Report No: ACS5393 . Republic of Burundi Burundi Public Expenditure Review Strengthening Fiscal Resilience to Promote Government Effectiveness . December 2, 2013 . Poverty Reduction and Economic Management Network (PREM) Great Lakes Unit (AFTP5) Africa Region The World Bank . . Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: . The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978- 750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover photo: Burundi drummers perform during a traditional ceremony in Ninga, Burundi. Photograph by Bruno De Hogues/Getty Images. Cover design: Printing and Multimedia General Services, World Bank. Republic of Burundi Public Expenditure Review Table of contents Abbreviations and Acronyms ........................................................................................................... I Acknowledgments.......................................................................................................................... XI List of Figures ................................................................................................................................. III List of Tables .................................................................................................................................. VI List of Boxes ................................................................................................................................. VIII Executive Summary ............................................................................................................... 1 Structure and Approach....................................................................................................... 12 Chapter 1 : Macroeconomic and Fiscal Developments .......................................................... 14 1.1 Recent Economic Developments and Trends ..................................................................... 14 1.1.1 GDP growth................................................................................................................... 14 1.1.2 Balance of payments .................................................................................................... 16 1.1.3 Inflation and monetary policy ...................................................................................... 18 1.1.4 Fiscal policy ................................................................................................................... 19 1.2 Risks and Challenges Ahead ................................................................................................ 26 Chapter 2 : Drivers of Fragility ............................................................................................. 33 2.1 High Dependency on Foreign Aid and its Volatility............................................................. 33 2.2 The Volatility of Budget Support ......................................................................................... 36 2.3 Growing Stock of Arrears .................................................................................................... 38 2.4 Critical Level of Reserves ..................................................................................................... 39 Chapter 3 : Fiscal Sustainability Analysis .............................................................................. 51 3.1 Key Findings and Assessment.............................................................................................. 51 3.2 The Baseline Scenario Analysis ........................................................................................... 52 3.2.1 Macroeconomic assumptions ...................................................................................... 52 3.2.2 Fiscal policy ................................................................................................................... 53 3.2.3 Baseline results ............................................................................................................. 53 3.3 Fiscal Risk Analysis............................................................................................................... 55 3.4 Policy Scenario Analysis ...................................................................................................... 60 3.4.1 Scenario 1. Partial policy implementation ................................................................... 60 3.4.2 Scenario 2. Reduction of tax incentives toward regional standards ............................ 62 3.4.3 Scenario 3. Options for civil service pay reform........................................................... 63 Chapter 4 : Public Expenditure Analysis ............................................................................... 67 4.1 Data Sources and Data Quality Issues ................................................................................. 67 4.2 Domestic Financing ............................................................................................................. 70 4.2.1 Public expenditures by economic classification ........................................................... 71 4.2.2 Public cxpenditures by administrative/sector classification ........................................ 74 4.3 External Financing ............................................................................................................... 78 III 4.3.1 Framework of aid co-ordination................................................................................... 78 4.3.2 Analysis of aid allocations and expenditures (2007-2011)........................................... 78 4.3.3 Aid management and reporting: key challenges.......................................................... 82 4.3.4 Key recommendations .................................................................................................. 83 4.4 Selected Non-salary Public Expenditures............................................................................ 84 4.4.1 Pro-poor priority programs .......................................................................................... 84 4.4.2 Areas aside from pro-poor ........................................................................................... 86 4.4.3 Domestically financed capital expenditures ................................................................ 87 4.4.4. Miscellaneous Expenditures ........................................................................................ 88 4.5 Pro-poor Budget Spending .................................................................................................. 89 4.6 Experience with MTEF: a stock-tacking exercise ................................................................ 91 Chapter 5 : Confronting Competing Pressures for Reform: a case-study of the wage bill and pay policy ............................................................................................................................ 98 5.1 The Scope of the Problem ................................................................................................... 98 5.2 Recent Developments in Government Wages and Employment ....................................... 99 5.3 Complexities and Inequities of Public Pay ........................................................................ 102 5.4 Inequities in Wages ........................................................................................................... 104 5.4.1 The proliferation of bonuses and allowances ............................................................ 106 5.5 Governance and Institutional Arrangements.................................................................... 107 5.5.1 Personnel management is not strategic..................................................................... 107 5.5.2 Career progression is not tied to performance .......................................................... 108 5.6 Management of Pay Policy................................................................................................ 109 5.7 Realistic and Feasible Reform Options: a proposal .......................................................... 110 5.8 Concluding Remarks .......................................................................................................... 114 Concluding Remarks .......................................................................................................... 115 References .................................................................................................................................. 116 Annex A : GoB-donor PER Technical Working Group – TOR ...................................................... 123 Annex B : Arrears and Pending Payment .................................................................................... 127 Annex C : Regression Results Underlying FSA............................................................................. 135 Annex D : Wage Bill Management Study – background data ..................................................... 139 Annex E : Budget Classification and Data Quality Issues ............................................................ 157 Annex F : Public Expenditure Analysis- background tables ........................................................ 161 Annex G : Aid Management – supplementary analysis .............................................................. 169 Annex H : Pro-poor Spending – supplementary analysis............................................................ 174 Annex I : Burundi Financial and Economic Statistics ................................................................. 180 Annex J : List of People Consulted during the PER preparation ................................................. 180 Map of Burundi ........................................................................................................................... 195 IV List of Figures Figure 1.1: Annual real GDP growth – Burundi in comparative perspective .............................................. 15 Figure 1.2: Burundi’s balance of payments, 2000-2012 ............................................................................. 17 Figure 1.3: Burundi’s inflation and monetary policy, 2004-2012 ............................................................... 19 Figure 1.4: Tax revenues breakdown (average 2007-2011) ....................................................................... 20 Figure 1.5: Tax revenues growth rate ......................................................................................................... 21 Figure 1.6: Tax revenues as percentage of GDP ......................................................................................... 21 Figure 1.7: Grants and Total Expenditures, 2006-2012 ............................................................................. 23 Figure 1.8: Total Expenditures, as percent of GDP ..................................................................................... 24 Figure 1.9: Composition of current spending, as percent of GDP .............................................................. 24 Figure 1.10: Civilian and security sectors' wage bills, as a share of total, 2001-2012 ................................ 24 Figure 1.11: Burundi Governance at a glance, 2011 data ........................................................................... 31 Figure 1.12: Burundi CPIA, 2005 - 2011x .................................................................................................... 32 Figure 1.13: Burundi CPIA in comparative perspective (2011) ................................................................... 32 Figure 2.1: Burundi in comparative perspective: ODA to GDP ratio........................................................... 34 Figure 2.2: Burundi in comparative perspective: Per capita Aid inflows .................................................... 34 Figure 2.3: ODA to Fragile States, 2011 data .............................................................................................. 34 Figure 2.4: Remittances inflows .................................................................................................................. 35 Figure 2.5: Aid inflows in Burundi ............................................................................................................... 35 Figure 2.6: Burundi growth rates of projects and program support, 2002-2012 ....................................... 36 Figure 2.7: Burundi: Budget Support as a Percentage of Total Budget, 2007 - 2012 ................................. 37 Figure 2.8: Budget Supports Received at the Central Bank (cumulative flow), 2003 - 2012...................... 36 Figure 2.9: Revenues and Budget support (flows), 2011 - 2012 ................................................................. 37 Figure 2.10: Government’s Net Position with Domestic Financial Corporation’s 1/, 2003 - 2012............. 37 Figure 2.11: Stock of arrears (millions of FBu) ............................................................................................ 38 Figure 2.12: Reserves Hold at the BRB and Government’s Position with the BRB, 2003-2012.................. 40 Figure 2.13: Reserves as a share of GDP ..................................................................................................... 41 Figure 2.14: Reserves as a share of M2 ...................................................................................................... 41 Figure 2.15: Reserves to Month of Imports ................................................................................................ 41 Figure 2.16: Composition of Burundi’s imports .......................................................................................... 42 Figure 2.17: Composition of Burundi’s exports .......................................................................................... 42 Figure 2.18: Destination of Burundi’s exports (share of total) ................................................................... 45 Figure 2.19: International price of Arabica coffee FBu), constant U.S. cents per Kg .................................. 46 Figure 2.20: Price received by Burundian coffee producers vs. world price (in constant FBu) .................. 46 Figure 2.21: Spread between produce and market prices, 2008................................................................ 47 Figure 2.22: Long term evolution of rainfall in Burundi (millimeter; yearly average) ................................ 49 Figure 2.23: Retail food prices; percent increases 2011-2012 ................................................................... 50 Figure 2.24: Retail price of beans in Burundi markets ................................................................................ 50 Figure 2.25: Retail price of cassava powder in Burundi markets................................................................ 51 Figure 2.26: Retail price of rice in Burundi markets ................................................................................... 51 Figure 3.1:Changes in Growth Forecasts .................................................................................................... 55 Figure 3.2: Fiscal indicators under Stochastic Coffee Price Shocks ............................................................ 57 Figure 3.3 Burundi: Annual Rainfall and Agriculture Growth Rates ........................................................... 57 Figure 3.4: Fiscal Indicators under Stochastic Rainfall Shocks .................................................................... 58 Figure 3.5: Fiscal Indicators under Stochastic Aid (Grant) Shocks .............................................................. 60 Figure 3.6: Government Revenue excluding grants (percent of GDP)........................................................ 61 Figure 3.7 Comparisons of Alternative Reform Scenarios: Baseline vs. Scenario 1 & 2 ............................. 61 Figure 3.8: Comparisons of Alternative Reform Scenarios: Baseline vs. Options 1 -4................................ 66 V Figure 4.1: Total Expenditures .................................................................................................................... 70 Figure 4.2: Composition of spending .......................................................................................................... 70 Figure 4.3: Breakdown by expenditure category (actuals) ......................................................................... 71 Figure 4.4: Execution rate (LR/LFI) .............................................................................................................. 72 Figure 4.5: Capital Expenditure by source of financing .............................................................................. 72 Figure 4.6: Mapping key donor’s engagement and disbursements (2010-2011)....................................... 80 Figure 4.7: PRSP focus areas, external funding........................................................................................... 81 Figure 4.8: Disbursements by Ministry, 2007-11 (million of USD).............................................................. 81 Figure 4.9: Where does the pro-poor budget go? ...................................................................................... 90 Figure 4.10: Execution rate of the pro-poor budget ................................................................................... 91 Figure 4.11: Inconsistencies within the MTEF document ........................................................................... 94 Figure 4.12: Percentage difference between expenditures estimates and approved budget. .................. 95 Figure 4.13: MTEF Estimates vs. Approved Budget (current expenditure) ................................................ 95 Figure 4.14: MTEF Estimates vs. Approved Budget (capital expenditure)................................................. 96 Figure 5.1. Sub-Saharan Africa Wage Bills, 2011 (percent of GDP) ............................................................ 99 Figure 5.2: Wage Bill vs. GDP real growth (2007=1.0) .............................................................................. 101 Figure 5.3: Compensation by Ministry, (percent; 2012) ........................................................................... 107 Figure B.1: Simplified explanation of treasury balance movements ........................................................ 129 Figure G.1: External funding evolution per priority sector ....................................................................... 173 Figure H.1: How relevant is the pro-poor budget? ................................................................................... 174 Figure H.2: How continuous is the pro-poor budget? .............................................................................. 175 Figure H.3: Where does the pro-poor budget go?.................................................................................... 177 Figure H.4: Current expenditure breakdown of pro-poor budget ............................................................ 177 Figure H.5: Execution rate of the pro-poor budget .................................................................................. 178 Figure H.6: Execution rate of the pro-poor budget in priority sectors ..................................................... 179 List of Tables Table 1.1: Economic Trends, 2007-2014 ..................................................................................................... 16 Table 1.2 Burundi: Revenue, Main Categories as a Percent of GDP, 2007-2012........................................ 21 Table 1.3 Monthly tax collection between January and May, 2010-2013.................................................. 28 Table 2.1 The world’s most aid-dependent countries (fragile states are in bold) ...................................... 33 Table 2.2 Burundi’s top 10 imports and exports, by HS 2-digit chapters (2008 – 2010) ............................ 43 Table 2.3 Burundi's food balance in recent agricultural seasons (1000 ton of cereal equivalent) ............ 47 Table 3.1 Baseline Projections: Macroeconomic Assumptions .................................................................. 52 Table 3.2 Medium-Term Fiscal Projection: Baseline Scenario .................................................................... 54 Table 3.3 Sensitivity of Burundi’s Debt to Small Changes in Macroeconomic Forecasts ........................... 55 Table 3.4 Salary Scale Adjustment Options ................................................................................................ 63 Table 4.1 Expenditure chain 2009 .............................................................................................................. 67 Table 4.2 Burundi: Actual expenditures, 2001-2012 .................................................................................. 71 Table 4.3 TOFE: Budget Execution Rates, 2009-2011 ................................................................................. 74 Table 4.4 Priority Sectors as Share of Total Budget (1) .............................................................................. 75 Table 4.5 Execution Rates, 2009-11 (billions of FBu) .................................................................................. 75 Table 4.6 Budget execution rates by Ministries (2010-2012) ..................................................................... 76 Table 4.7 Engagement and disbursement per donor (2010-2011)............................................................. 79 Table 4.8 Project recorded in the Finance Law and CNCA data………………………………………………………………81 Table 4.9 Executed project recorded in the Settlement Law and CNCA data………………………………………….82 Table 4.10 Ring fenced programs in Burundi's post-conflict context (billion FBu)………………………………….83 Table 4.11 Three expenditures with no prior authorization and official rents (billion FBu) ………………….. 85 VI Table 4.12 Burundi's capital spending in electricity, road and state transport (billion FBu)……………………86 Table 4.13 Prioritized expenditures with unclear pro-poor impacts (billion FBu)……………………………………87 Table 4.14 Pro-poor budget distribution by administrative classification of spending……………………………89 Table 4.15 Pro-poor budget allocation in priority sectors………………………………………………………………………89 Table 5.1 Size of the Wage Bill (percent of GDP) ...................................................................................... 100 Table 5.2 Size of wage bill in recurrent budget (percent)......................................................................... 101 Table 5.3 Public Sector Workforce............................................................................................................ 102 Table 5.4 Classification Schemes in Burundi, 2012 ................................................................................... 102 Table 5.6 Magistrates, Military, and Police pay grids, 2012 ..................................................................... 103 Table 5.7 Structure of the Workforce, 2012 ............................................................................................. 103 Table 5.8 Staffing and Salaries, 2010-2012 (FBu) ..................................................................................... 104 Table 5.9: Monthly Remuneration for Select Jobs, 2008 (FBu) ................................................................ 105 Table 5.10 Comparison between General Statute and Autonomous Agencies, 2012 (FBu) .................... 105 Table 5.11 Monthly Compensation of Political Appointments, 2012 (FBu) ............................................. 105 Table 5.12 Annual Wage Increases and Promotions, 2012 ...................................................................... 110 Table 5.13 Adjustment Scenarios, fiscal cost per year ............................................................................. 111 Table B.1. Balance of the account Fournisseur, 2008 ............................................................................... 127 Table B.2 Balance of the account Fournisseur, 2009 ................................................................................ 128 Table B.3 Pending payments: stocks and flows, 2008 - 2012 ................................................................... 132 Table B.4 Income, Expenditures, and Balances of the Burundi Government, millions of FBu ................. 134 Table C.1 Unit Root Tests: Augmented Dickey-Fuller and Phillips-Perron Tests ...................................... 137 Table C.2 Regression Results for Coffee Price Shocks .............................................................................. 137 Table C.3 Regression Results for Rainfall Shocks ...................................................................................... 138 Table D.1 Distribution of the workforce across the pay grade ................................................................. 139 Table D.2 Option 1 proposed wage scale ................................................................................................. 140 Table D.3 Option 2 proposed wage scale ................................................................................................. 140 Table D.4. Option 3 proposed wage scale ................................................................................................ 140 Table D.5 Option 4 proposed wage scale ................................................................................................. 141 Table D.6 Maximum salary adjustment .................................................................................................... 142 Table D.7 Average salary adjustment ....................................................................................................... 143 Table D.8 Reference salary adjustment .................................................................................................... 144 Table D.9 Combined adjustment .............................................................................................................. 144 Table D.10 Civil Service Census April 2008 ............................................................................................... 145 Table D.11 Public Staffing and Salaries, January 2012.............................................................................. 146 Table D.12 Salaries by Ministry, 2012 ....................................................................................................... 155 Table D.13 List of Bonuses and Allowances .............................................................................................. 