak interagency coordination, and the centralization of budget preparation all contribute to poor budget execution. Due to the weak project planning and preparation process, many projects in the budget are 20 often not ready for implementation (see above). Frequently, projects enter the budget not fully designed, and with no credible implementation plan (i.e., lacking procurement and staffing plans). Weak interagency coordination renders procurement and disbursement processes even more complex and time consuming. The general lack of accountability and incentives to take risks make it difficult to cut through bureaucratic bottlenecks. Finally, the overall budget process lacks a rigorous project selection process that would ensure that the best mix of demonstrably good projects is funded with the limited resources available. Line ministries do not have formal and well publicized guidance on the technical aspects of project appraisal appropriate to the capacity of ministries and departments. The integrity of investment planning process is also undermined by the lack of independent peer review. 3.23 Thus, a number of anomalies have affected the credibility of the PIP exercise. First, the PIP includes only projects financed by local resources or the counterpart funding of foreign-financed investment. As noted above, the projects financed by external assistance are not included in the BIP. The donors use to work directly with the line ministries, without consulting the ministries of finance and planning, to discuss and prepare projects for financing. Most of these projects were not discussed in the budget process, since the financing was already established. As a consequence, there was not strong ownership of the project by the Government and it was not clear whether these donor-dominated projects were in line with Government objectives and strategies. Also, more than 50 percent of the projects (mostly small projects of NGOs) are excluded from the PIP. Second, for a number of years, the PIP was adopted after the Finance Law was passed, when in theory it should be done before the law is passed. Thus, the PIP remains off-budget until it is reconciled with the Finance Law (this report uses PIP numbers off-budget).Until all the projects financed by foreign donors are inscribed in the Public Finance Law, the PIP will remain an incomplete document that is more of a wish list than the reality on the ground. 3.24 In practical terms, there are a number of salient features in the Burundi’s PIP. First, projects are characterized by a small average size (less than US$500,000) and by a short life cycle (fewer than 5 years). Second, the quality of projects varies according to the sector. Health and energy projects have different public investment traits according to their technical and social mandates. The common perception is that health and education projects are easier to manage than energy and agriculture projects, since their scale is smaller and the country has more recent experience with them. In particular, the health ministry in Burundi developed many viable projects during the postwar period. By contrast, energy projects require large overhead and coordination challenges. The goal of the PIP is to ensure that the investment budget fosters close coordination among sectors. Finally, recurrent costs are frequently included in the PIP, leading to a slight overestimation of project costs in terms of fixed capital formation (see below). 3.25 Budget classification has been misleading, especially with counterpart funding. These funds, which are often required in donor-financed projects, serve to cover the current account spending part of the projects. However, the budget classification does not take into account that part of the investment project, making it seems as if investment is too high compared to its impact, as well as compared to the level of investment in neighboring countries (see Chapter 4). 3.26 The PIP can be improved with a set of reforms that allow greater efficiency. Most important, there must be a more rigorous analysis of projects based on rates of return and a standard set of feasibility studies. At the same time, donors and line ministries need to coordinate to ensure that the PIP has a strong 21 database of all projects. Finally, a working group can be created within the Ministry of Planning to rigorously track project quality and insure a better link between the PIP and the investment budget, and with the recently established central MTEF. Through these reforms, Burundi can have a more powerful public investment program. 3.27 Overall, Burundi follows some of the general principles of good investment planning, although there are multiple deficiencies and challenges, partly due to the difficulties of post-conflict reconstruction and planning. On one level, the public investment program does emanate from sectoral strategies prepared by line ministries. For example, health projects are linked to identified problems in health infrastructure. Moreover, there is a basic project selection process that focuses on coherence and cost effectiveness. The problem arises in the dynamics of project selection. H. IMPLEMENTATION AND MONITORING OF PUBLIC INVESTMENTS Key Obstacles 3.28 Implementation of the investment budget has been improving in recent years, but there remain major challenges to its successful execution. The low rate of execution of donor-financed projects is a serious cause for concern, and can be explained by a number of factors. First, the weak selection process negatively impacts the implementation phase. Second, the ministries need to strengthen their capacities, especially at the sectoral level, to handle accounting and financial management (for example, since the Ministry of Education lacks expertise, it has not been able to handle the donor- financed Education Fund). Third, there is no common M&E mechanism to track the execution of domestic and donor-financed projects. This has undermined Government control of the budget. Fourth, most donors use their own procedures for procurement, and the introduction of the new procurement system has caused delays. Table 3.2 shows the compliance of the PFM to related Public Expenditure Financial Accountability (PEFA) criteria. 22 Table 3.3. Compliance of PFM with PEFA Criteria on Public Investment A. PFM-OUT-TURNS: Credibility of the budget PI-1 Aggregate expenditure compared to original approved Execution rates for nationally funded investment budget projects (including those funded with HIPC money) have improved significantly in recent years to more than 75 percent, although rates for donor-financed projects vary widely, from 30 to 80 percent, depending on the donor Composition of expenditure compared to original approved The separation between recurrent and investment budget expenditure follows the budgetary principles. B. KEY CROSS-CUTTING ISSUES: Comprehensiveness and Transparency Classification of the budget The investment budget follows international norms after recent reforms Comprehensiveness of information included in budget The investment budget is not comprehensive. documentation C. BUDGET CYCLE C(i) Policy-Based Budgeting Multi-year perspective in fiscal planning, expenditure policy, The MTEF is not yet fully incorporated into and budgeting investment budget planning. C(ii) Predictability and Control in Budget Execution Predictability in the availability of funds for expenditures Funds remain unpredictable due to donor procedures. Competition, value for money, procurement controls Procurement issues impede the efficiency of the public investment program. Effectiveness of internal controls for non-salary expenditures Controls are in theory but little ex post verification Effectiveness of internal audit Internal audit procedures are weak. C(iii) Accounting, Recording, and Reporting Timeliness and regularity of accounts reconciliation Accounts reconciliation is stronger for recurrent budget than for investment budget. Quality and timeliness of in-year budget reports Reporting on the execution of the investment budget, with the exception of HIPC resources, is weak. C(iv) External Scrutiny and Audit Scope, nature and follow-up of external audit External audit for investment budget is slow, but oversight mechanisms are being strengthened (i.e., the HIPC Committee) Legislative scrutiny of the annual Budget Law Annual Budget Law goes through the legislative process; investment budget is in the annex 23 3.29 The lack of coordination between donors and the Government with respect to investment projects is especially troubling. Donors and Government officials do not meet on a regular and systematic basis to discuss budget implementation. While formal coordination mechanisms between the Government and donors do exist, they need strengthening. The Cadre de Partenariat, which was established in recent years, anchors donor and government in a strategic partnership, although it is only the first step in a process. Furthermore, the lack of harmonization between the procedures (especially in procurement) used in donor projects and the Government’s own procedures has created difficulties for budgetary management. Reporting requirements have been uneven, and the SIGEFI does not have information on the projects executed by donors, since they do not pass through the government’s expenditure chain.14 Government officials are also not fully aware of donor procedures, and the lack of counterpart funding has rendered budgetary management difficult. Since many of the donor projects are not included in the budget, there is no allocation of counterpart funds. Ensuring greater Treasury management of aid flows and disbursements could help improve coordination. Table 3.4. Public Investment (% of GDP) 2007 2008 2009 2010 Burundi 12.7 15.1 16.3 17.1 Cameroon 3.9 5.7 5.7 6.0 CAR 3.6 4.5 5.0 5.3 Congo 2.4 3.7 9.3 18.0 DRC Kenya 4.7 6.8 8.0 9.2 Niger 11.6 10.3 12.3 10.2 Rwanda 8.6 11.0 12.0 14.0 Senegal 11.2 10.0 10.6 11.2 Tanzania 6.1 8.0 8.2 9.1 Uganda 5.7 5.6 5.5 7.2 Source: IMF dataset. 3.30 Key reforms also need to be undertaken to improve the aid management regime. Most important, the current partnership framework (Cadre de Partenariat) between the Government and development partners can be augmented by the establishment of a Financial Unit (FU) within the finance ministry in charge of centralizing financial and budgetary data on foreign-financed expenditures. This information could be used by the Ministry of Finance to collaborate with line ministries and CNCA. There needs to be a speedier information flow between the Government and the donors with regard to the projects, and in many cases, annual workshop involving donors and central authorities as part of the budget preparation process can be used to ensure continued partnership in preparing key documents. 3.31 The implementation experience of expenditures funded under HIPC and education SWAp illustrates some of the conditions surrounding public investment efficacy in Burundi. To better understand how things have evolved since the PFM reforms, we use HIPC-funded investment and the basket funds for education (“Fond Commun de l’Education - FCE”) to draw key lessons that could help 14 To overcome the problem of incomplete information in reports to the Treasury on externally financed investments, the Treasury aims to gradually transfer project accounts to the central bank. 24 improve the execution rates of Burundi’s capital expenditure in the coming years. On the one hand, the HIPC expenditure experience allows us to identify three key elements of successful implementation of public investment, namely: (i) a rigorous selection process of investment projects; (ii) effective institutional organs for M&E with full support of oversight institutions of procurement and budget; (iii) and use of systematic audits (and related action plans) to keep investment program on tracks. On the other hand, the experience of the FCE shows that poor preparation and lack of coordination between main stakeholders can seriously hamper the implementation of public investment (Annex 5). I. BURUNDI PUBLIC EXPENDITURE ISSUES 3.32 The Government needs to expand its fiscal space by finding a soft balance between current expenditures related to the post-conflict situation and investment expenditures critical for accelerated growth. While development spending, including public investment, has been increasing and now accounts for more than 15 percent of GDP, high defense and security spending and the high wage bill remain key obstacles to reallocating public resources toward priority economic and social sectors. As the country has recovered from war, it has attempted to reform its public finance regime and expand education and other social sector expenditures, as well as to contain the rapid growth of the wage bill. In addition, aid inflows have increased significantly, and the share of poverty-reducing expenditures increased in 2007-2009 due to the availability of HIPC resources and higher government revenue.15 For example, overall public spending on education rose from less than 4 percent of GDP in 2005 to close to 10 percent of GDP by 2010. The result of this new investment has been a major increase in primary school enrollments and higher utilization of health facilities. In parallel, however, there has been a constant growth in the wage bill to match the expansion in basic services. Such a situation prevents the expansion of public investment and the use of available resources for growth-oriented projects (see Chapter 4). 3.33 One major challenge for Burundi has been the implementation of the capital budget, as discussed above (Chapter 3). As the share of social and economic investment expenditures has increased, there has been growing pressure on the Government to improve the efficacy of the capital budget. Compared to other East African countries, Burundi’s capital investment is high, averaging about 15 percent of GDP over the last four years (Table 3.3). While it is largely financed by increases in donor funding, changes due to HIPC have led to some of these increases. However, said investment has had a relatively weak impact on growth, for several reasons: (i) high increase of investment budget in non- productive sectors; (ii) the misclassification of counterpart funding as investment expenses rather than current expenses for more than 5 percent of the investment; (iii) low execution rates due to internal planning and budget process-related issues (discussed in chapter 3); and (iv) a series of macroeconomic and fiscal challenges discussed below. 