156 Table E.1 Personnel payments, 2012 ........................................................................................................ 160 Table F.1. Comparison of budget executions by Source (billions of FBu)................................................. 161 Table F.2 Expenditure Chain, 2009 ........................................................................................................... 162 Table F.3 Expenditure Chain, 2010 ........................................................................................................... 163 Table F.4 Expenditure chain, 2011 ............................................................................................................ 164 Table F.5 Expenditure chain, 2012 (prel.) ................................................................................................. 165 Table F.6 Items that can be paid through special procedures .................................................................. 166 Table G.1 External disbursement evolution per ministry (2007-2011) .................................................... 172 Table H.1 Pro-poor budget distribution by administrative classification of spending ............................. 176 Table H.2 Pro-poor budget allocation in priority sectors ......................................................................... 176 Table I.1.a Nominal GDP in billion Fbu, 2005-2012………………………………………………………………..…………….180 Table I.1.b Real GDP in billion FBu, 2005-2012…………………………………………………………………………………....182 Table I.2 Monetary survey, 2005-2012…………………………………..……………………………………………………………184 VII Table I.3 Balance of payments, 2005-2012…………………………………………………………………………………………..187 Table I.4 Fiscal operations, 2008-2012……………………………………………………………………………………………..….189 Table I.5 Inflation rate………………………………………………………………………………………………………………………….190 Table I.6 Exchange rate………………………………………………………………………………………………………………………..190 Table J.1 Government of Burundi ............................................................................................................. 191 Table J.2 Development Partners ............................................................................................................... 194 List of Boxes Box 1............................................................................................................................................................ 22 Box 2............................................................................................................................................................ 25 Box 3............................................................................................................................................................ 39 Box 4............................................................................................................................................................ 40 Box 5............................................................................................................................................................ 47 Box 6............................................................................................................................................................ 69 Box 7............................................................................................................................................................ 73 Box 8............................................................................................................................................................ 91 Box 9............................................................................................................................................................ 93 Box 10 ......................................................................................................................................................... 97 Box 11 ....................................................................................................................................................... 100 Box 12 ....................................................................................................................................................... 112 Box 13 ....................................................................................................................................................... 169 VIII PREFACE The present Public Expenditure Review was prepared by the World Bank, in close collaboration with the Government of Burundi and with the financial support of the Royal Embassy of the Netherlands, under the overall guidance of the Ministry of Finance and Economic Development Planning. This collaboration was formalized through the establishment – by ministerial ordinance – of the “PER Technical Working Group”, which acted as a forum for discussion and dialogue, facilitating the preparation of the report and the agreement on possible ways forward. This public expenditure review highlights the significant progresses made by Burundi over the past few years, including: the approval of the new Constitution in 2005 emphasizing the principles of power sharing and local development; the successful management of two rounds of democratic elections in 2005 and 2010, coupled with the implementation of the Disarmament Demobilization and Reintegration processes of ex-combatants and rebel groups. Moreover, Burundi has maintained a stable macroeconomic environment to allow and encourage rebuilding core public sector institutions. The most notable milestone of the last few years was the improvement of indicators in the sectors of health, education and gender. In spite of recent progresses, the report notes that economic growth remains weak and GDP rates are below expectations. Compared to neighboring countries in the East-African Community (EAC) and other post-conflict countries, economic growth is modest. With a population growth averaging 3% and an expected economic growth rate in the medium term at 5%, the country will be unable to reduce the high poverty rate and improve life conditions for its population. Burundi is therefore vulnerable to different sources of shocks, especially external as a result of increasing prices of imported food and fuel, the price volatility of tea and coffee exports, climate change, and declining aid transfers. Given the vulnerable country context, social fragility could create political instability and therefore undermine all previous gains. In this context, the main objectives of this Public Expenditure Review is to assist the Government in its efforts to escape fragility by focusing on two important and complementary themes: (i) creating adequate fiscal space to reduce the impact of future shocks; and (ii) using the available fiscal space to improve public expenditure management and promote government effectiveness. Building on these topics, the report emphasizes six key messages : (i) Burundi has a narrow fiscal space given the structure of its public expenditures, composed mainly of salaries and transfers; (ii) the country should mobilize additional domestic revenues to increase its fiscal space; (iii) the role of donors in supporting effective management of public expenditures remain critical and can also be the source of additional fiscal stress, especially when budget support arrives late in a fiscal year; (iv) accessing high quality budget data remains a challenge due to major incoherencies between data from the Treasury database and data published by SIGEFI; (v) the aid management is not properly integrated into budget execution process; (vi) it is advisable to adopt realistic and pragmatic solutions to improve the performance of the public administration, while keeping the wage bill under control. IX The report then gives a series of policy recommendations to the Burundi Government in order to strengthen its fiscal policies, build resilience and escape the fragility trap. This report has benefited from valuable comments received from government institutions and Civil Society Organizations, resulting in an inclusive process worth noting. I therefore take this opportunity to acknowledge and thank all the associations and offices that collaborated on this project. For the Government, implementing the recommendations outlined in this report will be necessary to improve the structural weaknesses in the management of the public expenditures. It is also worth noting that their successful implementation will depend on donors’ capacity to honor their commitments and transfer aid timely. His Excellency Tabu Abdallah Manirakiza Minister of Finance and Economic Development Planning X ACKNOWLEDGMENTS The Burundi PER is part of a series of programmatic policy studies developed by the World Bank to assist the government in its efforts to promote accelerated and inclusive growth and improve transparency and accountability in the use of public expenditures. While the Government of Burundi is the primary beneficiary of the analysis, the studies are intended for a wider audience, including members of the donor community, business leaders, representatives of civil society groups and other stakeholder engaged in Burundi’s economy. The report has been prepared jointly with the Government of Burundi and the team is most grateful to its government counterparts for this fruitful collaboration. The team would like to thank His Excellency Mr. Tabu Abdallah Manirakiza (Minister of Finance and Economic Development Planning, Burundi) and the members of the PER ‘Technical Working Group’ for their critical contributions during several rounds of consultations and workshop discussions convened to validate the initial findings of this study. Specifically, we would like to thank Mme Immaculée Bigirimana (Secrétaire Permanente, MFEDP); Mr. Jacques Ngendakuman (Coordonnateur de la Cellule de Coordination des réformes des finances publiques, MFEDP) ; Mme Aude Toyi (Directrice du Bureau des Grands Contribuables, OBR) ; Mme Béatrice Samandari (Directrice du Budget) ; (Mme Rose Kankindi, Directrice de la Paie, Ministère de la Fonction Publique) ; Mr Emile Nimpaye (Économiste, CNCA) and Mr Gérard Muringa (Conseiller en coopération et en finances publiques, Ambassade du Royaume des Pays-Bas). The Bank task team was led by Marco Larizza (TTL, Public Sector Specialist, AFTP5) and included Aurelien Serge Beko (co-TTL, Poverty Economist, AFTP5), Ferdinand Bararuzunza (Resident Economist, AFTP5/AFMBI), Mizuho Kida (Economist, PRMED), Zachary Mills (Public Finance Specialist, ECSP4), Francisco Vazquez Ahued (Consultant, WBI), Diana Beatriz Ayala Pena (Consultant, WBI), Brendan James Glynn (Consultant, AFTP5), Jean-Claude Nsabimana (Local Research Assistant, AFTP5) and Lyse Kaniambo (Team Assistant, AFMBI). The Fiscal Sustainability Analysis benefitted from comments and feedback received by the following PRMED colleagues: Juan Pradelli (Economist); Ralph Van Doorn (Economist); Bill Battaile (Senior Economist) and Naoko Kojo (Senior Economist). The Wage Bill management study was discussed with members of the Public Employment & Management (PE&M) Thematic Group, and valuable comments were received by Nick Manning (Head, Governance & Public Sector Management Practice), Sanjeev Ahluwalia, Senior Public Sector Specialist (AFTP5), Michael Stevens (Consultant/Public Administration Specialist) and Malcom Green (Consultant/Public Administration Specialist). Credit is also due to a distinguished group of peer reviewers who provided comments on the draft report: Zahid Hasnain, Senior Public Sector Specialist (EASPW); Douglas M. Addison, Senior Economist (EASPW); Leif Jensen (Senior Public Sector Specialist, PRMPS); and Oral Williams (Mission Chief for Burundi, IMF). Overall technical oversight and guidance has been provided by Albert Zeufack (Sector Manager, AFTP5), Jacques Morisset (Lead Economist and Sector Leader, AFTP5) and Chiara Bronchi (Lead Public sector and Governance Cluster Leader). The team is also grateful to Rachidi Radji (Country Manager, AFMBI), whose advice and warm support has been extremely useful to move forward the preparation process and engage in effective dialogue with government authorities. The task team remains responsible for any errors or omissions. XI ABBREVIATIONS AND ACRONYMS AfDB African Development Bank API Agence de promotion des investissements (Investment Promotion Agency) BEI Budget Extraordinaire d’Investissement (extraordinary investment budget) BGTE Bureau de Gestion de la Trésorerie de l’Etat (State Cash Management Office) BO Budget Ordinaire (ordinary budget) BRB Banque de la République du Burundi (Central Bank of Burundi) CAS Country Assistance Strategy CBMT Cadre Budgétaire à Moyen Terme (Medium-term Budget Framework) CdC Cour des Comptes (Audit Office) CDE Comptables des dépenses engagées (Financial Controllers) CDMT Cadre de Dépenses à Moyen Terme (Medium Term Expenditure Framework) CEM Country Economic Memorandum CMU Country Management Unit CNCA Comité National de Coordination des Aides (National Aid Coordination Committee) CPI Consumer Price Index CPIA Country Policy and Institutional Assessment CSLP Cadre Stratégique de Croissance et Lutte Contre la Pauvreté (Strategic Framework on the Fight against poverty DBC Direction Générale du Budget et Control (General Directorate of Budget and Control) DDI Directions des Douanes et des Impôts; DDR Demobilization, Disarmament, and Reintegration DeMPA Debt Management Performance Assessement DFID Department for International Development DGBCP Direction Générale du Budget et de la Comptabilité Publique (General Directorate of Budget and Public Accounting) DGP Direction Générale de Planification (General Directorate of Planning) DGR Direction Générale de Recettes (General Directorate of Public Revenue) DPI Debt Performance Indicator DPO Development Policy Operation DSA Debt Sustainability Analysis DT Direction de la Trésorerie (Treasury Directorate) EAC East African Community EERC Emergency Economic Rehabilitation Credit EMSP Economic Management Support Project (Projet d’Appui à la Gestion Economique, PAGE) ERSG Economic Reform Support Grant EU European Union FDI Foreign Direct Investment FBu Burundi Franc FSA Fiscal Sustainability Analysis GDP Gross Domestic Product GFS Government Finance Statistics XII GGTE Groupe de Gestion de la Trésorerie de l’Etat (State Cash Management Group) GoB Government of Burundi HDI Human Development Index HIPC Heavily Indebted Poor Countries HIV/AIDS Human Immunodeficiency Virus/ Acquired Immunodeficiency Syndrome HRMIS Human Resources Management Information System ICA Investment Climate Assessment ICR Implementation Completion and Results Report IDA International Development Association IDF Inspection des Finances (Sectoral inspection body) IFMIS Integrated Financial Management Information System IGE Inspection Générale de l’Etat (State Inspectorate General) IGF Inspection Générale des Finances (Finance Inspectorate General) IMF International Monetary Fund IPT Income Personal Tax ISTEEBU Institut de Statistiques et d’Etudes Economiques du Burundi LCU Local Currency Unit LICs Low Income Countries LIFC Low-Income Fragile Country MDG Millennium Development Goal MDRI Multilateral Debt Relief Initiative MDTF Multi Donor Trust Fund MFEDP Ministry of Finance and Economic Development Planning MoEPR Ministry of Economic Planning and Reconstruction MFP Ministère de la Fonction Publique (Ministry of Public Service) MoF Ministry of Finance MTEF Medium-Term Expenditure Framework NGO Non-Governmental Organization NPF National Police Force OBR Office Burundais des Recettes (Burundi Revenue Authority) OCIBU Office de Café du Burundi (Coffee Board of Burundi) ODA Official Development Assistance OECD Organization for Economic Cooperation and Development OTBU Ordonnateur Trésorier du Burundi (Burundi Payment orders Head) PBCE Plan Budgétaire et Comptable de l’État (Charter of Accounts and Budget) PEFA Public Expenditure Financial Accountability PEM Public Expenditure Management PEMFAR Public Expenditure Management and Financial Accountability Report PER Public Expenditure Review PER-TWG Public Expenditure Review Technical Working Group PETS Public Expenditure Tracking Survey PFM Public Finance Management PIP Public Investment Program PLF Projet de Loi de Finance (Draft Budget Law) PNCE Plan National de Comptabilité de l’Etat (National Plan for Public Accounting) PNRA Programme National pour la réforme administrative (National Program for Administrative Reform) PPP Purchasing Power Parity XIII PRSP Poverty Reduction Strategy Paper SDR Special Drawing Rights SIGEFI Système de Gestion des Finances Publiques (Integrated Financial Management Information System) SODECO Coffee Milling and Processing Company (Société de Déparcharge et de Conditionnement) SOEs State-Owned Enterprises SOGESTAL Coffee Washing Stations Management Corporation (Société de Gestion des Stations de Lavage de Café) SSA Sub-Saharan Africa TA Technical Assistance TOFE Tableau des Opérations Financières de l’Etat (Table of Government Financial Operations) TOR Terms of Reference TOT Terms of Trade UN United Nations UNDP United Nations Development Program USAID United States Agency for International development VAT Valued Added Tax WB World Bank WFP World Food Program WHO World Health Organization XIV CURRENCY AND EQUIVALENT UNITS (Exchange Rate Effective May 15, 2013) Currency Unit = Burundi Franc (BFU) US$ 1.00 = BFU 1,515 FISCAL YEAR January 1 – December 31 Vice President : Makhtar Diop Country Director : Philippe Dongier Sector Director : Marcelo Giugale Sector Manager : Albert Zeufack Task Team Leader : Marco Larizza co-TTL : Aurelien Serge Beko XV EXECUTIVE SUMMARY Rationale and Objectives 1. Burundi has made tremendous progress over the last decade, and today, has the potential to transform its economic development trajectory from a fragile post-conflict equilibrium to a resilient growth-focused path. Since the Arusha Peace agreement (2000), Burundi has made remarkable achievements in consolidating peace and security, including the approval of the new Constitution (2005) emphasizing power sharing and local development, two rounds of democratic elections (2005, 2010), the demobilization and integration of former combatants and rebels groups, the achievement of the Heavily Indebted Poor Countries (HIPC) completion point (2009) with subsequent improvements on debt management, and the re- building of core public sector institutions. 2. The World Bank has continued to be closely engaged with the GoB national re- stabilization and development effort, with notable attention and resource commitment to (i) strengthening public financial management; (ii) improving the business environment; and (iii) restructuring and liberalizing the export crop sector. Series of dialogues have produced significant results, including the modernization of public finance management policies; the introduction of improved Public Financial Management (PFM) procedures and systems (budget framework law (loi organique des finances publiques), procurement code, computerized financial management information system); the reallocation of substantial share of available resources toward education, health, and other priority sectors; and the approval of comprehensive legislative reforms, which helped to create a more favorable business environment and triggered a modest growth of private investments. 3. Despite these achievements, Burundi still faces a range of risks conspiring to create a ‘slow-growth-poor governance equilibrium’ or ‘fragility trap’ into which the economy can fall (Devarajan et al. 2011). Chief amongst them are: (i) food price volatility linked to international markets and climatic conditions: (ii) oil prices that affect the import bill of the country; (iii) high level of dependency on aid inflows; (iv) low budget execution capacity coupled with weak state effectiveness and under-performing public institutions; (v) constraints on productive activity and a highly concentrated export profile; and (vi) enduring risks of political and social instability linked to the return of refugees and its impact on land conflicts and security threats. 4. The present PER builds upon previous studies 1 and aims to assist government’s efforts to escape fragility by focusing on the following ‘twin’ themes: (i) creating adequate fiscal space to reduce the impact of future shocks; and (ii) using the available fiscal space to improve public expenditure management and promote government effectiveness. Given its 1 This PER appears within a sequence of analytical studies undertaken in the post-conflict period to improve the quality of public spending and maximize the development impact in Burundi. Most recently, important findings and recommendations were added in the deepening of this process of constructive dialogue through inter al., the 2008 PEMFAR, the 2008/11 PEFAs and the 2010/11 World Bank Policy Note Series. 1 thematic focus, the PER is intentionally highly selective and organized by chapters, which are designed to read as stand-alone pieces while contributing to the overall PER objectives. Specifically, Chapter 1 reviews recent macro-economic and fiscal developments and outlines major risks to the economic outlook. Chapter 2 provides a systematic investigation of key drivers of fragility (structural and cyclical), while Chapter 3 analyzes the impact of these risks on the country’s deficit and debt, and outlines policy options to mitigate them in the medium-term. Chapter 4 analyzes the current status of public expenditure management and identifies key channels through which allocative efficiency and improvement in public spending can be achieved; chapter 5 provides an in-depth analysis of the largest category of current expenditures (wage bill) to better highlight the trade-offs between the imperatives of fiscal stability and the urgent need to address human resource capacity constraints in the public administration. Main messages 5. The report finds that Burundi does not have much fiscal room to maneuver. The baseline analysis for 2013–17 shows that the government will continue to struggle to create fiscal space and reduce debt from its currently high level. The dynamics are mainly explained by falling grants as a share of GDP combined with rigidities in current expenditure (mostly the wage bill). Although the government is expected to limit current expenditures growth in real terms, unless additional revenue measures are introduced to offset the falling grants, it will struggle to reduce its deficit. While the deficit is expected to be largely financed by external concessional loans, debt will remain high as a share of GDP, and will likely keep the country at “high risk of debt distress”. 6. External shocks relating to the reduction in aid, rainfall, and the volatility of coffee prices could significantly increase the country’s deficits and debts, further undermining fiscal sustainability. Stochastic simulations show that, under very adverse shocks to rainfall or coffee prices, the country’s debt could reach 39 to 41 percent of GDP by 2017—about 5 to 7 percentage points above the projected ratio of debt to GDP under the baseline. Aid shocks could have a bigger impact—with debt to GDP reaching as high as 50 percent under adverse shocks— but the estimates are the least statistically reliable because of the time series is short and because its historical volatility in part reflects endogenous shocks (e.g., reform implementation) as well as exogenous shocks (donors’ behavior). 7. To reduce vulnerability to future shocks the government should focus on building buffers, creating fiscal space. Both short-term and medium-terms options are explored. In the short term, the authorities should persist in implementing planned fiscal reforms, to widen the tax base and eliminate unaffordable exemptions and subsidies. The policy scenario analysis shows that future fiscal sustainability will substantially hinge on the government’s ability to stick to its plans to increase revenue. Also, smaller fiscal deficits would allow the government more flexibility to use discretionary policies to respond when hit by a shock. Lowering the risks of debt distress by implementing the reform actions proposed under the 2012 DeMPA is also expected to give Burundi greater access to both concessional and non-concessional emergency financing. In the medium term, sustained efforts to improve the productivity in the agriculture sector are critical to reduce dependency on food imports, thereby reducing the financing gap. In the coffee sector, a national stabilization fund might be considered to better protect domestic producers against the volatility in international coffee prices therefore setting the right incentives for higher productivity. Finally, the government should also prioritize the implementation of the 2 new mining code, establishing an adequate fiscal regime to govern royalties and mining income taxes, with the ultimate goal of maximize the growth-enhancing potentials of expected mining revenues and avoid the risk of the ‘resource curse’. 8. Once fiscal space is created, an additional challenge is to ensure its efficient use. Substantial improvements in the quality of public expenditure management are urgently needed to meet this goal. Over-executions are prevalent in the recurrent budget, indicating significant violations of the expenditure chain and non-compliance with the budget ceiling set by Parliament. While efforts to stabilize current expenditures need to continue, additional energies are needed to bring under control extra-budgetary expenses paid outside the control of the expenditure chain. An additional challenge is the extra-budgetary nature of donor-funded projects, whose disbursement are reflected in commercial bank accounts and not captured in the treasury account. This undermines the quality of budget planning and also affects the ability to monitor the execution of development projects, contributing to sub-optimal sector allocations (with trends deviating from PRSP-II targets), and eventually undermining aid efficiency in Burundi. An integrated approach to the budgetization and report of external financing is needed to improve budget executions and promote allocative efficiency. Finally, urgent actions are needed to improve the quality of the MTEF and make it a planning tool conducive to macroeconomic and budgetary stability, with observable impacts on allocative efficiency (inclusive of donor-financed spending). 