3.34 Burundi’s high rate of capital expenditure relative to GDP presents some macroeconomic and fiscal challenges. First, an overdependence on foreign finance leaves the country’s investment program vulnerable to fluctuations in donor assistance, especially given the large gaps between commitments and disbursements. Moreover, the recurrent cost implications of a large capital increase are problematic, especially in the absence of government counterpart funding. Thus, the capital increase may 15 As a result, the Government has complied with the commitment made to the international donor community to allocate savings from the HIPC initiative to pro-poor expenditures, as articulated in the PRSP. 25 distort the budget to some extent. Third, an aid surge may result in possible Dutch disease impacts if the aid interventions are not used productively to finance the expansion of the non-tradable sector. Finally, the procurement implications of the large capital budget may drain scarce national capacity and lead to overdependence on expensive and uncertain foreign supplies, with adverse effects on service delivery. 26 Chapter 4. Public Investment Prospects J. INTRODUCTION 4.1 The composition of government expenditure is dominated by a large share of recurrent expenditures.16 These recurrent expenditures accounted for more than 66 percent of total expenditure over 2005-2009, leaving about 30 percent for capital expenditure and less than 4 percent for exceptional expenditures (DDR and elections). In recent years, however, mainly due to an increase in foreign- financed capital expenditure, there has been a moderate and gradual shift toward increased capital expenditure in Burundi’s budget. The share of capital expenditure in total expenditure increased slightly over 2005-2009, from 22.9 percent in 2005 to 32.2 percent in 2009. In the meantime, the share of recurrent spending contracted, from 77.1 percent of total expenditure to 61.3 percent over the same period. Yet recurrent expenditure has still accounted for an average of about 23.5 percent of GDP over 2008-09, continuing to dominate the structure of expenditure, while capital expenditure (excluding foreign-financed outlays) has been relatively modest, at less than 4 percent of GDP over the same period (Table 4.1). Due to this dynamic, foreign aid has tended to be allocated for investment expenditure. Table 4.1. Economic Distribution of Public Expenditures, 2005-2009 percent of GDP Average Average 2005-06 2008-09 Current expenditure 23.5 25.3 Wages & salaries 9.2 11.8 Goods and services 7.2 6.2 Transfers and subsidies 3.9 5.4 Interest payments 3.2 1.8 Domestic 1.8 1.3 Foreign 1.4 0.5 Exceptional expenditure 3.7 2.8 DDR 2.3 2.7 Elections 1.4 0.1 Capital expenditure 11.2 15.3 Domestic resources 3.1 3.5 External resources 8.1 11.8 Source: IMF; Burundian authorities. 4.2 As noted earlier in the report, the recent significant increase in capital expenditure is largely due to increased foreign-financed expenditures. Due to increased international assistance, capital expenditure has consistently expanded, from an average of 11.2 percent of GDP in 2005-06 to 15.3 percent in 2008-09.17 About 79 percent of capital expenditure was financed by foreign assistance in 2008-09, compared to 72 percent in 2005-06. However, this assistance remains very volatile. Despite the availability of HIPC resources, the Government's capital spending remains quite modest (at 3.5 percent of 16 For wages and salaries, other goods and services, interest payments, and transfers and subsidies. 17 The upward trend in foreign assistance started at the end of 2003 with the signing of a global peace agreement which included factions that were not part of the 2000 Arusha peace agreement. 27 GDP in 2008-09, from an average of 3.1 percent of GDP in 2005-06). This is worrisome, since it gives the impression that Burundi’s growth strategy is greatly influenced by donors. But given its limited fiscal space, Burundi will likely continue to rely on foreign assistance to support its growth and development strategy, at least in the short to medium term. 4.3 The efficiency of public investment needs to be improved. Compared to other EAC countries, Burundi’s investment is one of the least efficient, with notably lower road and energy penetration rates (2 percent) and an Incremental Capital Output Ratio (ICOR) of 5, compared to 3 for Uganda and Rwanda (Box 4.1). 4.4 Unlike its EAC comparators, with insignificant private investment, Burundi’s total investment is heavily dominated by public investment. Moreover, the execution rates of public investment averaged 75 percent over the past 3 years for domestically financed projects, and about 60 percent for externally financed projects (see Chapter 3). Yet as discussed in this PER, the efficiency of Burundi’s public investment is still hindered by low execution rates as well as aid volatility. Given the importance of investment for economic growth and poverty reduction, this volatility, as well as the complementarity between public and private investment, means that the Government needs to shift budgetary allocation toward the domestically financed capital spending. This will only be possible through a more effective allocation of public resources, away from non-productive expenditures such as those related to security, and more toward spending on goods and services associated with capital outlays and social sectors. Box 4.1. Definition and Use of the Incremental Capital Output Ratio The incremental capital output ratio (ICOR) is defined as the ratio between investment in some previous period and the growth in output in the subsequent period. It is calculated based on constant price data. ICOR has been used since the 1950s, and is still used by the Bank and other international organizations, to measure the investment required to reach GDP growth targets. There are several critical points to be mentioned about this ratio: (i) growth in output can be due to several factors other than investment in new capital, such as growth in productivity or the production capacity utilization rate; and (ii) the lag between investment and increased output will vary. Thus, to obtain a reliable relationship, the ICOR would be measured for a longer period, perhaps three or four decades. Based on the above definition, a higher ICOR indicates that a county’s investment results in less robust growth, while a lower ICOR indicates that a country’s investment results in more robust growth. Example: If a country has an investment rate of four percent of GDP and an ICOR of four, growth in GDP will be one percent per year. If the population is growing faster than one percent per year, GDP per capita will fall. Thus if the targeted GDP growth is five percent next year, then required investment this year would be 20 percent of GDP. Source: Adapted from the World Bank Statistical Manual. 