9. To manage the trade-offs between fiscal stability and capacity constraints, the Government may need to be more flexible in its approach to civil service pay. The growing civil service wage bill has created a significant rigidity in government spending and exacerbated the economic impact of external shocks. Options for reforming the salary scale currently under discussion are aimed at restoring fairness and improving incentives for better public sector performance. However, some of these options might have substantial fiscal implications and create strong inflationary pressures – likely to take place given the pre-election budget cycle - , undermining recent efforts to bring the wage bill under control. An alternative approach might be moving beyond across-the-board salary increases and stimulate a targeted, incremental pay reform for key government departments and agencies. While keeping the fiscal aggregate under control, this approach is more likely to overcome obstacles in implementing the new pay policy reform by identifying a more pragmatic and politically feasible solution, exploiting the appetite for reform that might exist within certain areas of the government administration. This in turn is expected to promote government effectiveness and strengthen its capacity to deliver. Key policy issues under investigation Macro-economic developments and outlook 10. Economic growth has remained modest compared to EAC neighbors and other post-conflict countries. Burundi’s economy expanded gradually in the post-conflict period, but the growth rate remains considerably lower EAC average. So far, the country has yet to experience the same economic boom as other post-conflict countries such as Uganda in the early 1990s or Sierra Leone and Rwanda in the early 2000. In the last three years, GDP growth averaged 3.9 percent per year instead of the projected 7 percent. This is considered insufficient to significantly reduce poverty rates, given the high rate of population growth. 3 11. Inflation remains volatile and highly sensitive to imported food and fuel prices. Burundi has one of the highest pass-through in the East Africa Community for both food and fuel2. In 2008 and again 2011, inflation averaged 24 percent per year following food and oil price developments at the regional and international level. Also, the government’s continued practice to borrow from the Central Bank to finance its budget added to the inflationary pressure, bringing inflation up to 25 percent in March 2012. 12. The current account deficit has deteriorated sharply in the last 3 years and is expected to remain large, mainly as a result of diminishing international aid, a narrow export base and deteriorating terms of trade. Burundi has suffered a constant decline in its terms of trade driven by increases in imported food and fuel products and, more recently, a decline in the international coffee prices3. The combined effect of these trends has been a sharp deterioration of the trade balance between 2009 and 2010, passing from 14.3 to 30 percent of GDP. While the trade imbalance has been usually financed by official current transfer, these transfers (mainly donor programs/grants and budget support) dropped from an average of 24.7 percent of expenditure in 2008 to 18.2 percent in 2012. As a result of these negative developments, Burundi is expected to have the 11th worst largest current account deficit in Sub- Saharan Africa by the end of 2012. Beyond 2012, terms of trade are expected to worsen gradually as a result of declining coffee prices. Although the government is expected to limit wage growth in real terms, unless additional revenue measures are introduced to offset the falling grants, it will struggle to reduce the deficits in the medium term, and will have to finance it mainly with concessional borrowing from multilateral and bilateral creditors. 13. Though the Government of Burundi has managed to conduct a prudent fiscal policy, future fiscal sustainability will critically hinge on the government’s ability to stick to the revenue mobilization efforts and overcome setbacks following unexpected changes in the tax base. Since the creation of OBR in 2009, Burundi has been able to improve revenue collection efforts. While tax revenue reports on average a higher growth rate than other EACs, it remains lower as percentage of GDP. In terms of performance, the progress has been uneven, with taxes on goods and services providing more than 50 percent of the total tax revenues, followed by taxes on income, profits, and capital gains (30 percent), and taxes on international trade (10 percent). Moreover, recent and unforeseen changes to the tax base associated with the new income tax law, combined with ongoing discretionary practices for tax exemptions and generous provisions under the Investment Code have undermined previous results, suggesting that further efforts are needed to widen the tax base. As part of the Framework to Strengthen Public Finances, the Government has engaged in efforts to rationalize and reduce tax exemptions. Recently, a working group composed of the Ministry of Finance and the OBR has been set up in order to propose a new ministerial ordinance creating a clear framework for fiscal privileges granted through the Investment Code. A stochastic simulation shows that foregone revenue rises as GDP grows over the projection period, costing 62 billion a year in 2013 to 100 2 Food and fuel prices in Burundi tend to pass-through to domestic inflation with rates of 77 and 48 percent, respectively with lags of between three to six months (IMF 2012). 3 Coffee exports represent around 60 of the total exports (3 percent of GDP). Global prices of Arabica coffee—the prices that matter for Burundi—were firmer during the 2000s, but dropped 30 percent year-on-year in 2012 because of strong supply from top global producers. Producer price of coffee set by the Burundian government is to some extent unrelated to the world market price of coffee, and coffee farmers in Burundi have traditionally received a lesser price for its coffee than the international reference price, Over time, poor incentives have resulted in low investments and declining quality and quantity of production. 4 billion a year by 2017 (about 1.5 percent of GDP per year). Fiscal balance will widen from 1.7 percent of GDP in 2012 to 5.1 percent of GDP by 2017. With higher gross financing needs each year, debt to GDP will reach nearly 40 percent by 2017, over 6 percentage points above the baseline. 14. Tax exemptions continue to grow in absolute terms, calling for urgent efforts to streamline them. According to OBR official data, tax exemptions on VAT on imports, custom duties, and redevances administratives have passed from FBu 64.7 billion in 2009 to FBu 106.3 billion (about US$ 70 million) in 2012. This is equivalent to about 3 % of GDP or about one- fifth of the total revenue collected by the OBR. Preliminary data for the first three quarter of 2013 seem to confirm this trend, suggesting that current tax incentive regimes are overly- generous and are likely to contribute to a significant amount of lost revenues for the authorities. Beneficiaries of tax exemptions are NGOs, private sector investors, members of Governments, diplomatic missions, and Government service providers. OBR data also indicate that service providers and contractors are among the largest beneficiaries of the tax exemptions, and the category where most of the exemptions granted without Parliament’s approval are reported. Overall, these findings indicate the urgent need for the Government to rationalize the system of tax incentives. 15. Results from the fiscal sustainability analysis suggest that the government’s accelerated effort in this area could produce a significant impact on fiscal outcomes. On the assumption that the authorities will gradually reduce tax exemptions revenue excluding grants will rise from 14.8 percent to 17.6 percent of GDP. Fiscal balance will begin to improve after 2014, approaching near balance by 2017. The government could build up a fiscal buffer while lowering the level of public debt, potentially creating additional borrowing space for priority investments. 16. The country’s external debt situation has improved since Burundi reached the HIPC completion point in 2009, but the outlook remains under high risk of debt distress. Meanwhile, 2009 HIPC accruals did not benefit investment. According to the IMF’s DSA, the debt outlook has deteriorated since 2011 due to “(i) the revision of the macroeconomic outlook following recent adverse shocks, mainly due to the terms of trade; (ii) the addition of new loans; and (iii) the lower discount rate.” Key risks identified include a narrow export base, lower GDP growth, and unsustainable borrowing requirement (i.e. high fiscal deficits). This was particularly the case in 2012. Drivers of fragility and their fiscal impact 17. Levels of Aid-Dependency remain among the highest in the world, with marginal role for remittances or FDI: Burundi has the second highest aid- dependency amongst fragile states in terms of ODA to GDP ratio (on average 2000-2011). It has a less pronounced position as a per capita recipient of ODA amongst the fragile states group, and receives just 1% of the total, but high aid reliance is nonetheless a defining current characteristic of the national economy. The influence of remittances is more marginal in Burundi than in other developing countries and shows little increase potential. While the country has not experienced radical ‘aid 5 shocks’4 , donor support remains uncertain despite pledges of US$2 billion in assistance at a Geneva Conference (October 2012), as most donors reiterate their determination to tie continued support to tangible improvements in critical reform areas including good governance, public financial managements, and private sector development. 18. Capital investments remain largely volatile and dependent on project grants from the donors, while overall budget execution is highly sensitive to delays in disbursement of budget supports. Overall, total aid inflows (i.e., budget support, project support, and technical assistance) represent more than half of Burundi’s budget. While the government succeeded in reducing public spending following marked declines in foreign grants, the bulk of spending contraction fell on public investments, which in 2012 fell by 20 percent from the previous year’s level5. (Moreover, delays in disbursements of budget supports put additional pressures on budget execution, widening the deficit and creating serious liquidity constraints. To address the liquidity crisis, the Government has limited room of maneuver, asking cash advanced to the Central Bank (creating pressures to increase the monetary base) and/or accumulating pending payments. Eventually, arrears build up at the end of each year as a result of the unavoidable pressures placed on budget execution by late aid disbursements and there is a year on year accumulation. In 2008, the stock of arrears was FBu 351 million (0.02 % of GDP). At the end of 2011, the stock of arrears increased up to FBu 13,093 billion (0.4 % of GDP). This has substantially compromised previous efforts to reduce the stock with targeted budget support. 19. Lack of export diversification, including under-exploitation of mineral resources, undermines current efforts to widen the revenue base: Burundi’s goods export basket has been relatively stable in the 2000s. Export earnings remain highly dependent on primary commodities, predominantly coffee and, to a reduced extent, tea, which together represented almost 70 percent of total exports in 2008/10. Other traditional exports (hides and skins, cotton and sugar) add some degree of diversity - the share of these products increased from 2.5 to 5.6 percent of total exports between 2001/03 and 2008/10. However, the almost total absence of new higher-value exports in the last decade is striking. Moreover, Burundi possesses 5-6 percent of known world reserves of nickel as well as other minerals such as gold, cobalt, tungsten, and ore. While mining activities are undertaken in a small scale and gold exports amount at 14.1 percent of total exports over the 2008-2010 period, the revenue-generating potential of the extractive industries is under-exploited due to infrastructure constraints (energy and transport) and delays in reforming the governance of the sector. While the approval of the new mining code might catalyze private investments in the sector, further efforts are needed to establish an adequate fiscal regime– to ensure the GoB gets the best possible deal for the exploitation of its resources – and promote innovative approaches focused on ‘local content’ provisions 6 - to maximize the sector contribution to job-creating growth. 4 “Aid shocks” are defined as yearly change in the level of ODA greater than 10 percent (OECD 2012). Burundi has never experiences such high level of aid volatility since 2002. 5 Both current and capital expenditures show ‘relative’ decline, that is the increase in absolute terms but grow less faster than the GDP. 6 Except for the labor-intensive construction phase, extractive industries generally do not generate a high share of employment in the host country. However, domestic firms can find business in a range of supporting sectors, such as transport and equipment manufacturing or maintenance. Innovative approaches focused on ‘local content’ provisions are becoming increasingly popular across developing countries to maximize jobs for nationals, by binding or incentivizing companies to employ and train nationals through contractual obligations and incentives. 6 20. Low prices and declining production in the coffee sector further exacerbates the effects of export concentration, with significant impact on the country’s fiscal path. Despite its good quality, Burundian coffee farmers have historically received a lesser price for their coffee than the international reference price. This situation profoundly discourages farmers. Moreover, in addition to volatile coffee prices, the country’s export earnings from coffee depend on the quantity and quality of the coffee harvest each year, which have been declining. Export earnings from coffee also depend on the quantity and quality of the coffee harvest each year. The quantity and quality of output have suffered in recent years from tree disease, low input use, and low incentives from low producers prices. 7 The FSA shows that, other things being equal, a negative coffee price shock will result in lower coffee export values, which will then result in lower government revenue and higher fiscal deficits. The higher fiscal deficits cause a rise in gross financing needs, leading to higher debt. At an extreme of the simulation, under very adverse shocks in coffee prices, the government’s debt could spike 41 percent of GDP by 2017. 21. Wide fluctuations in domestic food production – linked to adverse impact of climate change – leave the country vulnerable to volatility in international food crops prices: Though subsistence agriculture engages the vast bulk of the population, domestic production is insufficient to meet the country’s food needs, and the country relies on import of food to cover the deficit, with adverse impact on the budget. Insufficient rainfall exacerbates this problem by leading to falling agricultural production, although stochastic simulation results suggest that rainfall shocks carry a smaller risk to fiscal outcomes than shocks to world coffee prices8. As a combined effect of these trends, the prices of essential food crops have been increasing in Burundi since the end of 2007. Starting in May 2012, the Government has decided to subsidize essential food products, widening the fiscal deficits. Public Expenditure Management 22. The inconsistencies in budget data sources are significant, and substantially limit the reliability of the results and implications emerging from the analysis. The data from the Treasury balances show budgetary over-executions in all the economic categories of the budget with the exception of capital expenditures. SIGEFI, to the contrary, shows under-executions for most ministries, including the priority sectors. These problems have already been highlighted in the 2012 PEFA report and reflect major accounting and budgeting problems requiring immediate attention from the authorities. 23. Current expenditures are systematically over-executed, reflecting poor budget execution process. Despite recent efforts to devote additional resources to productive investments, current expenditures represent 78 percent of the Government budget (excluding donors’ project support) over the 2009-2011 period, with salaries representing the largest single item of current expenditures. This suggests that the government is unable to undertake massive public investments in the future. Current expenditures are over-executed, suggesting (i) poor quality of the budget as a planning document and (ii) recurrent violations to the expenditure The state-run coffee agency, OCIBU, annually sets the producer price and is in control of marketing and export. According to some observers, state-led ownership has been a major negative factor on private initiative and incentive to invest (IMF 2008, CEM 2011). 8 The smaller revenue impact of rainfall shocks might be explained by the well-known difficulty of collecting taxes from the mainly rural, subsistent agricultural sector. Alternative explanations are provided in Chapter 3. 7 chain (credits are ceilings under the Burundian budget system).9 The average budget execution rate for the period 2009-2011 was 109.4 percent for recurrent expenditures and 73.1 percent for government own-funded capital expenditures. In the case of interest payments, however, over- executions are partly driven by exchange rate losses derived from the volatility of the Burundi Franc exchange rate. 24. Avoiding expenditure controls through special procedures are an important issue at the budgetary and treasury level. Ordonnance ministérielle 540/0015 of January 9th 2008 indicates the budget items that can be paid through special procedure (procedure exceptionnelle). An analysis of the data registered in SIGEFI shows that 13 percent of the budget 2012 could be paid without going through the regular controls, of which 12 percent corresponds to debt (principal and interest) payments. Besides fiduciary problems, these practices seriously undermine budget execution by exacerbating cash management problems. 25. The pro-poor budget has received progressively more resources, reflecting government effort to focus on priority sectors - Health, Education, Agriculture, Energy and Mining, Public Works), as well as other social services. This indicates that, at least at the allocation level, the pro-poor budget is consistent with the poverty reduction strategy (PRSP-II). 60 percent of the pro-poor budget was allocated in Education. The next largest concentration of pro-poor spending is the Ministry of Health and Fight Against AIDS10, which received, on average, 12.5 percent of the pro-poor budget. Despite this progress, other priority sectors (namely: agriculture) remains systematically under-funded, while the security sector continues to receive large share of budget resources, despite previous recommendations to shift resources away from the sector (World Bank Policy Notes 2013). 25. The fact that the performance of pro-poor spending is not translating more identifiably into realization of MDG goals suggest that further efforts are needed to improve the efficiency of public spending. According to the 2012 Joint Staff Advisory Note on PRSP-II, while the effective implementation of the program will help to converge toward the MDGs, very few of them will be met by the end of 2015. The document estimates that poverty, extreme poverty, and hunger rates are unlikely to be halved within the next 4 years in comparison to the 1990 benchmark. Moreover, none of the health and social protection MDG’s will be met despite notable progress observed in some important areas such as HIV/AIDS prevalence. 26. The quality of the MTEF in Burundi has deteriorated over time, raising concerns on the suitability of the tool’s use. There are significant inconsistencies within the MTEF document, as well as between the MTEF and the estimates shown in the budget framework letter. Moreover, these inconsistencies increase over time, raising concerns on the capacity of the Burundi authorities to adopt MTEF. This suggests that MTEF is not yet used directly as a budget planning tool, and Burundi has yet to reach a stage where MTEFs materially impact public expenditures. 27. The implications of non-application of the MTEF tool require a systematic investigation of the reasons accounting for the problem, followed by targeted and long- 9 This does not apply to 2009, when a revised budget increased the ceilings for most of the credits. 10 Before 2011 this ministry was divided in two ministries. We added the two ministries to make this comparison. 8 term technical assistance to overcome them. At minimum, a full independent audit of the Excel file used to produce the MTEF estimates is needed to correct internal inconsistencies. Differentiation between domestic and external financings should also be re-introduced. More broadly, it needs to be examined what specific form and modality of technical assistance package for effective operationalization would build the necessary levels of competence to the system in Burundi’s current institutional environment. Wage Bill Management and Options for Pay Reform 28. Although the size of the wage bill is within the normal range for the region, there is little fiscal space for expanding it. The wage bill accounts for nearly half of the recurrent budget, and has increased in recent years, before falling back to 8 percent of GDP in 2012. This volatility has been driven by changes in the size of the civilian workforce (mainly in the education sector). As part of the peace deal, many soldiers were added to the payroll. Concurrently, the Government has added teachers, nurses, and doctors to boost the prospect of achieving the Millennium Development Goals (MDGs). Wage expenditures, consequently, have outgrown GDP, and if this trend continues, drastic measures may have to be taken to reduce wage bill growth. In 2012, however, the government established a hiring freeze (in 2012) for all sectors except education, health, and justice. 29. A major source of job dissatisfaction in the public sector is inequitable pay, which undermines work motivation. Base salaries are determined solely by educational certificates, not by job responsibilities. Substantial inequities exist in the remunerations of different occupational groups. Teachers and some health workers receive significantly more pay than other civil servants with similar qualifications. For example, a newly recruited teacher is paid 50 percent more than another new civil servant recruit with equivalent education, though this cannot be justified by market pay levels. Inequities are further exacerbated by considerably higher salaries in the case of political appointees and throughout the autonomous agencies. 30. Although public salaries are high, on average, compared to per capita income, they are low in absolute terms. The average monthly salary of 150,000 FBu translates to approximately $96 U.S. dollars, while monthly consumption per capita in 2011 was estimated by the national statistical agency to be 83,248 FBu. Given that the average household contains five individuals (according to the latest census), the average wage level is relatively low (Government of Burundi, 2012). To cope with low-salaries, a variety of bonuses and allowances has been introduced and represent a significant component of the compensation package. 31. There are currently no coherent policies that guide pay increases, the size of the public workforce, or job promotions in Burundi. The current policy on annual pay increases has not been adjusted since 2006, which means that the real wages of new entrants is significantly less due to inflation. At the same time, there is no centralized authority that evaluates the optimal size of the public payroll, which has increased by nearly 20 percent since 2007. Furthermore, job promotions are solely dependent upon educational attainment and years of service, without any consideration to performance 9 32. The Government has recognized the need for reform and has adopted a National Program for Administrative Reform (PNRA),11 which has been endorsed by the Council of Ministers. The PNRA calls for rationalizing the civil service, promoting greater accountability, and strengthening dialogue between the Government and unions. A reform pay roadmap has also been recently developed and discussed with government officials, which contains a proposed timeline to implement a simplified and unified pay structure. 33. Experiences from other low-income countries, however, have shown that ambitious and across-the-board civil service reforms that seek to transform civil service management have often failed to match expectations. In addition, across the board pay harmonization exercises are typically met with strong resistance from unions and politicians, and do not necessarily improve performance. 34. In the short-term, a more pragmatic approach to improve performance may be to stimulate a targeted, incremental reform of key functions within government departments and agencies. An asymmetric approach, though departing from conventional approaches of public administration reforms by establishing a more complex and less homogeneous civil service, follows recent global trends and experiences in other developing countries. Hasnain and Manning (forthcoming) explain that this approach should prioritize “transformations and changes that would enhance the legitimacy of the center rather than undermine it, and which both exploit and nurture the appetite for reform that does exist within certain areas of the government administration.” Examples of this approach include diversified pay arrangements at the agency level, annual performance targets, and moves toward position-based (as opposed to seniority- based) systems which encourage lateral entry. 