28 4.5 The final section of this paper will deal with: (i) the progress in strengthening strategic budget planning since introduction of MTEF; and (ii) a quantitative evaluation of growth and poverty, through the Burundi Macroeconomic Model (MacMod_BI),18 of an alternative scenario of improved efficiency of public investment. Table 4.2. Comparison of Investment Volume and Efficiency in EAC, 2005-2009 2005 2006 2007 2008 2009 Average 2005-09 GDP Growth Rate Burundi 0.9 5.1 3.6 4.5 3.2 3.5 Tanzania 7.4 6.7 7.1 7.4 5 6.7 Kenya 5.9 6.4 7.1 1.7 2.5 4.7 Uganda 6.3 10.8 8.6 9.5 5.5 8.1 Rwanda 7.1 7.3 7.9 11.2 5.3 7.8 Total Investment (% GDP) Burundi 10.8 16.3 17.5 19.4 20 16.8 Tanzania 25.1 27.6 29.6 29.8 28.4 28.1 Kenya 16.4 19 18.9 17.7 18.8 18.2 Uganda 22.4 21.2 22.1 23.5 24.4 22.7 Rwanda 21.6 20.4 21 24.1 22.6 21.9 Public Investment (% GDP) Burundi 9.8 9.8 12.7 15.2 16.3 12.8 Tanzania 5.6 6.1 7.9 8.8 9.6 7.6 Kenya 4.4 4.4 6.8 6.3 6.8 5.7 Uganda 7.1 8.4 8.4 7.7 8.6 8.0 Rwanda 9.1 7.6 8.6 11 4.7 8.2 Efficiency (ICOR) Burundi 12 3 5 4 6 5** Tanzania 3 4 4 4 6 4 Kenya 3 3 3 11 8 4** Uganda 4 2 3 2 4 3 Rwanda 3 3 3 2 4 3 Sources: IMF, World Bank, and national authorities. Note (**): Outliers in 2005 for Burundi and 2008 for Kenya are mostly explained by weak economic performance in Burundi due to a severe drought, and the political instability that followed the 2007 disputed elections in Kenya. Both figures were excluded in the calculation of the average ICOR. 18 The Burundi Macroeconomic Model was developed for the preparation of the MTEF. The growth path in this model is determined by macroeconomic and sector policies that affect long-term productive capacity (human capital, infrastructure, and direct investment flows); short- and medium-term fluctuations in aggregate demand (monetary and fiscal policies); and the external environment (oil price and terms of trade). This endogenous growth model was not estimated econometrically. Rather, the model used elasticities estimated through cross-country regressions. Other components of the macroeconomic framework are determined by accounting equations and standard financial programming principles. The medium-term budget framework allocates resources to sector budgets based on Government priorities. 29 K. MEDIUM-TERM EXPENDITURE FRAMEWORK AND FUNDAMENTALS FOR IMPROVING THE MANAGEMENT AND EFFECTIVENESS OF PUBLIC INVESTMENT 4.6 The progressive establishment of a medium-term expenditure framework will consolidate the link between strategic policy objectives and budget allocations in three main ways. First, improving the reliability of macroeconomic and revenue projections will facilitate the setting of credible spending caps. Second, validation by the Council of Ministers of multi-year, inter-sector allocations will aid in achieving the objectives of the PRSP. Third, it will empower line ministries to propose intra-sector allocations based on their own strategies and priorities. With the help of other donors, the Government has already prepared sector MTEFs in health, education, and agriculture. 4.7 The development of three-year, rolling MTEFs in key sectors will address longstanding weaknesses in the link between recurrent and investment spending, and will improve the prioritization of intra-sector expenditures. Ultimately, this will lead to improved efficiency of public investment. A team of consultants financed by the Bank-executed Belgian Poverty Reduction Program helped the Government to define a common methodology for future sector MTEFs and a draft roadmap for extending the process to other ministries. Consequently, as of December 2009, the preparation of the central MTEF19 and the expansion of sector MTEFs have started in several other sectors, including Mining and Energy, Public Works, Water, Defense, and Public Security. A high-level seminar was organized in December 2009 to restitute the first phase of the MTEF deployment process (Box 4.2 presents a schematic description of MTEF). During this seminar, the authorities confirmed the Government’s commitment to MTEF and agreed to take steps to institutionalize this process. 4.8 Although the authorities have emphasized improving budget preparation, other weaknesses remain. With a few notable exceptions (i.e., health and education), most line ministries either lack a sector strategy, or have a strategy of such low quality that it is not a useful input into budget preparation. Nevertheless, following institutionalization of the MTEF which is expected to take place in 2010 at both the central (Finance and Planning) and sector levels, the introduction of the sector MTEFs will need to continue to cover all ministries. It is expected to be used as a planning tool in the preparation of the draft Finance Law for 2012. The preparation of a new PRSP in 2010-2011 will provide an opportunity to refine sector strategies and place them within the medium-term framework. 19 A model central MTEF for the period 2011-2013, supported by macroeconomic forecasts derived from the macroeconomic model of the Ministry of Planning, was recently developed jointly by the finance and planning ministries. 30 Box 4.2. Schematic Description of the Functioning of the Burundi MTEF CSLP et autres Documents stratégiques Heart of the M d l : Historic Data: Projections Resources Projections: Central MTEF and recent budget executions Sector 1: SOVEREIGNTY Sector 1: SOVEREIGNTY - Presidency, Vice-presidency - Presidency, Vice-presidency - Parliament - Parliament Distribution Sector 2: DEFENSE AND Sector 2: DEFENSE AND SECURITY SECURITY Parameters Sector 3: GOVERNANCE Sector 3: GOVERNANCE - JUDICIARY SERVICES - JUDICIARY SERVICES Sector 4: GOVERNANCE Sector 4: GOVERNANCE - ECONOMIC SERVICES - ECONOMIC SERVICES Allocation to: Sector 5: SOCIAL SERVICES Sector 5: SOCIAL SERVICES - First Necessities - Recurrent expenditure Sector 7: INFRASTRUCTURE Sector 7: INFRASTRUCTURE - New Measures TOTAL TOTAL ƒ Source: MOP L. INCREASING THE LEVEL AND EFFICIENCY OF PUBLIC INVESTMENT AND THE PROSPECTS FOR POVERTY REDUCTION 4.9 This section presents MTEF projections with a special focus on public investment, and analyzes the potential impact of this investment. These projections are based on the macroeconomic framework model jointly prepared with the authorities in the context of the 2010 Country Economic Memorandum (CEM) report, which is briefly described in Chapter 1 of this report.20 Assuming the implementation of the priority recommendations identified in previous chapters, the discussion on MTEF projections is followed by a series of simulations aimed at evaluating the impact of improved efficiency of public investment on growth, poverty reduction, and achievement of the MDGs. 20 This medium-term macroeconomic framework (2010-15) reconciles the macroeconomic framework of the International Monetary Fund’s Poverty Reduction and Growth Facility program with the Government’s Poverty Reduction Strategy (PRS) targets. It is based on recent economic trends and assumptions about continued progress of the reform process. 31 Medium Term Expenditure Framework: Importance of Public Capital Spending 4.10 Under the MTEF, capital spending is expected to increase modestly. Total expenditure (excluding debt service) is expected to increase initially to 47.1 percent of GDP a year in 2010–12, up from 36.9 percent of GDP between 2002 and 2008 and 50.8 percent of GDP in 2009. However, it will decline to 41.2 percent in 2013–15 (Table 4.3). Following a significant increase in 2009 to accommodate the implementation of the peace accord with the last rebel group, total expenditures are expected to rise again in 2010 (mostly because of the need for resources to prepare for elections). Reduced debt service following Burundi’s debt relief status in the context of HIPC also explains the increase in total spending. Capital expenditure increases to an average of 16.6 percent of GDP over the projection period (2010–15), which is consistent with priorities set in the PRSP. Economic and social sectors (education, health, and productive infrastructure) are expected to be the main beneficiaries of this increase in capital spending (Table 4.3). The budget share of capital expenditure is projected to increase modestly to 40.1 percent, on average, in 2010–15 (from an average of 33.7 percent in 2002–09). Although still below what is required to achieve the MDG targets, the projected trend is encouraging (see 2010 CEM report). Table 4.3. Baseline Scenario—Medium-Term Expenditure Framework, 2007–2015 Actual Estimated Projected Average Expenditure 2007 2008 2009 2010 2011 2012 2013 2014 2015 2010–15 (percent of total expenditure) Recurrent expenditure (excl. debt service) 63.2 60.6 60.4 59.1 59.0 59.3 58.8 58.9 58.8 59.0 Capital expenditure 36.8 39.4 39.6 40.9 41.0 40.7 41.2 41.1 41.2 41.0 Sectoral allocation (% of total expenditure, excluding debt service) Social sector 28.9 28.8 33.7 34.3 35.2 36.1 36.9 37.8 38.6 36.5 Education 23.2 20.3 23.8 24.0 24.4 24.9 25.3 25.7 26.1 25.1 Health and social development 5.7 8.5 9.9 10.3 10.7 11.2 11.6 12.1 12.5 11.4 Productive infrastructure 6.0 15.8 12.8 13.1 13.1 13.1 13.2 13.2 13.2 13.1 Mining and energy 1.0 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 Transport and telecommunications 0.4 0.4 2.9 3.0 3.0 3.0 3.1 3.0 3.1 3.0 Water, environment, and urbanism 0.3 0.5 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 Public works and equipment 4.3 12.7 7.0 7.2 7.2 7.2 7.3 7.3 7.3 7.2 Productive sectors 6.4 5.3 8.8 9.6 10.1 10.5 11.1 11.6 12.1 10.8 Agriculture 5.0 4.1 8.3 9.0 9.5 10.0 10.6 11.1 11.6 10.3 Industry and commerce 1.4 1.2 0.5 0.6 0.6 0.5 0.6 0.6 0.6 0.6 Other sectors 58.7 50.1 44.7 43.0 41.6 40.3 38.8 37.4 36.0 39.5 Sources: World Bank staff and Burundi authorities. 4.11 This increase is justified by the reallocation of nonproductive expenditure toward socioeconomic priority expenditures following the consolidation of peace, the reduction of military expenditure, and continued efforts to rationalize the national budget. The budget share of productive infrastructure increased markedly in recent years, from 6.0 percent in 2007 to 15.8 percent in 2008; the increase is mostly explained by growth in spending on public works and equipment (construction projects). The share of productive infrastructure is projected to be steady at 13 percent of GDP over the period of projection (2010–15). The budget share of the productive sectors (mostly agriculture), which decreased between 2007 and 2008, is estimated to increase to 8.8 percent in 2009, reflecting the authorities’ priorities for rural development. Furthermore, that priority is reflected in the projected budget share of these sectors, which is expected to rise by 2011 above the critical 10 percent level recommended by the New Partnership for Africa’s Development (NEPAD). 32 Simulating an Improved Efficiency and the Level of Public Capital Spending 4.12 Simulating the impact of an improved efficiency of public capital spending is based on the following assumptions: • Improving execution rates of public capital expenditure (averaging 72.4 percent in 2007-09) translate to a 2 percent increase in the annual growth rate of capital expenditure. • Controlling of the growth of military spending translates to a reduction of 2.5 percentage points of its weight in total recurrent and capital expenditures, respectively. This reduction is compensated by a one percentage point increase in the budget share of agriculture and public works, and a 0.5 percentage point increase for the energy sector. • Improving public expenditure efficiency translates to a slight increase in MTEF elasticities for the primary (food crop, export crop, livestock), and construction sectors.21 4.13 As indicated in the 2010 CEM, key projects to be undertaken over the short to medium term include the development of Burundi’s domestic hydroelectric potential (Kaganuzi, Mpanda, Kabu 16), and the provision of adequate power lines with necessary substations to all provincial capitals. The rehabilitation and maintenance of national highways and community and rural road networks will also contribute to improving Burundi’s infrastructure gap. There is also scope for a significant increase in growth in the medium and long term, should potential investments in the mining sector materialize. 4.14 The results of the above simulation (Table 4.4) indicate a significant increase of GDP growth, expected to reach 7.3 percent in 2015 compared to 6.1 percent under the baseline scenario, without an impact on inflation and the fiscal balance. This simulation also indicates that poverty incidence will decline to 35.8 percent in 2015 compared to 38.2 in the baseline scenario. The CEM estimated that the impact of increased infrastructure spending (as reflected in the Infrastructure Action Plan within the CEM), combined with accelerated reforms, could potentially lead to a higher growth rate, averaging about 8.5 percent for the period 2010–15. 21 MTEF elasticities increased from 0.025 to 0.03 for the food crop and export crop sectors, from 0.02 to 0.025 for livestock, and from 0.2 to 0.3 for construction. 