35. Given the difficulties in attaining comprehensive civil service reform, full harmonization and simplification of the pay structure should be pursued as a medium- term objective, with the full support of the unions, to find a fiscally sustainable solution to low pay levels. It is also recommended that any new structure attaches job responsibilities, such as decision-making and the difficulty of tasks, to each grade. This reform would boost performance by providing higher pay to those taking on more difficult tasks. 11 Government of Burundi (2012c). 10 Policy Recommendations 36. In support of the general directions called for in the report’s particular policy focus on (i) creating adequate fiscal space (ii) using this space to improve the efficiency in the management of public expenditures, the following recommendations have been identified: OBJECTIVE POLICY AREA PROPOSAL FOR ‘PRIORITY ACTIONS’ TIMELINE CREATE FISCAL Revenue Reexamine the new personal income tax (2012) redressing Short-term SPACE mobilization the effect of increased tax allowances. Include in the annual budget a reliable estimate of the tax Short-term exemptions expected to be incurred, to increase transparency on the quantum and types of tax incentives given. Revise existing tax incentives/exemption regime bases on a Medium- systematic assessment of the costs and benefits of existing term incentives. Aid Examine with key donors, the options to increase the Short-term predictability predictability of aid flows, including the possibility of creating contingency funds to be used in the event of exogenous shocks or other emergencies. Export Establish an adequate fiscal regime to ensure the GoB gets Medium- promotion/ the best possible deal for the exploitation of its resources, term Governance of including the adoption of innovative approaches focused on the Extractive ‘local content’ provisions. Industries PROMOTE Public Improve the quality and reliability of budget data, by (i) Short-term(i) GOVERNMENT Expenditure introducing a functional classification of the expenditures Medium- EFFECTIVENESS Management / and (ii) Integrate SIGEFI in the expenditure chain, in such term(ii) FM Information a way that the data extracted from the treasury balances System match those provided by SIGEFI. Reduce the level of extra-budgetary expenditures and Short-term comply with legislative provisions to minimize discretionary practices outside expenditure chain control Improve information on external financing by aligning Medium- CNCA reporting mechanism to the official nomenclatures term used for domestic budget MTEF Revise MTEF to distinguish between estimates of donor- Short-term funded expenditure and expenditures financed with own resources. Medium term Develop training module on MTEF, including macro- modelling and GFS 2001, and explore options for long- term TA. HRM / Pay Examine the case for targeted, incremental pay reform. Short-term Policy Reform Review public sector organizational structures to identify Medium- needs and improve HRM system. term 11 STRUCTURE AND APPROACH Chapter 1: Macroeconomic Developments and Prospects What are the key recent developments in the institutional and economic context of the country? What are the recent trends in economic growth and what remain the most critical vulnerabilities? What is the current debt outlook, and has previous debt relief been channeled to investment? What progress has there been on domestic revenue mobilization? What challenges remain? Chapter 2: Drivers of Fragility In the push to transform the economy and establish sustainable growth what are the key current impeding factors? How is Burundi’s particular aid dependency profile characterized? What are the implications for budget management of the volatility of budget support? What is the configuration of imports and exports and what challenges are posed by such configuration? How does domestic food production and climate change affect the economic environment? Chapter 3: Fiscal Sustainability Analysis What is the most likely scenario for the country’s medium-term fiscal outcomes? What are the magnitude of fiscal risk arising from the country’s exposure to well-known external shocks relating to aid, rainfall, and coffee prices? What are the potential medium-term implications of the specific revenue and expenditure policy reforms that are currently being considered by the authorities? Are they likely to help Burundi on a sustainable fiscal path and build resilience? Chapter 4: Public expenditure Analysis What is the quality of the budget data and how this constrains a thorough analysis? What are the stand-out constituent elements of public spending and what is the pattern of execution rates? What are the defining features of foreign financing? To what extend the management of aid is integrated into the domestic budget process? How is pro-poor spending characterized including in distribution and execution? What has been the experience of the introduction of MTEF and what lessons can be learned? Chapter 5: Confronting Competing Pressures for Reform: a Case-Study of the Wage Bill What is the nature and the scope of the problem? What have been the developments on wages and employment in Burundi? What inequities are present? How governance and institutional dynamics matter? What are the most feasible options for reform? Is an asymmetric approach suitable? 12 Defining the process of engagement 1. A broad consensus has emerged among practitioners and researchers that how problems and reform approaches are identified matters at least as much as what the solution is. This suggests that development practitioners should focus on bringing a broad range of stakeholders together and facilitate an extended process of engagement to collectively identify the problems and identify the solutions together with the client. 12 This implies that there should be no presumption to evaluate public spending in isolation. Rather, the emphasis should be on building government involvement, capacity and ownership so that policymakers undertake such an exercise themselves on an ongoing basis as an integral part of their public expenditure planning, budgeting, and evaluation (Pradham and Campos 1997). 2. The present study follows this approach and articulates the “process of engagement” through an extensive dialogue with the government and the development partners involved in this study. In particular, the Task team carried out extensive consultations at the early stages of the study (including two field work missions in September and November 2012, respectively), targeting key government counterparts to test their interest for this work. These consultations included technical discussions on the design of the PER and key thematic areas of focus, helping the team to identify aspects where more traction can be expected and where the study’s policy relevance can be maximized. During PER preparation, the team carried out four additional missions, including two rounds of data collection efforts (December 2012 and February 2013, respectively). Preliminary drafts of the analysis were shared during two subsequent missions (April 2013), to give the authorities the opportunity to react and provide feedback to the team. The validation of the data by the technical cadres of the Government was essential to reach a consensus on the interpretation of the findings of the analysis, as well as on their policy implications. 3. Client ownership and commitment. During the consultation process, the Permanent Secretary of the Ministry of Finance and Economic Development Planning (MFEDP) reiterated the strong interest of the Government for the Burundi PER. She also expressed her satisfaction and agreement to lead a joint Government-Donor “Technical Working Group” (PER-TWG) to support the study and improve the quality of the analysis. In the intentions of both parties, the PER-TWG was intended to act as a forum for discussion and analysis of preliminary findings, maintain a close dialogue, and jointly discuss possible ways forward. Thanks to regular interactions and exchanges with the World Bank team, the PER-TWG has successfully fulfilled this role, being instrumental to help the team identifying sources of problems and outline a set of realistic and feasible solutions. Government officials and members of the donor community have all expressed satisfactions with the functioning of the PER-TWG. In particular, stakeholders praised the extensive analytical and evidence-based policy discussions held throughout the PER preparation. It is hoped that this approach will help to promote government ownership of the findings from the analysis; and maximize the prospects of implementation of the policy recommendations presented here. 12 Andrews (et al 2012) describes a “Problem-Driven Iterative Adaptation” (PDIA) approach as a means of putting this idea into practice. 13 Chapter 1 : Macroeconomic and Fiscal Developments By EAC comparisons, Burundi’s growth has been modest (3.9 percent on average since 2007) and, crucially, just slightly ahead of population growth. Vulnerability to shocks is pronounced, arising from the country’s reliance on imports of food and fuel and reliance on coffee export. Heavy reliance on foreign aid also poses a risk as this source of revenue is highly unpredictable. Despite significant progress in strengthening tax administration, the effort to widen the tax base and enhance domestic revenue has been blunted by elite influence. Concerted effort on revenue mobilization, agricultural modernization, export diversification, PFM, and other institutional improvement is needed to achieve sustained growth and poverty reduction. 1.1 Recent Economic Developments and Trends 1.1.1 GDP growth 1. Economic growth has Figure 1.1: Annual real GDP growth – Burundi in remained lower than expected, and comparative perspective modest compared to EAC and 12 other post-conflict countries. Average Burundi’s economy expanded 10 Africa gradually in the post-conflict period, but the growth rate remained 8 EAC, considerably lower than neighboring excluding countries in the EAC community or 6 Burundi the SSA region. The results hardly 4 Burundi change if Burundi is compared with other post-conflict countries (Figure 2 1.1). The country still grew more Post- slowly - on average at 3.3 percent Conflict 0 each year - than the other low-income 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 post-conflict countries in the region -2 (4.2 percent on average). In the last three years, GDP growth averaged 3.9 Source: WDI The sample of SSA Low-Income Post-Conflict states includes Democratic percent per year instead of the Republic of Congo, Liberia, Rwanda, Sierra Leone, and Sudan. Countries projected seven percent. This is are compared annually after the end of the conflict in each country. considered insufficient to significantly reduce poverty rates, given the high population growth (three percent per year, on average). 14 2. The modest economic performance reflects the lack of economic diversification and fluctuating prices in the agricultural sector. Burundi has a small domestic market so it needs to rely on external markets as the main source to stimulate economic growth. However, the country is landlocked and faces very high transport and transaction costs to global markets so it has not yet managed to develop and diversify its export base. As a result Burundi’s exports/ GDP ratio – at five percent of GDP – remains among the lowest in the world. Finally, growth has been undermined by the structural problems in the agriculture sector, which has been affected by low productivity and weak investments, showing little shift to modern production methods or higher value agricultural products. These structural weaknesses have been compounded by the negative effects of climate changes and international shocks (volatility in international coffee prices). The low productivity of the agricultural sector has had a direct and significant effect on growth given the role played by the sector in the Burundi’s economy (representing about one third of GDP and employing more than 90 percent of the work force). 3. The economy recovered from the global economic downturn in 2009, but remains highly vulnerable to external and domestic shocks. In 2009, growth slowed to 3.5 percent compared to five percent in 2008. In the following years, economic activity shows modest signs of recovery, with GDP growth rates in the range between 3.8 and 4.2 percent in 2010-2012 (Figure 1.1). This is explained by the multiple effects of the financial crises, affecting growth through (i) a reduction in projected foreign direct investments; (ii) a decline in private transfers; and (iii) a reduction in expected exports. Exports depend on few commodities, namely, coffee and tea. Despite the importance of agriculture for the national economy, domestic production is insufficient to meet the country’s food needs, and the country relies on import of food as well as large import of fuel. As a result, Burundi is highly vulnerable to external economic shocks, including (i) global food and fuel prices, which often lead the Government to last minute adjustment in the budget either in the form of increased subsidies or a reduction in tax estimates as a result of cuts on staple food); (ii) volatility in international coffee and tea prices; (iii) changing climatic (rainfall) conditions; and (iv) aid flows. Burundi’s macroeconomic performance is also contingent on domestic factors, such as a weak implementation capacity coupled with weak governance institutions, and enduring risks of political and social instability linked to the return of refugees and its impact on land conflicts and other security threats.13 13 Access to land is limited and remains an important driver of conflict and fragility. Total surface of cultivable land shrinks under growing demographic pressure as the population has quadrupled in half a century. Also, more than 500,000 refugees have returned to Burundi over the past 10 years and are now claiming ownership of occupied land. It is estimated that between 70-90 percent of court cases are linked to land issues, with disputes often turning violent. 15 Table 1.1 : Economic Trends, 2007-2014 2007 2008 2009 2010 2011 2012 2013 2014 Population (million) 7.7 7.9 8.2 8.4 8.6 8.8 9.0 9.2 Population growth (%) 3.1 3.0 2.9 2.6 2.3 2.4 2.4 2.4 GDP (current prices, US$ billion) 1.6 2.0 2.2 2.0 2.2 2.5 2.8 3.0 Real GDP growth (%) 4.8 5.0 3.5 3.8 4.2 4.0 4.5 5.1 Inflation (average CPI, %) 8.4 24.5 10.6 6.4 9.7 18.1 7.0 7.5 Real effective exchange rate (% change) -7.5 15.4 0.4 1.8 0.7 n/a n/a n/a Terms of trade (% change) -23.4 3.4 19.6 51.7 -8.9 -17.9 -10.1 +3.8 Gross investment (% of GDP) 18.2 18.7 18.8 18.0 20 20 20 20 Gross national savings (% of GDP) -0.1 9.3 9.0 9.6 6.3 4.4 3.7 4.0 Overall fiscal balance (excl. grants, % of -26.2 -27.6 -23.5 -26.4 -24.6 -19.8 -15.8 -14.8 GDP)a Overall fiscal balance (after grants, excl. -2.6 -2.7 -5.0 -3.6 -4.0 -1.7 -1.7 -3.3 HIPC, % of GDP) (1) Current account balance (incl. official -6.1 -1.8 -11.5 -12.3 -14.8 -15.1 -14.5 -14.6 transfers, % of GDP) Overall balance of payment (% of GDP) 2.4 5.5 3.0 0.7 -1.6 -1.1 1.4 -0.2 Gross international reserves (months of 3.5 4.3 4.9 4.1 3.4 3.4 3.7 3.9 imports) Export growth (fob; US$ % change) 1.1 18.4 -1.8 48.0 22.5 4.7 -17.9 +9.2 Share of coffee in exports (%) 65.5 57.8 57.9 68.9 60.9 59.9 48.8 51.1 Debt service to exports ratio (%) 5.4 7.7 1.7 1.4 2.8 2.9 5.2 External debt stock (% of GDP) 115.1 97.5 20.7 22. 24 21 19 18 Source: Burundi authorities; World Bank; IMF. (1) Cash basis. 1.1.2 Balance of payments 4. The current account deficit has deteriorated sharply in the last three years, largely as a result of diminishing transfers, a narrow export base and deteriorating terms of trade. Since 2009, Burundi experienced a steady decline in current transfers (mainly donors’ programs and budget support), and deteriorating terms of trade. In 2008, current transfers reached an all- time high of 27.4 percent of GDP. Since then, this component of the current account has decreased gradually and reached 18.2 percent of GDP in 2012. Burundi has also suffered a constant decline in its terms of trade driven by increases in imported food and fuel products and, more recently, a decline in international coffee prices.14 The combined effect of these trends has been a sharp deterioration of the trade balance between 2009 and 2010, passing from 14.3 percent to 30 percent of GDP (Figure 1.2). Though Burundi’s current account has traditionally been lower than other countries in the region, the difference has grown larger since 2010. 5. The current account deficit is expected to remain large in 2012. In 2012, Burundi’s trade deficit was estimated at -29 percent of GDP. Freight and insurance costs on imports and the 14 Coffee exports represent around 60 of the total exports (three percent of GDP). Global prices of Arabica coffee— the prices that matter for Burundi—were firmer during the 2000s, but dropped 30 percent year-on-year in 2012 because of strong supply from top global producers. Producer price of coffee set by the Burundian government is to some extent unrelated to the world market price of coffee, and coffee farmers in Burundi have traditionally received a lower price for their coffee than the international reference price, Over time, poor incentives have resulted in low investments and declining quality and quantity of production (section 3.3 in chapter three further elaborates on this external shock in the context of the FSA). 16 salaries of foreign construction workers represented minus four percent of GDP, while current transfers reached 18 percent of GDP. As a result of these negative developments, Burundi is expected to have the eleventh worst largest current account deficit in Sub-Saharan Africa by the end of 2012. ToTs are expected to worsen gradually beyond 2012 as a result of declining coffee prices. Figure 1.2: Burundi’s balance of payments, 2000-2012 Current transfers have diminished sharply. Net exports stabilized at very low levels… 30.0 Current Transfers (as percent of GDP) Trade balance (as a percent of GDP) 0.0 25.0 -5.0 20.0 -10.0 -15.0 15.0 -20.0 10.0 -25.0 5.0 -30.0 0.0 -35.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 …driven by a continued deterioration in Current account balance is at historically terms of trade sharper than SSA’s average. low levels… Burundi: Terms of Trade (percent change with respect Burundi: Current Account Balance and Components (as a to the previous year) 40.0 percent of GDP) 60 30.0 50 20.0 40 10.0 30 0.0 20 -10.0 10 -20.0 Current Transfers 0 -30.0 Income (net) -10 Services (net) -40.0 Trade balance -20 -50.0 Current account balance -30 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2005 2006 2007 2008 2009 2010 2011 2012 2013 …and among the lowest in Sub-Saharan The world price of coffee continues its Africa. downward trend. Burundi, EAC, and Sub-Saharan Africa: Current Account Balance (as a Coffee, arabica (cents/kg) 5 percent of GDP) 700.00 0 600.00 500.00 -5 400.00 -10 300.00 -15 Sub-Saharan Africa (IMF 200.00 definition) EAC, excluding Burundi 100.00 -20 Burundi 0.00 1990M01 1990M11 1991M09 1992M07 1993M05 1994M03 1995M01 1995M11 1996M09 1997M07 1998M05 1999M03 2000M01 2000M11 2001M09 2002M07 2003M05 2004M03 2005M01 2005M11 2006M09 2007M07 2008M05 2009M03 2010M01 2010M11 2011M09 2012M07 -25 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Sources: IMF WEO Database (Vintage April 2013), IMF Africa REO Database (Vintage October 2012), World Bank Pink Sheet, and authors’ own calculations. 17 6. Despite an adverse context, the Burundi authorities have been able to accumulate reserves, largely as result of increased donor funding. The Government’s total reserves, excluding gold, passed from US$65.7 to 269 million (nine to 12 percent of GDP) between 2004 and 2012. The combined effect of HIPC and MDRI relief and increased donor funding acted as trigger to buildup foreign reserves. Despite this increase, Burundi’s coverage remains low in comparative amounts: at the end of 2011, its reserves-to-months-of-imports ratio was barely above three, which is considered a minimum “rule of thumb” level of insurance. 1.1.3 Inflation and monetary policy 7. Inflation remains volatile and highly sensitive to imported food and fuel prices. The fact that Burundi is a landlocked country increases its vulnerability. Econometric evidence produced by the IMF (2009) indicates that Burundi has one of the highest pass-through rates in the East Africa Community for both food and fuel. 15 In 2008, and again in 2011, inflation averaged 24 percent per year following food and oil price developments at the regional and international level. Also, the Government’s continued borrowing from the Central Bank to finance its budget added to the inflationary pressure, bringing inflation up to 25 percent in March 2012 (Figure 1.3). 8. While tightening the monetary base in 2012, the Central Bank has little room to maneuver and continues to face strong pressures to adopt expansionary policies. In 2012, the BRB adopted a tighter monetary policy and the Government decided to remove taxes on food. This helped to ease inflationary pressures, bringing down inflation to 12 percent (Figure 1.5). The Central Bank also responded to shocks by allowing greater exchange rates flexibility, with exchange rate declining from an average 28 percent during 2009-2011 to 12.7 percent in 2012. The continuing liquidity crisis of Burundi’s government (linked to the late arrival of budget support),16 however, implies that monetary authorities remain under pressures to address Government’s requests of cash advances by increasing the monetary base, which in turn creates inflationary pressures. 15 Food and fuel prices in Burundi tend to pass-through to domestic inflation with rates of 77 and 48 percent, respectively, with lags of between three to six months (IMF 2012). 16 Chapter two analyzes the fiscal and monetary effects of later disbursement of budget support in greater detail. 18 Figure 1.3: Burundi’s inflation and monetary policy, 2004-2012 Burundi’s inflation is above the EAC average. 35 Inflation (Y/Y percent changes) Average EAC 30 Burundi 25 20 15 10 5 0 -5 2004 Sep 2005 Sep 2006 Sep 2007 Sep 2008 Sep 2009 Sep 2010 Sep 2011 Sep 2012 Sep 2004 May 2005 May 2006 May 2007 May 2008 May 2009 May 2010 May 2011 May 2012 May 2004 Jan 2005 Jan 2006 Jan 2007 Jan 2008 Jan 2009 Jan 2010 Jan 2011 Jan 2012 Jan 2013 Jan Continued periods of inflation are linked with devaluations of the Burundi Franc… Inflation and Changes in the BIF/USD Exchange Rate 35 (Y/Y percent changes) Inflation 30 Y/Y changes in the exchange rate with the USD 25 20 15 10 5 0 -5 -10 -15 2005 Jan 2005 Sep 2006 Jan 2006 Sep 2007 Jan 2007 Sep 2008 Jan 2008 Sep 2009 Jan 2009 Sep 2010 Jan 2010 Sep 2011 Jan 2011 Sep 2012 Jan 2012 Sep 2013 Jan 2005 May 2006 May 2007 May 2008 May 2009 May 2010 May 2011 May 2012 May …leading to a continuous deterioration of the Real Effective Exchange Rate. Burundi: Real Effective Exchange Rate (Index, 2005=100) 120 100 80 60 40 20 0 2004 May 2005 May 2006 May 2007 May 2008 May 2009 May 2010 May 2011 May 2012 May 2004 Jan 2004 Sep 2005 Jan 2005 Sep 2006 Jan 2006 Sep 2007 Jan 2007 Sep 2008 Jan 2008 Sep 2009 Jan 2009 Sep 2010 Jan 2010 Sep 2011 Jan 2011 Sep 2012 Jan 2012 Sep 2013 Jan Sources: International Financial Statistics, and authors’ own calculations. 1.1.4 Fiscal policy 9. Though the Government of Burundi has managed to conduct a prudent fiscal policy, fiscal policy is still vulnerable to low revenue mobilization, high expenditure needs, and dependence on aid. Although Burundi has been able to increase revenue mobilization since the 19 creation of the Office Burundais des Recettes (OBR), the Government is still largely dependent on budget support and project grants from donors. The budget support still represents around 10 percent of the budget (three percent of GDP) and is one of the main sources of fiscal revenue and external reserves. Project grants, including technical assistance, represent 45 percent of the budget (15 percent of GDP). Overall, total external aid (i.e., budget support, project support, and technical assistance) represents more than half of the Government budget. Both budget support and project grants declined over the last three years, creating additional pressures on the country’s fiscal position. Revenues 10. Burundi revenue collection has Figure 1.4: Tax revenues breakdown improved significantly since the creation of (average 2007-2011) the Office Burundais des Recettes (OBR) in 60% 2009. The positive performance of OBR 50% 40% contributed to considerable improvement in 30% Burundi’s national tax administration, 20% efficiency of tax collection, and amount of % total tax revenues 10% taxes collected. (Box 1 provides an in-depth 0% international… Taxes on income, Taxes on property Taxes on goods Other taxes review of OBR performance). Also, the profits and… and services Taxes on Government facilitated the process by introducing VAT, a revision of the VAT law, a revision of the Income Tax law, and the introduction of Common Tax Procedures, which will take effect beginning in 2013. Burundi Average EAC Domestic revenue increased as a percentage of Source: Authors’ own calculations with data from IMF GDP, from 13.8 percent in 2009 to 15.3 percent GFS. in 2011, and declined again to 14.8 percent in 2012. The decline in 2012 reflected the temporary exemption of taxes on food products in May to December 2012, to limit the impact of food price inflation on the poor and losses in fuel excise taxes due to an emergency waiver issued by the Government to curb inflationary pressures.17 Taxes on goods and services provide more than 50 percent of the total tax revenues (Figure 1.4), followed by taxes on income, profits, and capital gains (30 percent), and taxes on international trade (10 percent). This is generally in line with the structure of the tax revenues in other EAC countries. 11. While tax revenues growth rates are higher than other EAC countries, they remain lower as percentage of GDP, suggesting that further efforts are needed to widen the tax base. In comparative terms, Burundi’s tax revenue reports on average a higher growth rate than the other EAC’s countries (with the exception of 2011). 18 It remains lower, however, as percentage of GDP. In terms of performance, the highest increase was registered in the tax collection of goods and services, with an average of 28 percent (Table 1.2). In the case of the 17 The authorities took other measures to partially offset the revenue losses. These included an increase in taxes on income, excise taxes on alcohol, tobacco, telecommunications, and duties on used cars 18 It must be noted, however, that Burundi was starting from a lower baseline. Also in terms of GDP, Burundi’s tax revenue is increasing, but remains below the average of EAC by about three percentage points. 20 EAC’s the strongest growth rate was observed in the tax collection of income, profits and capital gains (27 percent). Figure 1.5: Tax revenues growth rate Figure 1.6: Tax revenues as percentage of GDP 35% 18% (Annual growth rates) 30% 17% 16% (as % of GDP) 25% 20% 15% 14% 15% 13% 10% 12% 5% 11% 0% 10% 2008 2009 2010 2011 2008 2009 2010 2011 Burundi Average EAC Burundi Average EAC Source: Authors’ own calculations with data from IM F-GFS. Table 1.2 Burundi: Revenue, Main Categories as a Percent of GDP, 2007-2012 2007 2008 2009 2010 2011 2012 Taxes on income, profits, and capital gains 3.6 3.4 4.0 4.4 4.4 4.9 Taxes on goods and services 6.3 6.4 6.9 7.8 8.4 7.4 Taxes on international trade & transactions 2.3 2.1 2.1 1.6 1.5 1.1 Taxes not elsewhere classified 0.2 0.1 0.1 0.0 0.0 0.0 Other revenue 1.0 1.4 1.2 0.9 1.1 1.4 Budget support 6.0 7.1 3.8 5.0 3.9 3.2 Project support 9.7 9.7 8.1 7.3 11.0 9.0 Other Grants and Transfers (mostly 7.4 8.3 7.4 10.4 5.9 6.0 technical assistance projects) Total 36.5 38.5 33.5 37.3 36.1 33.0 Source: IMF 2012 preliminary data. 12. Several tax policy reforms have been recently introduced. The most prominent ones include the introduction of VAT in 2009 and the recent revision of the VAT law that will be effective early in 2013. A complete overhaul of the Income Tax law and the introduction of a Common Tax Procedures law beginning in 2013 were supposed to represent other dimensions of the extensive reform that Burundi is undertaking. However, the Government’s effort to broaden the tax base has been blunted by political resistance. At the end of 2012, a reform to the personal income tax was drafted aiming to simplify the tax brackets, broaden the base through reductions in the level of deductions, and, for the first time in Burundi’s independent history, make politicians and high-level officials pay their taxes. Estimates suggested that the reforms on the income tax would be revenue neutral, though the exemption on transport and housing allowances would shift the tax burden to the highest income brackets. However, after discussions within Cabinet and Parliament, the resulting law expanded the tax exemptions rather than reducing them. As a result, revenue collection in 2013 is expected to be significantly lower the projected. Available data shows that, between January and May 2013, the OBR collected FBu 216.7 billion, 21 which compares with FBu 219.2 billion for the same period in 2012. This is reduction of 1.13 percent in nominal terms (OBR 2013). Box 1 Burundi OBR: A success story in the region In Burundi, tax administration remained functional throughout the period of civil war, but the conflict distorted the tax collection process. It is a characteristic of the whole post-independence experience that the public service remained largely intact and functioning, though subject to atrophy. This is commonly attributed to the Belgian colonial administrative legacy. Tax administration in particular continued to be functional and the tax to GDP ratio did not drop. Indeed, it has been consistently higher than in all EAC countries except Kenya (ranging from 16 to 18 percent during the period 1998 to 2008 [AFDB 2010]). Common explanation for the phenomenon is the symbiosis of interests between the political and the administrative elites (revenue was needed to maintain privileges and pay salaries). The conflict exacerbated the tendencies of these interests to rent-seek and engage in corruption, thus further distorting the tax collection process. The situation confronted after the 2005 elections with regard to domestic resource mobilization was thus complex and corrective measures were a priority. Customs duties were being systematically evaded and multiple exemptions were being invoked. In 2006, 60 percent of imports were estimated to have entered the country with partial or total exemptions, costing the government an equivalent of 11 percent of GDP and 66 percent of tax revenue. Tax base remained exceptionally narrow, with 22259 large payers accounting for 80 percent of the domestic tax take and 1,500 small and medium-sized payers accounting for the remainder.19 Accession to the EAC in 2007 provided an impetus to accelerate revenue reforms. In 2009, the Office Burundais des Recettes (OBR) was created.20 The establishment of OBR demonstrated the Government of Burundi’s resolve to move in the direction of mobilizing a significantly enhanced contribution of domestic revenues for containing the fiscal deficit and delivering on its ambitious development goals. Symbolically, the Office became the first semiautonomous tax administration agency in Francophone Africa. The Adams Smith Institute, which has been providing technical assistance to OBR, identified problems in the existing system as “out of date legislation, weak governance structures, corrupt tendencies, excessive bureaucracy, opacity over organizational roles and low skills levels.” The Executive Senior Management Team quickly demonstrated leadership and set itself the task of addressing each of these shortcomings from the time the Office became effectively operational in the second half of 2010. A transparent recruitment program for 425 staff was quickly launched, buttressed by an 19 Figures are for 2008. 20 The creation of the OBR was also consistent with the actions of EAC Partner States. In creating the OBR, Burundi has contributed to achieving the EAC objective to harmonize their financial systems and to take measures to harmonize their tax systems with a view to removing tax distortions (Article 82 of the EAC Treaty). 22 intensive and incremental capacity building program. A new organization structure was introduced. Legislation was promoted and passed on VAT, income tax, tax procedures, and taxpayer registration. IT software was introduced and operators trained in customs clearance, back office supports, VAT and issuance of tax identification, and transnational information sharing. A Code of Conduct has been agreed including referral procedures for a disciplinary committee (11 employees, one a Director, have been dismissed under the procedure). A performance pay system based on collections is to be introduced. Delivery through improved performance is impressive. For example, revenue collected from July 2011 to February 2012 was approximately US$246.9 million against a target of US$231.5 million, an increase of seven percent above target and a 24 percent increase from the previous similar period. Systems improvements are notable in cargo clearance, TIN issuance, one stop customs posts, tax payment arrangements, and self-assessment. OBR has now defined its 2013-17 strategic plan. Its first six month target for 2013 matches the entire year’s take for 2009. The Commissioner General is confident of further professionalization and that within the five years of the plan, 95 percent of Burundi’s recurrent expenditure will be met by domestic resource mobilization. The Commissioner General’s powers are expected to be better defined in a decree, including adding the responsibility of Receiver General. Expenditures 13. After years of steady increases in Figure 1.7: Grants and Total Expenditures, public spending - largely financed by aid 2006-2012 grants - Government stabilized the ratio 30 45 Percent of GDP Percent of GDP of spending to GDP by 2010, further 25 40 reducing it in the following years. A large 20 35 15 30 part of the expansion in public spending 10 25 over the last decade has been driven by 5 20 foreign grants (Figure 1.7). Reflecting marked declines in foreign grants, Government reduced public spending in terms of percent of GDP by one percentage Grants (left) point (from 41 to 40 percent of GDP ) in 2011, followed by a sharp contraction in Total public expenditure (right) 2012 of another 5.5 percent of GDP (to 34.6 Data for 2012 is preliminary [Sources: IMF, and authors’ percent of GDP ). The expenditure own calculations. contraction in 2012 (Figure 1.8) more than compensated for the decline in foreign grants (of 2.5 percent of GDP), leading to a narrowing of the fiscal deficit by more than two percent of GDP to 1.7 percent. The bulk of the spending contraction fell on public investment (capital expenditure), which in 2012 it fell by 20 percent from the previous year’s level or more than three percentage points of GDP. Current expenditure declined by nine percent from the previous year’s level (two percentage points of GDP).Total expenditures declined from 45.3 percent of GDP in 2008 to 34.6 in 2012, mainly driven by cuts in goods and services (Figure 1.9). This has helped to reduce the fiscal deficit and maintain macroeconomic stability. However, transfers and subsidies have 23 increased considerably, largely driven by the fuel subsidy. A recent IMF staff report (2013) estimated that fuel subsidies in Burundi to be around 0.7 percent of GDP. Figure 1.8: Total Expenditures, as percent of Figure 1.9 Composition of current spending, GDP as percent of GDP 50.0 10 45.0 Salaries 40.0 8 35.0 6 30.0 Goods and 25.0 4 services 20.0 2 15.0 Transfers 0 10.0 and 2000 2002 2004 2006 2008 2010 2012 5.0 subsidies 0.0 Source: IMF data. Spending as percent of GDP. Source: IMF. Spending as percentage of GDP. 14. Salaries and subsidies remain Figure 1.10: Civilian and security sectors’ wage the single largest item in the national bills, as a share of total GDP, 2001-2012 budget. The wage bill has increased 80 rapidly from 2001 to 2009 (from six to 70 nine percent of GDP) but has since 60 declined slightly to eight percent of GDP. 50 Civilian Military Following the end of the conflict, and in 40 Police line with the agreements on 30 Demobilization, Disarmament, and 20 Reintegration (DDR) program, the share 10 of the wage bill going to the army has 0 been gradually reduced, while the share 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Data for 2012 is preliminary. paid to the security sector (army plus Sources: IMF, and authors’ own calculations. police) has stabilized at around 30 percent of the total wage bill (Figure 1.10). 1.1.5 Debt management 15. Burundi’s external debt situation has improved since the country reached the HIPC completion point in 2009, but the outlook remains fragile. In 2009, Burundi benefited from debt relief reducing its external debt from 94.2 percent to 21.6 percent of GDP.21 At the end of 2012, external debt was expected to stand at 21.4 percent of GDP. Although one of the objectives of the HIPC initiative was to free resources to allow poor countries to execute investment projects with their own resources, unfortunately the country’s investment expenditure 21 In 2009, the country became eligible for US$833 million (net present value) in debt relief (HIPC and MDRI), including US$425 million from IDA and US$28 million from the IMF under the HIPC Initiative, reducing its public and publicly -guaranteed debt by more than 90 percent in net present value terms, and scheduled debt service by some US$30-40 million per year for the next 30 years. MDRI relief cancelled an additional amount of debt outstanding of more than US$120 million in nominal terms to IDA, the IMF, and the AfDB. 24 did not significantly increase as a share of GDP. The latest IMF-World Bank Debt Sustainability Analysis (2013) classifies Burundi as a country with “high risk of debt distress.” According to the DSA, the debt outlook has deteriorated since 2011 as a result of revisions of the macroeconomic outlook due to adverse terms of trade shocks; additional loans contracted at less concessional terms;22 and the lower discount rate applied to calculate present value of debt.23 Key risks identified by the alternative and stress scenarios are narrow export base, lower GDP growth, and unsustainable borrowing requirement (i.e., high fiscal deficits). The document recommends the government to follow a prudent fiscal and debt policy, including the swift adoption and implementation of a reform strategy based on the recent Debt Management Performance Assessment (DeMPA) (see Box 2). Box 2 Burundi Debt Management Performance Assessment (DeMPA): key findings In April 2012, a World Bank team visited Bujumbura, Burundi, to undertake the Debt Management Performance Assessment (DeMPA)—a comprehensive assessment of government debt management operations by applying the tool developed by the World Bank. The DeMPA tool comprises of a set of 15 debt performance indicators (DPIs), which aim to cover a wide spectrum of government debt management operations, as well as the overall environment in which these operations are conducted. Each DPI has one or more dimensions, and in total 35 dimensions are evaluated. The tool assesses each dimension and assigns a score of A, B, or C. The score C reflects that the country’s current practice meets the minimum requirement for effective performance of the tasks being evaluated under the dimension. The score A reflects that the country satisfies sound international practice. The score B is an intermediate score, falling between meeting the minimum requirement and achieving the sound practice.24 The DeMPA assessment in 2012 revealed that Burundi met the minimum requirements or better (the score of A, B, or C) in five dimensions, and did not satisfy the minimum requirements (the score of D) in the remaining twenty-nine dimensions. One dimension— relating to a use of derivatives—was not rated because the country did not use them to manage its portfolio risks. Overall, of more than 80 countries assessed by the World Bank under the DeMPA program so far, Burundi’s debt management appears relatively weak compared to other low-income countries in Africa. Burundi’s debt management was assessed to be strong in the areas of (i) legal framework for debt management; (ii) coordination of debt management policy with fiscal policy; (iii) separation of debt management instruments from monetary policy instruments; (iv) 22 In 2012, the Burundian government signed a credit line of US$80 million with the EXIM Bank of India to finance the construction of a power plant. The loan has a grant element of 31.6 percent (source: IMF EBS/13/7, Supplement 1 January, 2013). 23 The increase in PV of debt due to the last factor is common to all countries for which Low Income Countries (LICs) DSA is performed and not specific to Burundi. 24 For more detail on the DeMPA methodology, see the DeMPA Guide (http://siteresources.worldbank.org/INTDEBTDEPT/Resources/468980-1184253591417/DeMPAGuide.pdf). 25 involvement of a legal counsel in external loan negotiations; and (v) availability of information to domestic market investors. Its relative weakness and areas for improvements included (i) capacity to formulate a debt management strategy without external assistance; (ii) frequent use of domestic financing from the Central Bank; (iii) transparency in procedures for approval of external borrowing, on-lending, and government guarantees; (v) management of operational risks; and (vi) recording and reporting of debt. The latest DeMPA was a follow up to the baseline assessment in 2009. Although the scoring of each dimension could not be directly compared because of the revision of the tool in the intervening years, improvements were evident in several areas, most notably in the areas highlighted as the strengths above. The ongoing reorganization of the Ministry of Finance and Planning, the report noted, also provided an opportunity to consolidate the entities involved in debt management and focus capacity building effort. 1.2 Risks and Challenges Ahead 1.2.1 Risks to the economic outlook 16. Burundi remains susceptible and vulnerable to exogenous and domestic shocks. As will be discussed in the following chapter, these shocks can come from the external economic environment such as rising prices on critical imports such as food and fuel products, falling prices on key exports such as coffee and tea, and declining foreign aid. The external current account balance is vulnerable to these shocks as the country has a very narrow export base that is exposed to volatile terms of trade and is already running large trade deficits. Another type of shock could be economic slowdown in other countries with Burundi emigrants (e.g., in Europe with its current fiscal contraction and potential worsening in the Eurozone) which could lead to further reducing the already low levels of workers’ remittances. The domestic price level is also vulnerable as inflation is sensitive to rises in international food and fuel prices as well as to depreciation of the Burundi Franc. Donors’ commitment to provide support depends on their perceptions about the Government’s commitment to governance reforms, as well as new developments in the Eurozone crisis. Likewise, the impact of climate change on the agriculture sector and the overall economy cannot be neglected, as insufficient rainfall can lead to falling agricultural production, income declines for farmers and food price increases for households, hurting in particular urban households. 17. Such exogenous shocks could also have a severe fiscal impact. Shocks will often lead to demands that the public sector cushions some of the impact on the private economy. For instance, rapid increases in prices on imported fuel and food products as well as costs of electricity production are in many countries only partially passed on to the consumers as Government reduces excise taxes on these products (the former) or expands a subsidy (the latter). This worsens the fiscal balance as it lowers fiscal revenue intake and increases public spending, and is not sustainable as a long-term measure. 18. The limited degree of export diversification remains a problem but is something the Government is focused on addressing, including as part of the ongoing regional integration 26 process. The current reform momentum that is improving the business environment could induce broad based growth and a more diversified economy, which will reduce the country’s vulnerability to exogenous shocks over time. Also, the ongoing expansion of electricity supply by Government investment (and in some cases PPPs) will over time help facilitate growth in a number of industries, in particular manufacturing. Finally, the private sector has the potential to act as the main driver of growth in the coming years, particularly if the Government continues the momentum of reforms in the business environment. 1.2.2 Boosting of tax revenue, harmonization of incentives, and reduction in exemptions 19. Financing of the 2012–15 PRSP II requires significant additional resources. Current expenditure is projected to rise over the period by an average of 8.2 percent per year. Capital expenditure is expected to rise by 18.9 percent per year (although expected to be largely funded by external financial assistance/year). Consequently, increasing tax revenue is a priority in the PRSP. However, there are some counter indications, casting doubt on the Governments’ commitment to this priority, including the extension of exemptions of fuel excise tax, and the broadening of exemptions in the new personal income tax bill. The OBR’s response is to expand the tax base, improve compliance, streamline collection processes and integrate a larger segment of the economy into the formal sector. From a year-on-year estimated rise in tax revenue by 17 percent in 2012-2013(from FBu 527.5 billion to FBu 615 billion), the OBR has set ambitious growth target for domestic resource mobilization: 16 percent for 2014-15 and 20 percent for 2016-17. 20. The distribution of revenue collection between income tax/goods and services taxes requires attention, reflecting increasing difficulties to increase the tax base in a fragile, informal and rural environment. Compared to the other members of the EAC, Burundi underperforms with regard to the rate of tax collection growth on income, profits, and capital gains. OBR is geared to expand the tax base and has recently registered 7,000 new taxpayers. However, baselines and annual targeted growth were yet to be identified in the strategic plan. Moreover, recent changes to the tax base associated with the introduction of new exemptions (under the 2013 income tax law), combined with generous tax provisions under the Investment Code have undermined previous gains and contributed to reduce revenue up to one percent of GDP (IMF 2013). This in turn can undermine government’s ability to invest in public goods and infrastructure projects that can have a positive, lasting impact on the investment climate. Table 1.3 shows that, between January and May 2013, the OBR collected FBu 216.7 billion, which is 1.3 percent less in nominal terms than the total tax collection for the same period in 2012. In May, the OBR collected FBu 36.9 billion, which is 19.71 percent less than forecast (FBu 45.98 billion). In order to partially compensate for the revenue foregone as a result of the new income tax law, the OBR has proposed to the Government to increase excises on alcohol and tobacco, as well as increasing the VAT on fuel, and suppress the duty exemption on imported sugar (OBR 2013). These proposals are consistent with similar requests made by the IMF mission (IMF 2013). 27 Table 1.3 Monthly tax collection between January and May, 2010-2013 2013 2012 2011 2010 1st Quarter 140,793,168,830 145,290,535,530 123,334,513,302 95,051,694,851 April 38,989,907,450 37,101,505,521 46,028,565,604 22,955,054,657 May 36,916,321,966 36,785,248,407 38,035,549,884 28,256,788,007 Total 216,699,398,246 219,177,289,458 207,398,628,790 146,263,537,515 Source: OBR. Figures are expressed in nominal FBu. 21. Recent data from the OBR show that tax exemptions continue to grow in absolute terms, calling for urgent efforts to streamline them. According to OBR official data (table 1.4), tax exemptions on VAT on imports, custom duties, and redevances administratives have passed from FBu 64.7 billion in 2009 (to FBu 106.3 billion (US$ 70 million) in 2012. This is equivalent to about 3 % of GDP or about one-fifth of the total revenue collected by the OBR. Preliminary data for the first three quarter of 2013 seem to confirm this trend, suggesting that current tax incentive regimes are generous and are likely to contribute to a significant amount of lost revenues for the authorities. Beneficiaries of tax exemptions are NGOs, private sector investors, members of Governments, diplomatic missions, and Government service providers. OBR data also indicate that service providers and contractors are among the largest beneficiaries of the tax exemptions, and the category where most of the exemptions granted without Parliament’s approval are reported. 25 Overall, these findings indicate the urgent need for the Government to rationalize the system of tax incentives. Table 1.4: Tax Exemptions by beneficiaries, 2009-201326 Beneficiary Amount (in % of total % of GDP % of FBus) exemptions total revenue 2009 NGO’s 5,049,337,360 7.8 Investors (exemptions granted through 19,811,737,133 30.61 Investment Code) Members of Government (exemptions on 71,553,764 0.1 goods/services purchased) Government service providers, including 28,397,617,645 43.9 investment projects Diplomatic missions 2,282,906,821 3.5 Others 9,108,269,960 14.1 2009 Total 64,721,422,683 100 3.02 21.49 2010 NGO’s 87414 006 0.2 Investors (exemptions granted through 10,843,242,773 25.4 Investment Code) Members of Government (exemptions on 45,761,118 0.1 25 “Exemptions granted without the approval of the National Assembly” refer to ad-hoc exemptions granted to contractors through ministerial ordinances, outside the formal mechanisms established by the current legislation. 26 The OBR changed the classification of sectors in 2012. For comparability purposes, the category “Government services providers, including investment projects,” included in 2009-2011, is equivalent to the sum of the “Donor- funded projects” plus the “Investment projects financed with the State’s own resources” category. 28 goods/services purchased) Government service providers, including 24,129,139,445 56.4 investment projects Diplomatic missions 7,405,302,426 17.3 Others 263,284,776 0.6 2010 Total 42,774,144,544 100 1.71 11.79 2011 NGOs 1,461,320 0.001 Investors (exemptions granted through 18,643,703,869 18.4 Investment Code) Members of Government (exemptions on 34,006,101 0.03 goods/services purchased) Government service providers, including 49,969,997,041 49.3 investment projects Diplomatic missions 32,540,259,909 32.1 Others 74,390,335 0.07 2011 Total 101,261,818,575 100 3.43 21.47 2012 NGOs and diplomatic missions 5,869,761,339 5.5 Non-Profit Organizations 1,420,644,985 1.3 Investors (exemptions granted through 25,700,424,589 24.1 Investment Code) Donor- funded projects 21,399,096,514 20.1 Purchase of medicines 8,170,393,836 7.7 Investment projects financed with State’s 41,603,461,251 39.1 own resources Others 2,195,164,556 2.1 2012 Total 106,358,947,070 100 2.98 20.20 2013 (Q1-Q3 data) NGOs and Diplomatic missions 6,038,775,420 7.4 Non-Profit Organizations 1,694,968,215 2.1 Investors (exemptions granted through 14,957,381,058 18.4 Investment Code) Government, external funding, grants 26,944,718,623 33.1 Construction projects from external 7,555,805,028 9,3 funding Purchase of medicines 6,793,292,948 8.4 Others 3,796,329,083 4.7 Exemptions granted by the Government 13,525,945,825 16.6 2013 Total 81,307,216,200 100 N/A 19,23 Source: OBR. Please note that the 2013 data only include data from Q1-Q3. Moreover, these data do not cover all the exemption categories provided under the current tax policy regime. Therefore, further studies are needed to provide a detailed analysis of all tax exemptions (by categories and beneficiaries). 22. Tax exemptions/incentives have been justified on the basis that they will help boost private sector development and alleviate the cost of living of vulnerable households. 29 However, the effectiveness of such regimes remains to be assessed. While the number of registered Small and Medium Enterprises (SMEs) has grown in Burundi, it is not clear whether the tax incentives have been the main driver of this process. Moreover, no assessment have been made on the performance of the new enterprises in terms of quality and quality of jobs created, and capability to contribute to overall economic growth. A broad consensus is therefore growing among key policy-makers and stakeholders (including Development Partners) that a closer look should be given to the efficiency of the existing tax regimes and their contribution in promoting sustainable investments in Burundi. 