33 Table 4.4. Impact of Improved Efficiency of Public Investment, 2009-2011 2009 2010 2011 2012 2013 2014 2015 INDICATORS Real GDP growth (%): Baseline 3.5 4.0 4.4 4.9 5.5 5.9 6.1 Real GDP growth (%): Alternative scenario 3.5 4.1 4.5 5.7 6.5 7.0 7.3 Inflation (CPI): Baseline 8.8 6.9 8.7 8.7 9.4 9.7 9.9 Inflation (CPI): Alternative scenario 8.8 6.9 8.7 8.7 9.4 9.7 9.8 Baseline Overall fiscal balance (% of GDP) 58.5 -8.4 -6.9 -5.2 -4.6 -2.4 -2.1 Fiscal balance, excluding grants (% of GDP) -31.4 -31.7 -26.3 -25.5 -24.3 -20.8 -19.4 Alternative scenario Overall fiscal balance (% of GDP) 58.5 -8.4 -6.9 -5.1 -4.5 -2.3 -2.0 Fiscal balance, excluding grants (% of GDP) -31.4 -31.7 -26.4 -25.5 -24.1 -20.6 -19.0 Poverty Reduction Poverty incidence (%): Baseline 56.8 54.4 52.0 49.1 46.0 42.7 38.2 Poverty incidence (%): Alternative scenario 56.8 54.4 51.6 48.3 44.6 40.8 35.8 Source: MACMOD_BI.. M. CONCLUDING REMARKS 4.15 Burundi’s current level of total investment is quite insufficient to generate the growth necessary to reduce poverty incidence. Burundi’s total investment of around 20 percent of GDP is the lowest among its EAC counterparts. Consequently, while other EAC countries enjoyed much higher growth during 2005-09, Burundi’s economic growth stagnated around 3.5 percent. Unlike in Burundi, total investment in other EAC countries is dominated by private investment. Inefficiencies in public investment and poor private investment is the main mainly explain low level of total investment. 4.16 The macroeconomic simulations conducted in this chapter confirm the importance of improving the efficiency of public investment. Burundi’s fiscal space should also be improved through improved efficiency of public expenditure. For instance, the demobilization of the security forces will control of the growth of security expenditures, which will be compensated by growth in the productive sectors. Combined with improved budget execution rates, this is expected to lead to interesting growth outlook. The recommendations provided in the report should help the Government to achieve these objectives. 34 ANNEXES Annex 1. Economic Trends, 2005-2009 2005 2006 2007 2008 2009 Population (million) 7.9 8.2 8.5 8.8 n/a Population growth (%) 3.9 3.8 3.7 3.0 n/a GDP (current prices, US$ billion) 0.8 0.9 1.0 1.2 1.3 GNI per capita (Atlas method, current prices, US$) 100.0 100.0 120.0 140.0 150 Real GDP growth (%) 0.9 5.1 3.6 4.5 3.5 Inflation (average CPI, %) 13.4 2.8 8.3 24.4 10.7 Real effective exchange rate (% change) 17.1 -2.7 -5.7 16.0 2.0 Terms of trade (% change) 11.1 -2.9 -23.4 3.4 39.8 Gross investment (% of GDP) 10.8 16.3 17.5 19.4 20.8 Gross national savings (% of GDP) 9.6 1.8 1.8 7.1 6.3 Overall fiscal balance (excluding grants, % of GDP) -16.8 -19.3 -19.8 -25.6 -32.1 Overall fiscal balance (after grants, excl. HIPC, % of GDP) -7.2 -1.8 -3.3 -3.7 -5.2 Current account balance (Incl. official transfers, % of GDP) -1.2 -14.5 -15.7 -12.3 -15.9 Overall balance of payment (% of GDP) 4.6 1.5 3.0 7.4 4.1 Gross international reserves (months of imports) 3.0 3.3 3.8 5.0 5.2 Export growth (f.o.b; US$; % change) 19.5 2.6 -9.7 15.1 21.5 Share of coffee in exports (%) 70.8 67.7 59.8 65.8 63.8 Debt service to exports ratio (%) 28.4 10.7 6.8 3.5 1.9 External debt stock (% of GDP) 183.0 165.1 155.0 134 27 Source: Burundi authorities; World Bank; IMF. *Revenue minus noninterest current expenditures and net lending, and foreign-financed capital expenditures, excluding exceptional spending. 35 Annex 2. Burundi Foreign-Financed Projects, 2007-2010 (US dollars) Titre du projet Décaissement 2007 2008 2009 2010 Prévus Réalisés Prévus Réalisés Prévus Réalisés Prévus ACBF 199,985 95,022 Allemagne 20,739,435 18,430,032 9,205,941 16,501,824 14,225,739 Arabie Saoudite 639,243 BAD 7,910,464 884,628 13,892,504 22,736,274 948,874 4,083,548 11,830,415 BADEA Banque Mondiale 57,372,473 85,009,180 56,858,845 93,018,677 43,444,296 52,871,772 105,648,267 Belgique 11,912,820 4,000,000 35,905,927 23,920,013 22,218,224 39,558,994 Canada 3,866,821 1,254,167 2,179,839 1,239,789 2,943,774 78,827,274 2,114,165 80,778,808 1,831,023 9,829,588 18,975,524 Commission de l'Union Européenne Danemark 2,584,773 2,575,844 194,885 Egypte 9,773 Espagne 424,000 238,000 278,198 229,124 FIDA 7,852,580 2,016,697 11,782,263 7,225,877 27,926,529 30,050,310 6,925,680 Finlande 638,374 Fonds Monétaire International 4,754,110 15,095,701 8,495,701 21,881,738 8,681,738 21,881,738 France 10,786,505 16,978,473 13,023,445 18,294,763 11,362,381 7,179,173 5,528,172 Irlande 2,748,225 437,637 Italie 56,564 308,975 2,371,541 Japon 16,184,186 17,483,000 411,564 845,044 2,346,982 Luxembourg 589,015 810,000 Maroc 100,000 Nations Unies 28,899,601 63,444,185 12,346,773 45,092,393 43,281,538 47,326,170 14,776,586 Norvège 11,383,311 22,654,018 200,000 21,835,513 ONG 1,013,729 11,681,906 OPEP 1,034,986 7,139,174 5,419,445 4,749,607 6,467,647 Pays-Bas 11,427,209 1,158,925 24,152,369 4,893,664 4,289,833 2,638,463 Privé 40,382 36,172 13,008 Royaume Uni 7,047,545 21,942,577 15,801,688 15,801,688 18,344,551 17,973,204 18,451,540 Slovenie 34,674 Suisse 1,275,952 2,609,154 3,521,066 3,791,244 4,219,987 4,856,174 6,454,745 Suède 2,194,493 1,135,701 2,177,068 2,177,068 États Unis d'Amérique 30,051,931 41,600 30,137,234 164,625 198,678 TOTAL 126,673,666 393,083,207 151,315,143 469,033,442 221,187,396 258,069,456 275,931,950 36 Annex3. Execution Rates and Breakdown of Public Expenditures in the National Budget, 2006-2009 2006 2007 2008 2009 Average Breakdown Breakdown Breakdown Breakdown Breakdown Invest. Budget Invest. Budget Invest. Budget Invest. Budget Invest. Budget (budget) (budget) (budget) (budget) (budget) E/C 98.8 98.8 0.5 16.3 16.3 2.7 95.6 94.3 6.6 71.0 71.0 4.9 70.4 70.1 3.7 Justice P/C 97.5 97.5 1.4 16.3 16.3 0.5 71.5 72.6 6.6 68.4 66.0 4.0 63.4 63.1 3.1 P/E 98.7 98.7 1.4 100 100 0.6 74.8 77.0 5.4 96.3 93.0 4.0 92.4 92.2 2.8 E/C 41.2 41.2 46.5 99.6 99.6 32.9 91.6 91.6 21.9 53.4 52.4 16.2 71.4 71.2 29.4 Education and Research P/C 40.1 40.1 50.1 99.3 99.3 38.1 81.4 81.4 21.4 53.4 52.4 9.9 68.5 68.3 29.8 P/E 97.3 97.3 50.7 99.7 99.7 41.6 88.9 88.9 20.2 100 100 10.5 96.5 96.5 30.7 E/C 7.7 26.4 35.3 99.8 95.8 34.1 98.2 98.7 43.5 66.2 99.5 46.3 68.0 80.1 39.8 Health and HIV/AIDS P/C 7.7 24.7 24.4 99.4 92.7 38.0 98.5 97.5 45.8 65.6 98.6 53.5 67.8 78.4 40.4 P/E 100 93.6 23.8 99.6 96.8 40.3 100.4 98.8 47.9 99.0 99.2 56.2 99.7 97.1 42.1 E/C 54.0 52.7 3.9 94.9 94.9 10.1 98.3 98.3 9.4 94.5 94.6 11.2 85.4 85.1 8.7 Agriculture P/C 50.0 49.0 5.3 65.0 65.3 11.2 97.1 97.2 9.9 71.4 72.2 12.3 70.9 70.9 9.7 P/E 92.5 93.1 5.2 68.6 68.9 8.4 98.8 98.9 10.3 75.6 76.3 10.0 83.9 84.3 8.5 E/C 33.3 33.3 6.6 34.1 34.1 10.7 63.8 63.8 7.9 76.2 76.3 10.3 51.8 51.9 8.9 Water, Energy, Mining P/C 33.3 33.3 5.7 25.8 25.8 4.2 62.3 62.3 5.4 59.5 59.7 9.1 45.2 45.3 6.1 P/E 100 100 6.0 75.6 75.6 3.5 97.7 97.7 5.6 78.1 78.2 7.6 87.8 87.9 5.7 E/C 65.4 65.4 6.2 57.1 57.1 5.9 94.9 94.9 3.6 84.0 83.2 6.9 75.3 75.1 5.6 Public Works P/C 65.4 65.4 10.5 14.2 14.2 3.9 94.9 94.9 3.7 84.0 83.2 6.7 64.6 64.4 6.2 P/E 100 100 11.0 24.8 24.8 1.1 100 100 3.9 100 100 7.1 81.2 81.2 5.7 E/C 100 100 1.0 - 98.5 3.6 - 96.6 7.1 97.4 92.6 4.2 98.7 96.9 4.0 Miscellaneous P/C 100 74.5 2.6 - 96.8 4.1 - 84.0 7.3 89.9 90.8 4.6 95.0 86.5 4.7 P/E 100 74.5 2.1 - 98.3 4.5 - 87.0 6.7 92.3 98.1 4.7 96.2 89.5 4.5 E/C 43.4 38.3 100 82.5 86.0 100 90.6 93.8 100 72.3 86.1 100 72.2 76.0 100 Total P/C 42.4 36.8 100 73.4 78.4 100 84.0 88.5 100 65.2 81.2 100 66.2 71.2 100 P/E 97.6 96.0 100 89.0 91.2 100 92.7 94.4 100 90.1 94.3 100 92.4 93.9 100 Source: Bank staff estimates based on data from the MOF. C: Credit P: payment E: Commitment 37 Annex 4: Public investment in Burundi-Evolution of credits by source of financing 2008 2009 2010 