22. The administration of tax incentives in Burundi is also difficult, as provisions on tax incentives are dispersed across several laws and granted by different institutions. Tax exemptions are divided into three main categories (Africa Research Institute 2013): (i) exemptions that are granted by law or through adherence to an international agreement ratified by Parliament and entered into law. They have legal standing in domestic law; (ii) exemption granted to encourage a particular type of investment under the Investment Code. The API is legally entitled to offer exemptions on domestic taxation without prior approval from Parliament; and (iii) ad-hoc or ‘discretionary’ exemptions written in the contracts of agreement between the government and a particular contractor. These exemptions are granted through ministerial ordinances, outside the current legal framework27. These practices create confusions for investors and undermine cooperation among government institutions, suggesting that further efforts are needed to strengthen cooperation among key actors (MoF, OBR, API)28 and improve the administration of tax expenditures. Further efforts are also needed to revise some generous provisions under the Investment Code, as well as to incorporate the calculation of tax expenditure in the yearly budget discussion, to foster a public debate on the expected costs and benefits of tax incentives. 27 The Ministry of Finance has sent an official letter to all ministries stating that exemptions of this nature should not occur, and if they are to do so they require written approval from the Minister. Despite these efforts, evidence suggests that these discretionary exemptions still occur and represent a significant amount of total exemption granted by the Government. (Source: authors’ interview with OBR officials). 28 In 2012, OBR and API have established a joint committee in charge of overseeing the administration of tax incentives under the investment code, to ensure a smoother management and interface with investors 30 1.2.3 Improving Government Effectiveness 23. Despite remarkable Figure 1.11: Burundi Governance at a glance, 2011 achievements in the post-conflict data period, 29 the quality of governance Voice and remains a key development challenge Accountabi Percentile for Burundi, undermining its lity Rank Burundi 40 Political (0-100) economic and social potential and Control of 30 Stability/A 20 increasing the risks of conflict and Corruption 10 bsence of Violence fragility. The 2011 World Governance 0 Governme Percentile Indicators provide a quick snapshot of Rule of nt Rank Africa Law Effectivene the challenges confronting Burundi. The ss regional country lags well behind the African Regulatory average (0- Quality regional average on most dimensions of 100) governance. In particular, Burundi ranks in the lowest tenth percentile on Source: World Governance Indicators, 2011. Government Effectiveness, the Rule of Law, Regulatory Quality, and Control of Corruption. While Political Stability has substantially improved, concerns over security remain. 30 However, the country performs relatively well on Voice and Accountability, in itself a major achievement in a fragile post-conflict setting. Overall, the picture suggests that the Government has been right to identify good governance and the rule of law as strategic pillars for development and means to preserve the gains from improved stability.31 24. Economic performance will depend on improvements in state effectiveness. The performance of the public administration represents another area of concern and has been identified by the government as an area requiring urgent reforms. This is reflected in the World Bank’s CPIA scores: despite the fact that Burundi scores better than the average fragile states in the region (CPIA score 3.1 against an average of 2.7), the country remains still below the average IDA-eligible country in terms of overall institutional capacity, as well as below the average of EAC countries (Figure 1.13). This is largely explained by the low performance on PFM and governance (Figure 1.12). 29 Since the 2000 Arusha Peace agreement, the country has made remarkable achievements in consolidating peace and security. In 2005, the Parliament approved the new Constitution emphasizing power sharing and local development. In 2006, a cease fire agreement with signed between the government and the last rebel movement, effectively ending the hostilities. In 2010, the country successfully held the presidential, parliamentary, and local elections. The reconciliation process has been strengthened by the creation of governance related institutions, such as the Ombudsman’s Office, the National Independent Human Rights Commission, and the Independent Electoral Commission. 30 In early 2012, for example, the U.N. Security Council expressed concerned over alleged extrajudicial killings by security forces. Reflecting broader security concerns, the Government has deployed a battalion at the Congo border. 31 This approach has been formalized in 2011 with the approval of the National Strategy for Good Governance and Fight against Corruption. 31 Figure 1.12: Burundi CPIA, 2005 - 2011 Figure 1.13: Burundi CPIA in comparative perspective (2011) 3.6 4 3.7 Economic 3.4 3.5 3.1 3.4 Management 3 2.7 Structural 3.2 2.5 policies 3.0 Social 2 Inclusion 1.5 2.8 PFM/governa 1 2.6 nce 0.5 CPIA 2.4 0 aggregate Burundi LIFS EAC IDA Africa* average average Source: World Bank Open Data. Source: World Bank Open data, FY11.32 32 Average calculated out of 77 IDA-eligible countries (full list) and 25 fragile states. Fragile states are defined as having either a) a harmonized average CPIA country rating of 3.2 or less, or b) the presence of a U.N. and/or regional peacekeeping or peace building mission during the past three years. 32 Chapter 2 : Drivers of Fragility The manifest over reliance on aid in Burundi carries a significant element of medium- term risk. It similarly retains the potential to undermine incentives for necessary domestic measures and behaviors. Volatility in budget support can be partially addressed by uprating treasury management but should be matched by more realistic budgeting. The abiding issue critically impacts liquidity (and variable in-year levels of reserves) in turn stoking a worsening arrears problem. The possibility of oil price shocks remains, with disruptive ramifications. Diversification of export production and trade outlets is a key policy zone. Food production issues are problematic and more pressing when linked with climate change. 2.1 High Dependency on Foreign Aid and its Volatility 1. Burundi is one of the Table 2.1 The world’s most aid-dependent world’s most aid-dependent countries ( fragile states are in bold) countries measured by its ODA to Rank Country ODA to GDP GDP ratio. The situation is ratio conditioned by the country’s less (2000- pronounced position as a per capita 2010 recipient of official development average) assistance (ODA) amongst the fragile 1 Liberia 72.4% states group, but high aid reliance is 2 Tuvalu 43.7% nonetheless a defining current 3 Micronesia* 42.2% 4 Burundi 38.9% characteristic of the national 5 Marshall Islands* 37.0% economy. While, in absolute terms 6 Afghanistan 36.9% Burundi is receiving only about one 7 Timor-Leste* 36.5% percent of total ODA to fragile states 8 Solomon Islands* 33.6% (Figure 2.3), it is the fourth most aid- 9 Sierra Leone 31.4% dependent country in the world and 10 Congo, Dem. Rep. 27.1% the second among the low-income 11 Mozambique 25.7% fragile states group, after Liberia 12 West Bank and Gaza* 25.5% (Table 2.1 and Figure 2.1, 13 Sao Tome and Principe 24.8% 14 Iraq* 22.8% respectively). Dependency is also 15 Palau 22.6% high in terms of per capita aid 16 Guinea-Bissau 22.3% inflows, with Burundi ranking third 17 Eritrea 22.2% within the LIFC group (Figure 2.2). 18 Malawi 21.1% 19 Rwanda 19.9% 20 Haiti 17.1% Note: (*) denotes middle-income fragile state. Source: OECD (2010). 33 Figure 2.1: Burundi in comparative Figure 2.2: Burundi in comparative perspective: ODA to GDP ratio perspective: Per capita aid inflows 80% 400,000 70% 350,000 60% 50% 300,000 40% 250,000 30% 200,000 20% 150,000 10% 100,000 0% 50,000 0 Source: OECD (2010) Data reflects 2000 - 2011. Source: OECD (2010) Data reflects 2000 - 2011, Benchmarks include EAC and a selection of African LIFC. Figure 2.3: ODA to Fragile States, 2011 data Afghanistan Pakistan Iraq Bangladesh Uganda Rwanda Nepal Niger Burundi Bosnia-… Sierra Leone Kyrgyz… Angola Eritrea Marshall… Kiribati 0 2000 4000 6000 8000 (US$ million current) Source: OECD (2012). 2. Aid dependency is not mitigated by the contribution of either remittances or foreign direct investment. Remittances as a source of external finance have become increasingly important for fragile states. Kenya and Sudan in the general region are in the top ten of such aggregate contributions. Their influence is more marginal in Burundi (Figure 2.4) and shows little increased potential foreign direct investment (FDI) likewise is notably low at less than one percent of GDP and only recorded US$3.4 million in 2011. There is more room for future optimism in tourism and mining but this will be sensitive to further governance and sector reforms. Trade revenue also does not dent dependency and (as discussed in the previous chapter) is susceptible to commodity price fluctuations and other system shocks. 34 Figure 2.4: Remittances inflows Figure 2.5: Aid inflows in Burundi 10.0% 200 8.0% (as % of GDP) 100 6.0% (as % of GDP) 4.0% 0 2.0% 0.0% 2008 2009 2010 2011 2012 -100 20042005200620072008200920102011 Burundi EAC (average) FDI net inflows (RHS) LIFS (average) ODA Source: WB, IMF and OECD data; World Bank Migration and Remittances Data and IMF WEO data. 3. The OECD 2012 report underscores the expected continuing fall (it declined 2.7 percent in 2011 in real terms excluding debt relief) in country programmable aid worldwide. Burundi’s own trajectory of receipts has generally not evidenced perturbation in the post-conflict period. Further, it is projected to experience a modest growth in annual average 2012-15 (0.7 percent). However, the country is in no way insulated from the global trend and recently in 2012 experienced some lessening of capital spending support as well as delays in the finalization of commitments following the Geneva donor conference. 4. In making good its commitment to decoupling from aid dependence, the Government must embed behavioral change, demonstrating that declining ODA can be used to trigger other flows. The country’s structural and governance reforms in the post- conflict period have been continually assessed as positive. Efforts in institution building, including with OBR, on PFM, in commercial arbitration, statistics management, public procurement, et al., have signaled that the Government is moving ahead with the necessary reforms to bolster the business environment (the recent marked progress on the “Doing Business” criteria reflect this). A tendency remains, as elsewhere, however, to readily entreat the development partners for continuing support with financing shortfalls. Such is evident, for example, in the framing of the PRSP II and, most recently, in the IMF Memorandum on Economic and Financial Policies. Maintaining the momentum on the Government’s part for the reduction in transaction costs on basic business procedures and other institutional bottlenecks, while generally promoting increased liberalization, will open the way to stimulating higher growth and obviating the need for excessive (and likely fruitless) future calls on aid. 35 2.2 The Volatility of Budget Support 5. Donor-funded projects and budget Figure 2.6: Burundi: Growth Rates of support are extremely volatile, undemanding Projects and Program Support, 2002 - 2012 implementation of service delivery and public investments. Project aid is particularly volatile. 350% 300% As Figure 2.6 shows, expressed in real terms per 250% capita, growth rates of project and budget 200% 150% support are highly volatile for both flows. 100% 50% 0% -50% -100% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Projects Program Source: IMF data Figure 2.7: Burundi: Budget Support as a 6. Budget support represents 10 percent of Percentage of Total Budget, 2007 - 2012 Burundi’s budget. Though its share of the budget has declined over the last years (Figure 2.6), budget 20.0 support still represents 10 percent of Burundi’s total 18.0 16.0 budget and stands at around three percent of GDP. 14.0 Budget support is still a crucial component of 12.0 Burundi’s budget and is critical for adequate 10.0 service delivery. The providers of budget support in 8.0 2012 were the African Development Bank, France, 6.0 Norway, and the World Bank. 4.0 2.0 0.0 2007 2008 2009 2010 2011 2012 Source: IMF data 7. Budget supports arrive late in the Figure 2.8: Budget Supports Received at the year. The providers of budget support wire Central Bank (cumulative flow), 2003 - 2012 the funds to the Government’s account in 140,000 the Central Bank of Burundi. The monthly 120,000 series of wires received shows that budget 100,000 supports arrive either in the last quarter of a given year, or even at the beginning of 80,000 the next one. In order to compensate for 60,000 the temporary shortfall of revenue, the 40,000 Government requests cash advances to the Central Bank, which must then increase 20,000 the monetary base to fulfill the 0 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Government’s financing needs without reducing the liquidity of the economy. Source: Banque de la République du Burundi. Unit : Burundi’s M2/GDP ratio is extremely low Millions of Burundi Franc 36 (25 percent at the end of 2011), which means that the economy is highly sensitive to changes in the monetary policy. By way of illustration, a budget support supposed to arrive in 2011 arrived in January of 2012 but was not used to execute programmed expenditures. Sources in the Treasury and in the Central Bank indicated it was used to repay the cash advances provided by the Central Bank during 2011, suggesting a further link to seasonality of revenue collection. Two more budget supports arrived later in 2012, but they were not large enough to provide enough liquidity to the Treasury. The largest budget support operation programmed for 2012 was actually disbursed in February 2013 and was used, again, to pay the cash advances provided by the Central Bank during the previous year. Figure 2.9: Revenues and Budget support Figure 2.10: Government’s Net Position (flows), 2011 - 2012 with Domestic Financial Corporation’s, 60,000.0 Revenue collected by OBR 2003 - 2012 Appuis budgetaires 0 50,000.0 -50,000 40,000.0 -100,000 30,000.0 -150,000 20,000.0 -200,000 10,000.0 - -250,000 2007 Aug 2012 Aug 2005 Dec 2006 Oct 2010 Dec 2011 Oct 2004 Apr 2007 Mar 2009 Apr 2012 Mar 2003 Jan 2003 Jun 2004 Sep 2005 Feb 2008 Jan 2008 Jun 2009 Sep 2010 Feb 2003 Nov 2005 Jul 2006 May 2008 Nov 2010 Jul 2011 May Source: Banque de la République du Burundi, and Office Under the IFS nomenclature, the term “Financial Burundais de Recettes. Unit : Millions of Burundi Franc. Corporations” refers to the Central Bank and the banking institutions. Source : IMF International Financial Statistics, IMF. Unit : Millions of Burundi Franc. 8. Several factors explain why budget support is disbursed late. Consultations with representatives of donors who provide budget support highlighted the following as the potential reasons behind the delays in disbursing: (i) donors’ internal administrative procedures prior to disbursement; (ii) macroeconomic data and monitoring reports needed to process the budget support are not produced on a timely basis by the Burundi administration, making it difficult to assess progress against agreed targets, policy actions, and the like; and, (iii) the IMF mission, and hence the comfort letter required to process the budget support, takes place late in the year. 9. The intra-year volatility of budget supports calls for better treasury management and more realistic budgets. Some of the factors underlying the date of disbursement of donors’ budget support are exogenous to the Burundi public administration and are unlikely to change, particularly the ones related with donors’ internal procedures. The seasonality pattern identified requires a better planning regimen. But if budget expansion pressures continue to burgeon, planning will still be frustrated. Less ambitious budgets that take into account the structural constraints of the Burundi administration are also necessary. 37 2.3 Growing Stock of Arrears 10. The unavoidable pressures placed on budget execution by late aid disbursements, mean that the Government has payments pending at the end of the year any or all of which potentially become arrears. Data extracted from the Treasury balances presented in the Loi de Règlement indicates that this is exactly what has happened and the arrears level (restes à payer) has grown increasingly large. 11. High frequency data indicates that the Government faces a permanent liquidity constraint motivating the fall back to arrears. Burundian policymakers have three non- excludable options to face the liquidity situation: (for a full analysis, see Annex B): • Ask for cash advances to the Central Bank to cover the deficit. This option has the inconvenience of exacerbating the inflationary and exchange rate pressures suffered by the Burundi economy. • Execute the budget with parsimony and forget about the pending payments, effectively accumulating arrears. In a best case scenario, the Government will dispose of the pending payments and the arrears at the arrival of the donors’ budget support or whenever it receives some extraordinary revenues such as privatizations. In a worst case scenario, the Government will dispose of the arrears very slowly. This option has the inconvenience of putting stress on the real economy, as arrears are mostly owed to providers of goods and services. • Dispose of some or all of the pending payments and unavoidable commitments (salaries, interest payments) while postponing nonessential expenditures (i.e., investment projects) and delaying payments to the providers of goods and services. This has the double inconvenience of putting stress on the real economy and deactivating infrastructural development. Data extracted from the treasury accounts Figure 2.11: Stock of arrears (millions of FBu) presented in the Loi de Règlement for the 14,000 period 2008-11 shows that the Government 12,000 is accumulating arrears (Figure 2.11). 33 In 10,000 2008, the stock of arrears was FBu 351 (FBu mn) 8,000 million. At the end of 2011, the stock of 6,000 arrears increased up to FBu 13,093 million. Arrears owed to providers of goods and 4,000 services (fournisseurs) and contractors 2,000 (fournisseurs – immobilisations) amounted to 0 FBu 4,588.8 million at the end of 2011. Labor 2008 2009 2010 2011 related arrears (Salaires – Pensions et Source: Loi de Règlement 2008-2011. Annex 1: Balance indemnities, INSS) have traditionally been low Cumulée des Comptes. with the exception of 2011, when they increased considerably. At the end of 2011, the largest creditor of arrears was “Other Organizations” followed by providers of goods and services (fournisseurs). 33 See Annex B for accounting and methodological details used to estimate the arrears. This methodology has been discussed with counterparts from the Ministry of Finance. Data for 2012 was not available at the time this report was written. 38 Box 3 A History of the Treatment of Arrears in Burundi1 Prior to 2008, arrears and payments pending at the end of the year were dealt in an unorthodox way. Prior to the 2008 reform, the budget included an item called “payment of arrears.” This item canceled, in theory, the operations that went through the ELO steps of the chaîne de la dépense in previous years. Under this methodology, expenditures were reengaged (reengagement), liquidated (liquidation), and committed (ordonnancement) as if they took place in the current year. This process was problematic for five reasons: a. “Arrears” is not an economic category in the GFSM-2001 budget classification, which the Government committed to implement fully in 2005. b. This process was not transparent from a budgeting and accounting perspective, as it did not allow assessment of whether the expenditures carried out under the line “arrears” were, indeed, payments made to orders previously authorized (contrôlé et prise en charge) in a regular fashion. c. This process does not respect budget orthodoxy. Expenditures are authorized by an annual budget law and executed on an annual basis. Under the French accounting system, the expenditures that go through the chaîne de la dépense should not be reintroduced in a budget law after the fact. d. This process did not allow producing regular or reliable statistics regarding the amount, age, or nature of arrears. With the modification of the 2005 Plan Budgétaire et Comptable (approved in 2008 and fully in place since 2010), the Ministry of Finance modified the process to treat arrears payments, which is now in line with international accounting standards. The modification of the Plan included the suppression of the economic category “arrears,” which had not appeared in the budget previsions since 2010. Under the current regime, the balances of the accounts debited in year N are reengaged in the same accounts under the rubric “previous years,” which respects the principle of budget annuity. The payment of arrears and pending payments is now an operation of treasury under the responsibility of the main accountant (the Direction de la Comptabilité et du Trésor, formerly OTBu). 2.4 Critical Level of Reserves 12. The reserves accumulation policy of the Central Bank is likewise compromised by the demands of cash advances from the Government. Though cash advances do not generate interests to the Government, the Central Bank needs to increase the monetary base, causing pressure on reserves to provide the economy with liquidity. The Central Bank thus loses reserves throughout the entire year and is able to cumulate them only after the budget support arrives. Budget support is the main driver of reserves accumulation in Burundi. Reserves passed from 39 US$287.24 million on February 25, 2013, to US$322.53 million the day after, when the budget support of the World Bank arrived to the Treasury account in the Central Bank.34 Figure 2.12: Reserves Hold at the BRB and Government’s Position with the BRB, 2003-2012 400 300 350 Total reserves excluding gold (millions of USD) Government's position with BRB (billions of FBU, RHS) 1/ 250 300 Correlation coefficient: 0.4748 200 250 200 150 150 100 100 50 50 0 0 2005 Dec 2010 Dec 2003 Jan 2003 Jun 2004 Sep 2005 Feb 2006 Oct 2008 Jan 2008 Jun 2009 Sep 2010 Feb 2011 Oct 2005 Jul 2003 Nov 2004 Apr 2006 May 2007 Mar 2007 Aug 2008 Nov 2009 Apr 2010 Jul 2011 May 2012 Mar 2012 Aug For ease of reading, the Government’s position is shown with a positive sign. BRB: Banque de la République du Burundi (Central Bank of Burundi). Source: IMF International Financial Statistics. Box 4 Burundi’s Reserves in a Comparative Perspective Emerging Markets (EMs) and Low-Income Countries (LICs) have increased their monetary reserves at a fast pace in the past few years. In LICs, the fast pace of reserves buildup is the consequence of previously low levels of reserves, as well as “increasing openness of economies, a favorable global environment, and (…) policy choice among countries to build precautionary reserves to insure against balance of payments risks (IMF 2011, 34).” The metrics that are used the most to measure reserves adequacy are: • Reserves to GDP. • Ratio of reserves to short-term (12 months on less) debt. Unfortunately, the LICs’ data on short term debt is mostly unreliable. • Ratio of reserves to M2 or broad money. This ratio is the most relevant for countries with a managed exchange rate. 34 The analysis provided here is based on interviewers with representatives from the Central Bank staff. 40 • Ratio of reserves to months of imports. The consensus among policymakers and researchers is that this is the most appropriate metric for LICs. The rule of thumb is to consider a ratio of three months as the minimum appropriate. The usefulness of these ratios is enhanced if they are used in an appropriate cross country context. In the case of Burundi, a pertinent group of comparators would be the members of the G7+ Group, a coalition of fragile post-conflict states of which Burundi is actually a member. In line with most LICs, the Government of Burundi has increased its stock of reserves considerably in the last years to insure against macroeconomic shocks. Figure 2.13: Reserves as a share of GDP Figure 2.14: Reserves as a share of M2 30.0 100.0 Reserves as a Share of GDP Reserves as a share of M2 90.0 25.0 80.0 70.0 Sample average 20.0 60.0 15.0 Sample average 50.0 40.0 10.0 30.0 20.0 5.0 10.0 0.0 0.0 Guinea Timor Leste Chad DRC Central Burundi Papua New Haiti Sierra Leone Comoros Haiti Central Burundi Chad Papua New Sierra Leone Comoros Liberia Afghanistan Timor Leste African Guinea African Guinea Republic Republic Source: IMF International Financial Statistics. Source: IMF International Financial Statistics. Figure 2.15: Reserves to Month of Imports 12.0 Reserves to Months of Imports 10.0 8.0 Sample average 6.0 4.0 2.0 0.0 Source: IMF International Financial Statistics. A comparison between Burundi and other G7+ countries shows that, at the end of 2011, Burundi’s ratios were systematically lower than those of the rest of the group. The reserves-to-months-of imports ratio is a particular source of concern. Burundi was not only below the average, but was barely above the critical level of three months. Burundi’s relatively low level of reserves should be a source of concern. Burundi is a net importer of oil and food while exports are heavily concentrated in only one sector, coffee, which accounts for 66 percent of the total. Also, contrary to other G7+ countries, remittances and FDI represent small shares of GDP, and there is no portfolio investment. Excluding exports, Burundi’s main source of hard currency is donors’ money. 