Average Source of Financing Amount (BIF) Breakdown (%) Amount (BIF) Breakdown (%) Amount (BIF) Breakdown (%) breakdown (%) Others (financed by domestic resources) 2,606,365,000 5.1 1.1 8,828,423,780 10.8 2.2 6,614,111,350 6.2 1.8 7.4 1.7 National budget (investment appearing in Pro poor (PP, financed by budget support) 11,260,352,499 22.2 4.6 28,751,155,514 35.2 7.3 20,109,528,701 18.8 5.5 25.4 5.8 HIPC 23,673,840,515 46.7 9.8 27,134,631,335 33.2 6.9 73,630,918,142 68.7 20.2 49.5 12.3 Capital Subvention (counterpart funds of projects) 13,151,469,773 25.9 5.4 16,955,289,138 20.8 4.3 6,838,616,200 6.4 1.9 17.7 3.9 LOF) From "Others" 452,117,063 0.9 0.2 476,800,000 0.6 0.1 2,487,180,761 2.3 0.7 1.3 0.3 From "Pro poor" 9,771,653,774 19.3 4.0 10,822,489,138 13.3 2.7 4,251,435,439 4.0 1.2 12.2 2.6 From "HIPC" 2,927,698,936 5.8 1.2 5,656,000,000 6.9 1.4 100,000,000 0.1 0.0 4.3 0.9 Total LOF 50,692,027,787 100 20.9 81,669,499,767 100 20.7 107,193,174,393 100 29.4 100 23.7 Development Projects Financed by Donors "Tirage sur Prets de Développement" 38,061,132,293 15.7 76,793,812,130 19.4 189,126,637,680 51.9 29.0 "Tirages sur Dons de Développement" 153,987,388,306 63.4 236,736,286,866 59.9 67,744,312,040 18.6 47.3 Total Tirages 192,048,520,599 79.1 313,530,098,996 79.3 256,870,949,720 70.6 76.3 Overall Public Investment (Domestic + External) 242,740,548,386 100 395,199,598,763 100 364,064,124,113 100 100 Source: Bank staff estimates based on data from the MOF. 38 Annex 5: HIPC and FCE cases HIPC: Case Study 1 1. One area where it is possible to monitor the execution and implementation of the investment budget is with regard to the HIPC funds. The HIPC process provides a case study to understand the implementation process, since it has been relatively well-monitored and significant data exist relating voted expenditures to actual outturns. HIPC represents a significant component of public expenditure for which there is appropriate budgetary information. 2. The National Committee for Tracking HIPC Expenditures (NCTHE) is an oversight body that maintains a dialogue with the Government on the use of HIPC funds. It was created in June 2007 to improve execution with the assistance of experts from the Directorate of Budget, and maintain a dialogue with the Government on the necessity of adhering to audit recommendations to ensure the effectiveness of the projects.22 3. The Technical Committee for Tracking HIPC Expenditures (TCTHE) plays a central role in the execution of HIPC investment projects. Created in July 2008, it is the linchpin of the NCTHE, as it provides relevant financial documents and activity reports on HIPC investment projects. It meets at least twice a month to discuss and advise, in close collaboration with the Directorate of Budget and the National Procurement Control Directorate, on possible ways to alleviate any constraints that might hinder the execution of HIPC-funded projects.23 The committee also aims to ensure that HIPC voted credits are committed, that they finance real needs, and that the projects short-listed for financing are consistent with the eligibility criteria (Box A). The committee also follows up on the implementation of audit recommendations. The institution of regular meetings between the 2nd Vice-president’s office and high- level authorities from ministries receiving HIPC funds is another initiative that has proved useful to alleviate bottlenecks to execution. 22 However, one must note that the work of the NCTHE is undermined by the irregularity of the meetings that made it very difficult to undertake planned field visits to monitor physical execution. 23 More specifically, the members (sector focal points) regularly update their knowledge with specialists from the Budget and Procurement Directorates, warn the authorities in the Ministry of Finance (MoF) about the bottlenecks in the execution of HIPC investment projects (e.g., identifying weaknesses in sector procurement cells, delays on procurement processes, etc.), share good practices in designing and applying budget management tools (e.g., dissemination of good practices in annual procurement plans“Plans Prévisionnels de Passation des Marchés Publics.” 39 Box A. Eligibility Criteria for HIPC Project Financing For the 2009 and 2010 budgets, the TCTHE used the following criteria, validated by the NCTHE, to select projects for HIPC funding: • The activity must be integrated into the sector strategy of the beneficiary ministry. If it is a pure investment project, it must be listed in the PIP. • HIPC investment projects must have an impact on the MDGs and be aligned to the poverty reduction strategy. HIPC investment expenditures should clearly contribute to the achievement of one or more MDGs. • A preliminary technical study is required to support the HIPC activity. Requests for funding should be made based on a reliable technical analysis. • The Technical Committee ensures the eligibility of all project expenditures, in conformity with the budget law. • For multi-year projects, any previous audit recommendation should be satisfied before the commitment of new funds. • Priority should be given to the payment of projects arrears before new activities are financed. • To ensure that project activities continue after the end of the HIPC financing facility, sustainability measures must be included in funding requests. • The project must demonstrate that its activities are directed to the poorest people. Priority will be given to projects in areas with high poverty rates. • The project must have a clear gender impact. 4. As of 2009, the Government has allocated HIPC resources to supervise the physical quality and timeliness of HIPC investment projects. This decision follows a recommendation of an external audit that was further re-iterated by the NCTHE. These resources are meant to increase the capacity of beneficiary sectors to undertake field supervision missions to monitor (directly or through private firms) the quality of public works being executed by contractors. The rate of execution of these specific non-investment expenditures was initially slow and erratic. The best estimates indicate that credits allocated to these expenditures reached about BIF 1.43 billion, but the cash and commitment-based execution rates were only 51.0 and 55.9 percent, respectively, at the end of fiscal year 2009. Both execution rates increased dramatically after the budget revision occurred in August and September 2009, but remained lower than 30 percent on a monthly basis.. 5. The experience of HIPC focal points shows the main barriers to efficient execution of the investment budget. First, execution was delayed by procurement procedures, requiring that a works contract be in place before a company was identified to monitor the works. These delays led to cost inflation and disruptions to the various (technical and financial) monitoring services expected. Second, contracting out such services was a new activity for the ministries, and there was a steep learning curve. Finally, few bidders came forward, either because of the tight deadlines imposed by the Procurement Code or because the services in that Sector of activities are still embryonic. 