41 2.5 High Dependency on Import and Limited Export Diversification 13. Manufacturing imports play a major role in the Burundian economy. Burundi’s imports cover a wide range of goods and are dominated by machinery, transport equipment, construction materials, other manufactured goods and chemical products (Figure 2.15). Manufactures represent about two-thirds of Burundi’s goods imports. This import composition largely reflects the lack of a domestic manufacturing industry and fuel resources, as well as the limits of a small domestic market that prevents economies of scale. Moreover, the country is a net food importer, mostly of cereals like malt, rice, and wheat. 14. Due to the unprecedented global increase in oil prices, the importance of oil in Burundi’s imports has also increased. Data by the Central Bank (BRB) suggest that oil imports averaged over US$100 million per year between 2009 and 2011, bringing the share of fuels in total imports to 20 percent in 2011. 35 Surprisingly, this trend is not reflected in the Comtrade database, which reports around US$5 million for Burundi’s oil imports per year between 2008 and 2010 (explaining the drop of “Petroleum oils” in Figure 1.1). Partner countries report around US$4 million of annual oil exports to Burundi during the same period. It is unclear what is behind the discrepancies between Comtrade and BRB data;36 but such inconsistencies are illustrative of the huge data problems that obstruct trade analyses in Burundi.37 Figure 2.16: Composition of Burundi’s imports Figure 2.17: Composition of Burundi’s exports Coffee 100% 20% Pharma. / Other Other Tea chemicals 10% 10% Petroleu m oils 1% 0% Metal ores & Gold scrap Motor Cement vehicles Cotton, sugar, Hides and Cereals plants skins Metals and foodstuffs 2001-2003 2001-2003 2008-2010 2008-2010 Source: Comtrade/WTS, HS 2 and 4 digits. 35 See www.brb.bi/se/docs/bulmens/iv2.pdf. 36 This may be due to shortcomings in the reporting of regional exports and reexports of petroleum products. Most oil imported by Burundi corresponds to reexports from Kenya, Tanzania, or South Africa. 42 Table 2.2 Burundi’s top 10 imports and exports, by HS 2-digit chapters (2008 – 2010) Imports Exports 1. Motor vehicles (HS 87) 11.6% 1. Coffee and tea (HS 09) 69% 2. Machinery and mechanical appliances 10.3% 2. Gold (HS 71) 14.1% (HS 84) 3. Electrical equipment (HS 85) 10% 3. Raw hides and skins (HS 41) 2.5% 4. Pharmaceutical products (HS 30) 9.5% 4. Metal ores (HS 26) 1.8% 5. Cement and other mineral products (HS 5. Machinery and mechanical appliances 5.5% 1.7% 25) (parts) (HS 84) 6. Paper and related products (HS 48) 4.4% 6. Cotton (not carded or combed) (HS 52) 1.7% 7. Iron and steel (HS 72) 4.1% 7. Sugar (raw cane sugar) (HS 17) 1.4% 8. Malt and products of the milling industry 4.1% 8. Iron and steel (waste and scrap) (HS 72) 1.3% (HS 11) 9. Articles of iron and steel (HS 73) 3.1% 9. Motor vehicles (HS 87) 0.9% 10. Various textile articles (HS 63) 2.8% 10. Electrical equipment (parts) (HS 85) 0.9% Top 10 share of total imports 65.5% Top 10 share of total exports 95.3% Source: Comtrade/WITS, mirror data for exports. 15. Burundi’s deposits of mineral resources are a potential source of growth, but their contribution to exports remains low and underexploited. Burundi possesses five to six percent of known world reserves of nickel as well as other minerals such as gold, cobalt, tungsten, and ore. Currently, the mines are exploited informally and about 50,000 workers are employed in this sector. Even though mining activities are undertaken in a small scale the sector does contribute to the country’s GDP, with gold exports amounting at 14.1 percent of total exports over the 2008-2010 period (Table 2.2). With the improvement in internal security and transport links, there is investor interest in the mining of nickel and other metal ores. The development of the mining sector, however, remains slow due to infrastructure constraints (energy and transport) and delays in reforming the governance of the sector. To address the latter, the Government is working on improving the legal and regulatory framework by revising the mining code, currently pending approval by the Parliament. In this context, further efforts are needed to establish an adequate fiscal regime to ensure the Government gets the best possible deal for the exploitation of its resources and to promote innovative approaches focused on “local content” provisions - to maximize the sector contribution to create job growth.38 16. Burundi’s very narrow export base and product concentration have not changed notably over the last decade. Burundi’s goods export basket has been relatively stable in the 2000s. Export earnings remain highly dependent on primary products predominantly coffee and tea, which together represented almost 70 percent of total exports in 2008-10 (Figure 1.16). Other traditional exports (hides and skins, cotton and sugar) add some degree of diversity - the share of these products increased from 2.5 percent to 5.6 percent of total exports between 2001- 03 and 2008-10.39 However, the almost total absence of new higher value exports in the last 38 Except for the labor intensive construction phase, extractive industries generally do not generate a high share of employment in the host country. However, domestic firms can find business in a range of supporting sectors, such as transport and equipment manufacturing or maintenance. Innovative approaches focused on “local content” provisions are becoming increasingly popular across developing countries to maximize jobs for nationals, by binding or incentivizing companies to employ and train nationals through contractual obligations and incentives. 39 See World Bank (2011) for more details. 43 decade is striking. Cut flowers emerged as a promising export in 2001, but they never surpassed half a million dollar per year and collapsed in 2008, as they were supplied to only one buyer who cancelled orders in response to low demand during the global crisis (World Bank 2011). The Herfindahl Index (HI)40 for Burundi has remained high throughout the last twenty years. In 2008 - 2010, the top three and 10 export categories represented over 85 and 95 percent of total exports respectively (Table 2.2). 17. Burundi’s exports have not Figure 2.18: Destination of Burundi’s exports shifted away from its traditional (share of total) European partners toward fast growing emerging markets, though the EAC has become an increasingly important region for Burundi’s exports Pakistan since 2001. As opposed to other EAC 30% United countries, which have increased their DRC Arab… 20% exports to Asia (for example, Tanzania’s exports to China and India have increased United 10% EAC significantly in recent years) or to the States 0% Middle East and North Africa (for example, Uganda’s exports to several Belgium Germany MENA countries registered a big jump) (IMF 2011), Burundi’s exports to Other Other EU15 emerging markets remains modest. The BRIC countries’ share in Burundi’s 2001-2003 2008-2010 exports only increased from two to 3.5 percent between 2001-2003 and 2008- 2010. Finally, exports to the U.S. have Source: Authors’ own calculations based on UnctadStat data. decreased over the decade, indicating that Burundi has not managed to take advantage of its eligibility to the African Growth Opportunity Act (AGOA) duty-free quota-free access to the American market since 2006. The share of Burundi’s exports going to EAC countries has increased (Figure 2.20), a dynamic which started in the years preceding Burundi’s accession to the EAC. Official figures underestimate the true levels because of informal cross border flows. 40 The HI measures the export concentration and is calculated as ‫∑ = ܫܪ‬ே ௜ୀଵ ‫ݏ‬௜ where si is the share of the ଶ considered variables (in this case, exports) and N is the total number of products exported. Two factors can lead to a lower HI, a more even distribution of shares, and an increase in the number of products. The HI lies between 0 and 1, where 1 represents exports that are extremely concentrated (only one product), and 0 stands for a completely diversified export basket. Economies with an index below a threshold 0.1 are considered to have a highly diversified export basket. 44 18. In addition to volatile Figure 2.19: International price of Arabica international prices, the country’s export coffee, constant U.S. cents per Kg 600 earnings from coffee depend on the 500 quantity and quality of the coffee harvest 400 each year, which have been declining. 300 Burundi has a biennial output cycle because 200 of aging coffee trees. The quantity and 100 quality of output have suffered in recent 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 years from tree disease, low input use, and low incentives from low producers prices.41 Source: World Bank Global Economic Monitor. 19. Despite its good quality, Burundian coffee farmers have historically Figure 2.20: Price received by Burundian coffee received a lower price for their coffee producers vs. world price (in constant FBU) than the international reference price. The producer price of coffee set by the Burundian government is to some extent unrelated to the world market price of coffee (Figure 2.19). 42 Compared to its neighbors, the gap between the producer price and the market price has been much larger (Figure 2.20). 43 “The situation profoundly discourages farmers” (USAID 2010, p.4). Due to the country’s law, farmers are prohibited from destroying coffee plants. Very Source: Kimonyo and Ntiranyibagira (2007). often, farmers are choosing to neglect their coffee plants in favor of food crops. The state-run coffee agency, OCIBU, annually sets the producer price and is in control of marketing and export. According to some observers, state-led ownership has been a major negative factor on private initiative and incentive to invest (IMF 2008, CEM 2011). 42 The state-run coffee agency, OCIBU, annually sets the producer price and is in control of marketing and export (IMF 2008). The agency argues that Burundian producers have been protected from fluctuations in international prices. Although this is true, the farmers have also not benefited from periodic price increases (USAID 2010). 43 There are different reasons for this, including (i) OCIBU’s control of coffee sales and bad marketing practices; (ii) the small volume of the country’s production; (iii) lack of access to the sea; (iv) the conflict, and according to some; (v) collusion by international traders operating in the Burundian market to maintain low prices (Kimonyo and Ntiranyibagira, 2007). 45 Figure 2.21: Spread between produce and market prices, 2008 Source: AfDB (2010). 20. Despite the ongoing coffee reforms, the sector is suffering from declining output and deterioration of quality. Moreover, aging coffee trees entail a highly cyclical yield - the production attains satisfactory levels only every other year,44 affecting buyers’ perceptions about Burundi’s reliability as a coffee supplier. Finally, the impact of the expected increase in production due to the ongoing coffee reforms is likely to be at least partially offset by the expected volatility in international coffee prices. Restructuring of the tea sector institutions and privatization of the public tea factories are expected to be carried out in the coming years. The tea sector has the potential to contribute more to export earnings, but the current situation suggests that major efforts will be required to improve the quantity and quality of Burundian tea. 2.6 Climate Change and Volatility in Domestic Food Production 21. Wide fluctuations in domestic food production weigh significantly on the budget of the Government. In order to fill the country’s food deficit, the Government has allocated internal public resources to purchase essential food crops for amounts averaging annually 29,200 cereal equivalents (EC) tons since 2010. In particular, during agricultural seasons A where the food deficit often ranges from 68-75 percent of total food usage, the Government’s efforts even went up to 50,000 EC tons of food (MINAGRIE 2012). Yet, the uncovered food deficit is still important despite contributions from donors such as the World Food Program (WFP) and the Food and Agriculture Organization (FAO). The pressure on the Government is large as persistent poor performance in agricultural seasons A have resulted in very low food availability for the population, averaging annually 30 kg per inhabitant since 2010. This is far below the minimum 95 kg per inhabitant (within a six month period) that is advised by the World Health Organization (WHO) for an individual to get a daily food intake of 2,100 kcal. In agricultural seasons B, the food deficit has been reduced in recent years, from 73,900 EC tons in 2010B to 28,600 EC tons in 2012B. 44 Production (expressed in thousands of bags) went from 499, to 133, to 412, to 112, to 347 in crop years 2006/07 to 2010/11. In calendar year 2011, coffee exports strongly decreased in volume, from 307 to 191 thousands of bags (www.ico.org/new_historical.asp?section=Statistics). 46 Table 2.3 Burundi's food balance in recent agricultural seasons (1000 ton of cereal equivalent) 2010A 2010B 2011A 2011B 2012A 2012B Domestic production 261.0 767.2 255.0 792.0 228.0 812.3 Cereal 81.0 211.4 82.0 224.4 70.0 227.0 Leguminous 30.0 172.8 28.0 178.3 26.0 186.3 Root and tuber 116.0 315.9 112.0 321.0 100.0 330.0 Banana and plantain 34.0 67.0 33.0 68.3 32.0 69.0 Total food usage 816.0 936.0 838.0 907.9 816.0 933.2 Food needs 785.0 809 808.0 807.8 802.0 830.4 Seed needs 31.0 127.0 30.0 100.1 14.0 102.9 Food import needs (deficit to cover) 555.0 168.8 583.0 115.9 588.0 120.9 Planned imports by Government 50.0 20.0 35.0 25.0 25.0 20.0 Food aid (WFP) 35.0 28.91 31.1 13.0 26.0 15.4 Other sources 18.0 46.0 18.0 55.0 16.0 57.0 Uncovered food deficit 452.0 73.9 498.9 22.9 521.0 28.6 Food per habitant (kg/habitant) 32 91 30 93 29 102 Source: Compiled from joint food evaluation reports by MINAGRIE, FAO, UNICEF, WFP, and other donors. 22. Food production uncertainty has led the Government to bolster improved seed varieties, but technical shortcomings and the persistence of disease in household produced seeds has undermined efforts. 23. Fragility in the execution of public expenditures originates as well in the difficulties encountered by rural households to access and apply agricultural inputs subsidized by the State. This situation has resulted in a waste of limited public resources, as the Government’s efforts to spread fertilizers and pesticides in a year clearly are not fully translated into effective application on the ground by farmers who lack purchasing power and knowledge. In terms of allocated credits, these efforts almost quadrupled in recent years, moving from a yearly average of FBu 2.5 billion in 2008-2010 to FBu 9.6 billion in 2011-2012. Increasing costs of direct transportation, poor transportation logistics and strategies, poor information campaigns, and resistance to previously unknown fertilizers further undermines the Government’s efforts. Box 5 The Effect of Climate Change on Food Production Regularly stressed by rural households during seasonal food assessments, climate change has a strong bearing on food production in Burundi’s rural areas. Over a period of 33 years of rainfall records, average annual rainfalls have declined, loosing close to 50 millimeters between 1980 and 2012 (Figure 2.21). As a subsistence agriculture of which yields depend primarily on the adequacy of rainfalls, this long decline, which is coupled with wide annual fluctuations, is negatively affecting the outcomes of programs aimed to promote food security and marketable agricultural surpluses. Rainfall variability affects most of the natural regions of Burundi, requiring the 47 Government to allocate funds to specific food security programs. Overall, climate perturbations in Burundi have manifested through the following phenomena: • Uneven distribution of rainfall across the country whereby areas with just normal to excess rainfalls are often accompanied by hail and strong winds that cause significant crop damages in localized areas. • Frequent heavy rains causing flooding, erosion and landslides that result in important direct damage for such food crops (i.e., sweet potato, corn, rice, beans, potato, vegetable, etc.), and even indirect damage related to transportation difficulties due to rain-related landslides in main provincial roads. • Great insufficiency of seeds and fertilizers distributed by donors and NGOs, which the Government attempts to complement with increases in credit allocations destined to cover emergency situations. Figure 2.22: Long-term evolution of rainfall in Burundi (millimeter, yearly average) 1600 1500 1400 (millimeter) 1300 1200 1100 1000 900 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Rainfall (annual average) Source: Bank staff based on IGEBU database. 48 24. The prices of essential food crops have been increasing since the end of 2007. Burundi is a net food importer so any surge in world food prices, albeit with some delays, impacts on the welfare of poor households and child nutrition. Domestic retail prices for basic staple crops have recently increased from 15 percent (beans) to 45.5 percent (wheat) in August 2012, as compared to August 2011 (Figure 2.23), contributing to increases in headline inflation. Beans and cassava powder (Figures 2.24 Figure 2.23: Retail food prices, percent increases 2011- and 2.25, respectively) saw further 2012 significant increases in retail prices from November 2011 to January 2012. Rice, a food crop that is mostly 50.0% procured in EAC neighboring countries like Tanzania and in international markets to complement small local production, experienced 0.0% more rapid and direct price increases Maize Rice Wheat Beans - Retail as of May 2011 (Figure 2.26). To mitigate inflationary trends, Source: FAO estimates. Retails prices computed in BFu. Government decided to subsidize food items with tax relief measures. While this measure achieved short-term positive effects easing social tensions45 and stabilizing consumer price, the elimination of taxes led to a drop in revenues estimated at FBu 10 million, contributing to further reductions in the available fiscal space and financial resources for public investments (IMF Country Report 2012: 58). Figure 2.24: Retail price of beans in Burundi markets 1750 1600 1450 (FBu/kg) 1300 1150 1000 850 700 Bujumbura (the capitale) Province of Ngozi Province of Gitega Province of Kirundo Province of Makamba Source: Bank staff using ISTEEBU’s retail price database. 45 On 27 March, 2012, civil society and trade unions conducted a large strike to protest increases in electricity tariffs and fuel prices. 49 Figure 2.25: Retail price of cassava powder in Burundi markets 900 800 (FBu/kg) 700 600 500 400 300 Bujumbura (the capitale) Province of Ngozi Province of Gitega Province of Kirundo Province of Makamba Source: Bank staff using ISTEEBU’s retail price database. Figure 2.26: Retail price of rice in Burundi markets 2750 2500 2250 2000 (FBu / kg) 1750 1500 1250 1000 750 Province of Kirundo Bujumbura (the capitale) Province of Gitega Province of Ngozi Province of Makamba Source: Bank staff using ISTEEBU’s retail price database. 50 Chapter 3 : Fiscal Sustainability Analysis On the fiscal front, Burundi does not have much room to maneuver. Even with a gradually improving external environment and relatively strong growth, the baseline projection is for the fiscal deficit to rise and for the ratio of debt to GDP to remain high. Moreover, external risks relating to aid, rainfall, and coffee prices could significantly increase the country’s deficits and debt. Changing policy to reduce the expected primary deficit would help build resilience and reduce the country’s exposure to risks stemming from large financing needs. The authorities should persist in implementing planned fiscal reforms— to widen the tax base, eliminate unaffordable exemptions and subsidies, and agree on a sustainable policy for civil service wages. 3.1 Key Findings and Assessment 1. Burundi does not have much fiscal room to maneuver. The baseline analysis for 2013–17 shows that the government will continue to struggle to create fiscal space and reduce debt from its currently high level. The dynamics are mainly explained by falling grants as a share of GDP combined with rigidities in current expenditure, mostly the wage bill. Although the government is expected to limit wage growth in real terms, unless additional revenue measures are introduced to offset the falling grants, it will struggle to reduce its deficit. While the deficit is expected to be largely financed by external concessional loans, debt will remain high as a share of GDP, and will likely keep the country at “high risk of debt distress.” 2. External risks relating to aid, rainfall, and coffee prices could significantly increase the country’s deficits and debt. Stochastic simulations show that, under very adverse shocks to rainfall or coffee prices, the country’s debt could reach 39 to 41 percent of GDP by 2017—about 5 to 7 percentage points above the projected ratio of debt to GDP under the baseline. Aid shocks could have a bigger impact—with debt to GDP reaching as high as 50 percent under adverse shocks—but the estimates are the least statistically reliable because of the time series is short and because its historical volatility in part reflects endogenous shocks (e.g., reform implementation) as well as exogenous shocks (e.g., donors’ behavior). 3. The analysis suggests that the government should focus on building buffers, increasing flexibility, and reducing exposure to risk. Countries that are exposed to shocks should build buffers to cope with the shocks. Smaller fiscal deficits would allow the government more flexibility to use discretionary policies to respond when hit by a shock. Larger foreign reserves would function as a shock absorber in response to negative external shocks. Lower and more stable inflation would give the authorities greater scope for monetary loosening and fiscal stimulus without triggering inflationary expectations and wage demands. Countries with lower risk of debt distress will have greater access to both concessional and non-concessional emergency financing. 51 4. To these ends, the authorities should persist in implementing planned fiscal reforms to widen the tax base and eliminate unaffordable exemptions and subsidies. The policy scenario analysis shows that future fiscal sustainability will hinge on the government’s ability to stick to its plans to increase revenue. The authorities’ effort to widen the tax base has recently suffered a setback, however, as the Parliament introduced last minute exemptions in the tax reform bill. Burundi’s tax exemptions as a share of GDP remain among the highest in the region. 5. The authorities should also agree on a sustainable policy for civil service wages. Fiscal space can be created not just through revenue increases but also through expenditure cuts and improved efficiency. The growing civil service wage bill has created a significant rigidity in government spending and exacerbated the economic impact of external shocks. Options for reforming the salary scale currently under discussion are aimed at restoring fairness and improving incentives for better public sector performance, and as such, will not generate savings in the short run. A very preliminary estimate of the potential fiscal implications of the wage reform shown in this chapter suggests that some reform options could be too expensive. In addition to considering the one-off adjustment costs of the salary scale options under consideration, the authorities should carefully study the dynamic costs and savings under each option once all the details have been worked out. 3.2 The Baseline Scenario Analysis 3.2.1 Macroeconomic assumptions 6. The assumptions underlying the projections in this analysis of fiscal sustainability are in line with the IMF’s most recent review under the Extended Credit Facility (March 2013). The main macroeconomic assumptions for the baseline are shown in Table 3.1. Table 3.1 Baseline Projections: Macroeconomic Assumptions Act Prelim Proj 2011 2012 2013 2014 2015 2016 2017 2013-17 Real GDP growth rate (%) 4.2 4.0 4.5 5.1 5.5 5.5 5.5 Inflation (%) 14.3 15.4 11.1 7.8 6.0 5.5 8.3 Nominal exchange rate (LCU/ USD) 1,261.1 1,440.6 1,596.6 1,622.1 1,637.0 1,653.6 1,641.6 1,630.2 Coffee price (US cents per lb) 273.2 187.6 158.6 171.9 177.9 158.0 158.0 Oil (US$/barrel) 104.0 105.0 102.7 98.5 94.7 91.9 89.4 Source: IMF (March 2013) and Joint World Bank–IMF Burundi Debt Sustainability Analysis (January 2013). 7. The macroeconomic scenario reflects improved external conditions, with moderate recovery in global growth, continued easing of international food and fuel prices, and strong regional economic activities. Real GDP growth is expected to accelerate over the five- year projection period starting in 2013—from four percent in 2012 to 4.5 percent 2013, 5.1 percent in 2014, and 5.5 percent in 2015–17. Agricultural output (35 percent of GDP) will continue to dominate economic performance. Recent developments in agro-industry, the construction of the hydroelectric dam in 2013, and the construction of several road projects in the following years, will sustain moderate growth in the secondary sector. Ongoing integration with the East African Community will support growth in wholesale and retail trade while also spurring investments in the tertiary sector more generally.46 Annual average inflation is expected 46 Background information provided by EIU, CPIA, and IMF. 52 to fall from 15.4 percent in 2012 to 11.1 percent in 2013 and to an average of 7.7 percent thereafter. The nominal exchange rate is expected to depreciate against the U.S. dollar at an annual average rate of 2.7 percent. The coffee price is expected to bottom out in 2013 and then to fluctuate around the medium term average of $1.65per pound. The oil price is expected to fall throughout the period, from US$105 per barrel in 2012 to US$89.4 per barrel in 2017. 