6. Overall, the quality and effectiveness of HIPC investment projects depend on ministries having the capacity to properly prepare and execute the budget. For sector ministries to come up with consistent project proposals, the Government (through the members in the TCTHE) agree 40 on the necessity to clarify the missions (and acknowledge the contributions) of HIPC focal points to support the identification of project proposals through an official notification. Moreover, high- level authorities in sector ministries have to be informed of the eligibility criteria to ensure that HIPC eligibility criteria are fully consistent with the eligibility criteria used in the elaboration of the “Lettre de Cadrage”. Finally, ministers and chiefs of cabinet should be reminded of the critical role of HIPC focal points in monitoring execution performance and facilitating the tracking of results during audits. 7. The HIPC process could be improved through a combination of measures. Regular field supervision missions are critical to guarantee that infrastructure is built as planned. Such activities could benefit from the following prior initiatives: • Provide HIPC focal points with minimal management capacity through a hands-on training program on the use of basic project management tools, including tableaux de bord to monitor multi-year projects, as well as procurement plans and key elements of financial and economic project appraisal. • Provide HIPC focal points with periodic training on procurement practices (led by experts from the National Procurement Control Directorate), which they can apply directly as members of procurement units. • Facilitate annual field visits by allowing the HIPC budget voted for field visit purposes to be executed as the project progresses. FCE: Case Study 2 8. The Fonds Commun de l’Education (FCE) or Common Fund for Education is one of Burundi’s first Sector-Wide Approach (SWAp) initiatives. It is supported mainly by the Belgian aid agency, with DFID and the World Bank in charge of managing its Secrétariat Exécutif Opérationnel (SEO). The overall objective of the FCE is to support the Government in implementing its Sector Plan for the Development of Education and Training (PSDF). Having started in January 2008 with no institutional arrangements or rules regarding procurement or project execution, the FCE performed poorly in 2009 (Box B); for example, many textbooks for primary school were unavailable in time for the school year. A lack of clear communication between the donors and the education ministry, especially in relation to procurement, compounded the difficulties and worsened Burundi’s progress towards the MDGs. Since mid-2009, the execution of operations has increased significantly, reaching about 23 percent of the 4 million euros in the pivot account as of February-March 2010. Future consolidation of this current trend would be a good signal of improved donor coordination in Burundi. 41 Box B: The SWAp in the Education Sector in Burundi The objective of the Common Fund for Education (FCE) is to support implementation of the Government’s Sector Plan for the Development of Education and Training (PSDEF). This strategy has two major objectives: • Improve access to all levels of education, particularly at primary level, to achieve the goal of education for all by 2015. • Improve the quality of teaching in all its dimensions through the targeted provision of educational resources and training for teachers. In complementarity to domestic funds, global budgetary support and other external support, the resources of FCE are intended to finance all levels of education, from preschool to higher education. This includes financing the construction and rehabilitation of schools; procurement of furniture and equipment, textbooks, and other instructional materials; training of teachers and administrators, and supporting school operations, including the payment of teachers’ salaries. The Secrétariat Exécutif Opérationnel (SEO) is the backbone of the FCE’s technical and financial activities. Composed of five members, it is responsible for coordinating the preparation of activities and acts as a focal point for all donors (“Partenaires Techinques et Financiers” - PTF). It coordinates the development of work programs (activities) and annual budgets (Plans de Travail et Budget Annuel - PTBA), strengthens the action plans of various PTF, oversees the implementation of activities, and maintains the accounts of external funding for control and auditing. Under the coordination of the SEO, the “Comité Technique d’Exécution” (CTE) is responsible for preparing annual business plans and corresponding budgets and implementing programs of activities. The CTE consists of four units, each responsible for a series of specific activities under a different branch of the ministries of education and training. The PTF contributes to the FCE for a period of three years. The amount of annual contributions are secured through grant/funding agreements. Contributions from PTF are paid to a joint account, called a "pivot account," at the central bank and denominated in euros. The account feeds into an "operations account," also housed at the central bank and denominated in euros. Funds are replenished through a refunding request submitted by by the SEO. In the absence of institutional arrangements, administrative proceedings, and rules for execution of operations and expenses, very few of planned activities were conducted in 2009. The implementation rate of the fund was about 4 percent after about three years of operation. The main causes of such poor performance were: • Delays in procurement. At the end of 2009, out of the 21 activities planned, only 4 had been realized. Many, including support for training, operation of services, compensation (incentives and bonuses), and grants to schools, were never launched. • Ministries’ feelings of exclusion in the execution of operations, despite their involvement in the development of action plans. • Lack of staff training for the Projet d’Appui à la Reconstruction du Système Educatif (Education Support Project, PARSEB), to enable them to properly manage their workload. • Inertia caused by poor sharing of responsibilities between ministries and the SEO, especially for procurement. The institutional design places all players at the same level of decision making. • Lengthy delays in obtaining notices of no-objection from the DNCMP (e.g., 85 days for the purchase of printers and copiers), and from committees analyzing bids (a purchase of benches and desks has been awaiting approval since July 2009). • Delays in the transmission of files. • A burdensome procurement process that requires specific technical advice in complex cases to support decision making. • Underuse of the technical and financial capacities in the ministries charged with implementing PSDEF, which are far superior to those available in PARSEB. 42 43