3.2.2 Fiscal policy 8. The Government is expected to continue implementing fiscal and structural reforms. Tax revenue is assumed to grow at 5.2 percent a year on average in real terms for the next five years. Nontax revenue—mainly revenues from the sale of state assets, especially in the coffee sector—will generate additional revenue of around one percent of GDP in each year. Grants are assumed to decline as a share of GDP. The Government is expected to limit the growth of wages for the civil service and the military. Total expenditure will decline as a share of GDP reflecting declines of similar magnitude in both current and capital expenditures. Ongoing public financial management reform is expected to enhance the effectiveness of public expenditure, offsetting the impact of the (relative) decline.47 9. The fiscal deficit over the next five years is expected to be financed mainly by grants and concessional borrowing from multilateral and bilateral creditors. Grants are expected to decline as a share of GDP as the country graduates from the post-conflict status and humanitarian assistance winds down. The Government is expected to finance the remaining deficit mainly with highly concessional loans with grant element of at least 50 percent. Any remaining financing gap will be closed by a mixture of less concessional external borrowing and domestic borrowing. 3.2.3 Baseline results 10. Even with strong growth in the economy and commensurate growth in tax revenues, the Government will continue to struggle to create fiscal space and reduce debt from its currently high level (Table 3.2). Burundi is projected to run a deficit throughout the period, expanding from 1.7 percent of GDP in 2012 to 3.8 percent in 2014 and then moderating to 3.7 percent in 2015–2017. The dynamics are explained by the falling grants as a share of GDP and rigidities in current expenditure, mostly the wage bill. The combination of the two, in particular, expands the fiscal deficit in 2014. Although the government is expected to limit wage growth in real terms, unless additional revenue measures are introduced to offset the falling grants, it will struggle to reduce the deficits in the medium term. Gross financing needs will rise with the deficit, but are expected to be financed mostly with external concessional loans. Interest costs as a share of GDP will gradually fall, but with widening deficits, debt will remain high as a share of GDP. 11. The results suggest the authorities should persist with prudent fiscal policy as there is not much room to maneuver. Slippages in either revenue or expenditure targets will further worsen the projected fiscal deficits. Room for additional borrowing is limited without causing macroeconomic sustainability: the country is already rated as being at high risk of debt distress by the latest Joint World Bank-IMF Debt Sustainability Analysis (DSA), and further domestic 47 That is, both current and capital expenditures are growing but not as fast as the nominal GDP. 53 borrowing will worsen the inflationary pressure, increase exposure of domestic banks to the public sector, and risk crowding out private sector borrowing. Table 3.2 Medium-Term Fiscal Projection: Baseline Scenario (in percent of GDP unless otherwise indicated) Prelim. Proj. 2012 2013 2014 2015 2016 2017 Revenue and Grants 33.0 29.4 26.6 25.9 25.8 24.2 Revenue 14.8 14.8 14.8 14.8 14.8 14.7 Tax revenue 13.8 13.8 13.8 13.8 13.8 13.8 (1) Nontax Revenue 1.0 1.0 1.0 1.0 1.0 0.9 Grants 18.2 14.6 11.8 11.1 11.0 9.4 Total expenditure 34.6 31.0 30.4 29.5 29.4 27.7 Current expenditure 22.2 19.9 19.7 18.7 18.6 17.5 Wages and salaries 8.0 7.3 7.5 7.4 7.3 7.0 Interest 0.9 0.8 0.8 0.7 0.7 0.6 Domestic interest 0.8 0.7 0.6 0.6 0.6 0.5 External interest 0.1 0.1 0.1 0.1 0.1 0.1 Other 13.3 11.7 11.5 10.6 10.6 9.9 Capital Expenditure 12.4 11.2 10.7 10.8 10.9 10.2 Overall surplus/ deficit -1.7 -1.6 -3.8 -3.6 -3.7 -3.5 Primary surplus/ deficit -0.7 -0.7 -3.1 -2.9 -3.0 -2.9 Financing 1.7 1.6 3.8 3.6 3.7 3.5 External borrowing, net 0.4 1.5 3.6 3.2 3.8 3.1 New borrowing 0.9 2.2 4.5 4.5 5.1 4.4 Amortization (- entry) -0.5 -0.8 -0.9 -1.3 -1.3 -1.3 Domestic borrowing, net 0.8 -0.1 0.2 0.4 -0.2 0.4 New borrowing 1.3 0.1 0.7 0.8 1.0 1.0 Amortization (- entry) -0.6 -0.2 -0.4 -0.4 -1.2 -0.6 (2) Changes in financial assets -0.5 -0.2 0.0 0.0 0.0 0.0 Financing gap 0.0 0.0 0.0 0.0 0.0 0.0 Memorandum items: Nominal GDP (in BFu bn) 3,566 4,138 4,690 5,242 5,833 6,662 Government Debt (% GDP) 35.2 33.7 33.9 34.1 34.5 33.6 External debt 21.4 21.9 23.2 24.1 25.7 25.4 Domestic debt 13.8 11.8 10.7 10.0 8.8 8.1 Source: IMF (March 2013), DSA (January 2013), and the author's calculation. Note: (1) Includes sale of fixed capital assets. (2) Negative sign denotes a reduction of financial assets. 54 3.3 Fiscal Risk Analysis 12. The above analysis is deterministic and Figure 3.1: Changes in Growth Forecasts its assessment of fiscal sustainability is 6.5 predicated on a realization of given 6.0 macroeconomic and fiscal policy projections. Macroeconomic outcomes in Burundi tend, 5.5 however, to be very unpredictable because of the 5.0 country’s exposure to large shocks, both external and domestic. Forecasts are correspondingly 4.5 April 2013 WEO October 2012 WEO volatile (Figure 3.1). 4.0 April 2012 WEO September 2011 WEO April 2011 WEO 13. The sensitivity analysis in Table 3.3 3.5 shows how small changes in growth or the 2011 2012 2013 2014 2015 2016 2017 exchange rate would affect debt. If real GDP Source: IMF World Economic Outlook, April 2011– growth was one percentage point lower than we October 2012. have forecast each year up to the year ending in 2017, tax revenue would be around 1.4 percent of GDP lower each year as a result. Lower GDP and the additional cumulative deficit would cause debt to rise to 39 percent of GDP, instead of 34 percent projected under the baseline. Similarly, if the exchange rate depreciated one percentage point faster than we have forecast each year up to 2017, the government’s grant revenue would be higher while debt service on external debt would also be higher. Because of the offsetting benefits and costs, the impact on fiscal outcomes would be relatively muted, with the ratio of debt to GDP remaining at around the same level as baseline in case of faster depreciation. Table 3.3 Sensitivity of Burundi’s Debt to Small Changes in Macroeconomic Forecasts 2013 2014 2015 2016 2017 Debt to GDP under the baseline 33.7 33.9 34.1 34.5 33.6 Impact of change in real GDP growth rate Lower GDP growth (-1%) 34.2 35.2 36.4 38.1 38.8 Highe GDP growth (+1%) 33.3 33.1 32.9 32.9 31.8 Impact of change in Exchange rate depreciation Lower depreciation rate (-1%) 33.7 33.9 34.4 35.3 35.2 Higher depreciation rate (+1%) 33.8 33.9 34.1 34.5 33.7 Source: Authors’ calculations. Stochastic analysis 14. The structural vulnerabilities of Burundi’s economy to external shocks are well known, namely, volatilities in global coffee prices, rainfall, and aid flows. Each type of shock is defined as deviation from the central trend and is assumed to be normally distributed with a mean of zero and a standard deviation estimated by the variable’s historical data. Using Monte Carlo simulation, each shock is drawn randomly from its respective probability distribution, and is applied to the relevant endogenous variables to generate a set of alternative fiscal outcomes. 55 The process is repeated 10,000 times to produce a probability distribution of possible fiscal outcomes. Fan charts are used to illustrate the results. 48 15. What would be the impact of coffee price shocks on the country’s fiscal paths? Other things being equal, a negative coffee price shock will result in lower coffee export values, which will then result in lower government revenue and higher fiscal deficits. 49 The higher fiscal deficits cause a rise in gross financing needs, leading to higher debt. 50 Positive coffee price shocks would have the opposite impact on Burundi’s fiscal prospects. 16. Stochastic simulation suggests that, under very adverse shocks in coffee prices, the Government’s debt could reach 41 percent of GDP by 2017 (Figure 3.2). The fan chart shows percentiles of the probability distribution of public debt, centered on the baseline projection. In 2017, for instance, the seventy-fifth percentile value corresponds to a ratio of public debt to GDP of 36 percent, which means that there is a 75 percent chance that public debt is 36 percent of GDP or lower (or a 25 percent chance that it would be higher). Put another way, there is a 25 48 The analysis is a type of sensitivity analysis—the shocks are applied on a ceteris paribus basis, i.e., while one shock (e.g., to the coffee price) is being applied, other shocks (e.g., to rainfall) are not taking place. In addition, only the direct impact of the shock on the endogenous variables is considered, not indirect effects, including “feedback” and policy responses. For example, while applying shocks to coffee prices, the direct effect is on the values of exports, and thereby GDP and government revenues. No secondary effects or feedback are considered, such as possible adjustments in imports, interest rates, or exchange rates. We do not estimate the combined effect of all three types of shocks. Usually such a combined effect is estimated simply by adding up the shocks (e.g., the DSA uses this approach). This is often misleading because (i) some shocks mitigate each other; (ii) there will be secondary adjustments (e.g., interest rates, exchange rates); (iii) there will be policy response by the Government; and (iv) there will be donor responses to shocks. This is why some studies estimate the impacts of combined shocks using the variance-covariance matrix estimated via VAR (Vector Auto Regression), which will summarize, in reduced form, the historical patterns of the combined responses from (i)–(iv). For Burundi, we cannot implement this approach due to lack of long-term data. VAR usually takes a data series of 20 years or longer to be meaningful (many countries try to increase the sample by using the quarterly data). With VAR dummy running around 10 years (which causes structural breaks), Burundi's covariance matrix cannot be estimated robustly. 49 The impact of coffee price shocks on government revenue is econometrically estimated using a simple reduced form time series regression model (see Appendix 1). Based on the world price of Arabica coffee in U.S. cents per kg (1980–2012), a shock to the coffee price is defined as the change in the logarithm of the price at time t relative to t - 1, and is assumed to be normally distributed, with mean zero and the standard deviation estimated by the historical data (0.37). We found export volume of coffee in Burundi to be completely inelastic to changes in the world price— perhaps due to the limited correlation between the world price and the producer prices in the country (see Kimonyo and Ntiranyibagira, 2007). Because coffee exports can contribute to the central government’s revenue in multiple ways (e.g., through trade tax, income tax, revenue from SOEs, etc.), we have estimated the impact of coffee price shocks directly on the government revenue instead of inferring it indirectly from estimating the impact first on exports or on GDP. We abstract from other policy responses like donor support, fiscal adjustment, and exchange rate depreciation. 50 The projected fiscal deficit under the baseline is expected to be fully financed by projected grants and concessional borrowing (see the absence of financing gaps in Table 3.2). Financing gaps that could emerge as a result of shocks are assumed to be closed by a mixture of less concessional external borrowing and domestic borrowing (see paragraph 9). However, it has been suggested that in reality financing gaps would likely be met by a mixture of (i) cuts in capital spending; (ii) building up of arrears; (iii) borrowing from the Central Bank; (iv) borrowing from the domestic banking system, as well as borrowing from less concessional external sources (e.g., short-term bank loans). Such financing strategy would have knock on effects on growth, inflation, exchange rates, as well as the vulnerability of the banking system. However, as explained above in the sensitivity analysis such secondary effects or feedback are not considered. 56 percent probability that coffee price shock in 2013–17 will raise the public debt to a GDP ratio of 2 percentage points over and above the 34 percent level expected under the baseline scenario.51 Figure 3.2: Fiscal indicators under Stochastic Coffee Price Shocks Total Debt (%GDP) Gross Financing Needs (%GDP) Source: Authors’ own calculations. Figure 3.3: Burundi Annual Rainfall and 17. Rainfall shocks are a central Agriculture Growth Rates concern in a country that is mainly 1,700 mm Annual precipitation, 50% rural, with agriculture dominating as a millimeter Agriculture GDP growth means of livelihood for a majority of its 1,600 (RHS) 40% population. Around 90 percent of the 1,500 30% labor force is said to be employed in 1,400 20% agriculture. With the exception of 1,300 commercial tea and coffee production— 10% which contribute less than five percent of 1,200 0% GDP—agriculture remains largely 1,100 subsistence, rain-fed farming. Rainfall in 1,000 -10% Burundi is highly volatile, and droughts 900 -20% are recurrent events (Figure 3.3). Higher 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 precipitation clearly contributes to faster growth in agriculture. Importantly, growth in agriculture is said to have a large Source: IGEEBU. spillover to the rest of the economy: recently, the IMF has estimated that a one percent increase in agricultural sector activity leads indirectly to 0.7 percent increase in real GDP growth in other 51 The probability distribution of fiscal outcomes depends on the assumed distribution of coffee price shocks, which we assume to be normally distributed. There is, however, a large empirical literature that shows that distribution of commodity price shocks are generally not normal but “fat tailed,” which implies that extreme shocks happen more frequently than would be the case under the normal distribution. For example, under the assumption of normal distribution, our result suggests that the “extreme outcome” of government debt exceeding 41 percent of GDP would occur with less than one percent probability. If instead the shocks are distributed with “fat tails,” then such extreme outcomes would occur with greater than one percent probability. 57 sectors.52 Thus, weather shocks can have a large impact on Burundi’s economic performance, and hence fiscal outcomes. 18. Other things being equal, a negative rainfall shock will lower growth in agriculture, and through the direct and indirect effects, lower the overall growth.53 Lower growth leads to lower government revenues, higher fiscal deficits, and higher gross financing needs, resulting in higher public debt. Positive weather shocks will have the opposite impact. 19. Stochastic simulation results - reported in Figure 3.4 - suggest that weather shocks pose a smaller risk to fiscal outcomes than shocks to world coffee prices. The smaller revenue impact of rainfall shocks might be explained by the well-known difficulty of collecting taxes from the mainly rural, subsistent agricultural sector.54 In the event of very severe weather shocks, the Government’s debt could reach 39 percent of GDP by 2017. With 25 percent probability, weather shocks in 2013–17 could raise the public debt to 36 percent of GDP; with 10 percent chance, it could reach 37 percent of GDP. Figure 3.4: Fiscal Indicators under Stochastic Rainfall Shocks Total Debt (%GDP) Gross Financing Needs (%GDP) Source: Authors’ own calculations. 20. Volatility of aid represents another area of potential shock. In 2012, grants represented 18 percent of the country’s GDP and 55 percent of total Government income. In 2008–12, they covered on average 85 percent of overall fiscal deficit excluding grants. Aid flows 52 IMF (2006). 53 The response of agricultural GDP to rainfall shocks is estimated using a simple reduced form time series regression (see Appendix 1). A shock to rainfall is defined as the logarithm of rainfall in time t relative to its long- run average, and is assumed to be normally distributed with mean zero and the standard deviation estimated by the historical data (0.27). The estimated response of agricultural GDP to rainfall shocks is then used to estimate their impact on GDP using a multiplier of 0.5—i.e., we assume that a one percent increase in agricultural GDP will lead to 0.5 percent increase in GDP—somewhat higher than the share of agriculture in GDP (0.34) but lower than the total elasticity implied by the IMF’s (share of agriculture in GDP + 0.71). As before, we abstract from policy responses such as potential increase in external support in the event of severe drought and fiscal policy responses (e.g., changes in tax exemptions, subsidies). 54 It is also possible that the multiplier effect of agricultural GDP is much larger than assumed here (0.5), but substantially increasing the assumed size of the multiplier would unlikely make the impact bigger than that of the coffee price shock. See Appendix 1. 58 are generally subject to significant delays and are less predictable than other sources of revenue (World Bank 2012). Heavy reliance on this volatile source of income has been identified as an important risk to Burundi’s fiscal sustainability in the most recent Country Assistance Strategy and in the previous Public Expenditure Review. And the risk is unlikely to fall in the near future. In 2012, donors pledged US$2.5 billion in additional aid to support the country’s PRSP-II— more than double the originally anticipated amount (US$1 billion).55 Most of the aid will be in the form of grants and highly concessional loans, given the country’s high risk rating in the latest IMF-World Bank Debt Sustainability Analysis. Most donors reiterated their determination to tie continued support for the government to tangible improvements in the identified areas of reform including governance, corruption, democratization, and private sector development. Burundi has experienced sharp drops in donor disbursement before in response to deteriorating political situations or governance and fiduciary concerns. 21. Other things being equal, grant shocks will reduce the government’s income on a one-to-one basis. 56 Delay in disbursement (negative shock) in any given year will reduce government revenue below the baseline projection, widen fiscal deficits, and increase gross financing needs. The deficit will be financed by new borrowing resulting in higher debt. A positive grant shock—disbursement over and above that projected in the baseline—will have the opposite effect on fiscal outcomes. 22. Stochastic simulation suggests that, under very adverse shocks, the Government’s debt could reach 61 percent of GDP by 2017 (Figure 3.5). In the fan chart, the ninetieth percentile corresponds to a ratio of debt to GDP of 50 percent in 2017, implying that there is a 10 percent chance that the country’s debt will exceed that level. The seventy-fifth percentile corresponds to a ratio of debt to GDP of 43 percent in 2017, implying that there is a 25 percent chance that the country’s debt will exceed that level. The results must be interpreted with caution. Apart from the small sample (short time series), the historical volatility in grants may, on one hand, underestimate the true fiscal impact because it does not capture delays in aid disbursement within year, but on the other hand, overestimate the true fiscal impact because it reflects not only exogenous shocks (donor behavior) but also endogenous shocks (the Government’s behavior, e.g., implementation of agreed reforms). 55 The World Bank accounts for nearly one-quarter of the total but other donors such as the African Development Bank, Belgium, the Netherlands, and Germany also remain important players (EIU 2013). Detailed negotiations are expected in coming months. 56 The impact of grant shocks on government revenue is assumed to be one-to-one, i.e., a $100 reduction in grants in time t will reduce government revenue by $100 times the FBU/USD exchange rate in time t. Based on the data (in U.S. dollars) in 2008–17, a shock to grants is defined as the standard deviation of grants as a percent of GDP in time t, and is assumed to be normally distributed with mean zero and a standard deviation estimated by the data (0.06). Grants before 2008 are an order of magnitude smaller than grants thereafter, and using a longer time series would have exaggerated the estimated volatility in grants. The projected exchange rate in the baseline is used to convert grant shocks to the local currency equivalent and as percent of total revenue. As before, we abstract from other possible changes like fiscal adjustment and exchange rate depreciation. 59 Figure 3.5: Fiscal Indicators under Stochastic Aid (Grant) Shocks Total Debt (% GDP) Gross Financing Needs (% GDP) Note: The fiftieth percentile outcomes are not centered in the distributions because (a) the gross financing needs are truncated to values greater than equal to zero, and (b) there is no assumption that the fiscal surplus will be used to pay down existing debt faster than it is scheduled in the baseline. Source: Authors’ own calculations. 3.4 Policy Scenario Analysis 23. To mitigate the risks shown in the analysis above, the authorities should develop buffers. In this connection, some of the reforms that have been initiated in the last few years are promising. However, endogenous risks of partial policy implementation, policy inertia, and inability to control short run expenditure pressures to balance against long run development objectives are some of the risks that have plagued the fiscal situation in Burundi in the past, and are the risks that need to be monitored closely over the political cycle in the country. This section will consider the specific revenue and expenditure policy reforms that are currently being considered by the authorities and examine their potential medium-term impact on fiscal outcomes. 3.4.1 Scenario 1. Partial policy implementation 24. The first policy scenario examines the risk arising from partial implementation of revenue reforms. Domestic revenue mobilization remains a top priority for the authorities’ fiscal reform agenda. While good progress has been made, sustained improvement in domestic revenue mobilization has eluded the authorities in the past (Figure 3.6). Two pending issues have been recently highlighted in the policy dialogues with the Bank and Fund. The first relates to ending the fuel subsidy (exemption from excise taxes), currently costing an estimate of about 3/4 percent of GDP. The authority has not agreed on the definite timing of the phase out.57 The second relates to the revision of the income tax code (corporate and individual). In the course of discussion in Parliament, new exemptions were introduced in the revised bill, which would reduce the tax revenue by an estimate of 3/4 percent of GDP compared to the baseline.58 57 See IMF (March 2013). 58 See Chapter one for a discussion of these issues. 60 Figure 3.6: Government Revenue excluding grants (percent of GDP) Source: IMF. 25. The simulation results suggest that the government’s failure to adhere to the revenue reform would pose a significant fiscal risk. Results are reported in Figure 3.7 under the “lower tax revenue” scenario. Compared to the baseline, annual revenue shortfalls resulting from the incomplete revenue mobilization effort would have the same detrimental impact on the fiscal position as adverse GDP growth shock by 1.5 percent each year. Foregone revenue rises as GDP grows over the projection period, costing FBu62 billion a year in 2013 to FBu100 billion a year 2017. Fiscal balance will widen from 1.7 of GDP in 2012 to 5.1 percent of GDP by 2017. With higher gross financing needs each year, debt to GDP will reach nearly 40 percent by 2017, over six percentage points above the baseline. Thus, with expected decline in external grants, future fiscal sustainability will critically hinge on the government’s ability to stick to the revenue mobilization efforts. Figure 3.7 Comparisons of Alternative Reform Scenarios: Baseline vs. Scenario 1 & 2 a) Revenue excluding grants (FBu billion) 1,200 1,100 1,000 900 800 700 600 500 2012 2013 2014 2015 2016 2017 Baseline Lower tax revenue Higher tax revenue Source: Authors’ own calculation. 61 b) Fiscal balance to GDP (% GDP) 0.0 2012 2013 2014 2015 2016 2017 -1.0 -2.0 -3.0 -4.0 -5.0 -6.0 Baseline Lower tax revenue Higher tax revenue Source: Authors’ own calculation. c) Debt to GDP (% GDP) 41 39 37 35 33 31 29 27 25 2012 2013 2014 2015 2016 2017 Baseline Lower tax revenue Higher tax revenue Source: Authors’ own calculation. Note: “Low tax revenue” and “high tax revenue” scenarios are explained in detail in paragraphs 24–25 and 26–27, respectively. 3.4.2 Scenario 2. Reduction of tax incentives toward the regional average 26. The second alternative scenario explores a strategy in which the Government boosts revenue collection by broadening the revenue base primarily by reducing tax exemptions. Burundi has a large potential for boosting government revenues by rationalizing its extensive tax incentives. According to OBR official data, the forgone revenue due to tax incentives was around FBu 106.3 billion (US$ 70 million) in 2012. This is equivalent to about 3 % of GDP or about one-fifth of the total government revenues. Compared to the neighboring countries, the figures are high, and since the exemptions are mostly given exclusively to domestic firms, it is said to be not effective for attracting additional investments in the country. With a move to harmonize tax incentives high on the EAC regional integration agenda, rationalizing tax incentives thus has the potential of not just boosting government revenue but helping toward improving the investment climate and accelerating regional integration. 62 27. The simulation results suggest that the government’s accelerated effort in this area could have a big impact on fiscal outcomes. Results are reported in Figure 3.10 under “higher tax revenue” scenario. We assume the authorities will gradually reduce tax exemptions by 2017. Reducing tax incentives would result in additional revenues of FBu 17.9 billion in 2013 gradually rising to FBu 127.4 billion in 2017 above those projected in the baseline. Revenue excluding grants will rise from 14.8 percent to 17.6 percent of GDP. Fiscal balance will begin to improve after 2014, approaching near balance by 2017. The government could build up a fiscal buffer while lowering the level of public debt, potentially creating additional borrowing space for priority investments. With lower gross financing needs each year, the country’s debt to GDP will fall from 35.2 percent in 2012 to 25.8 percent in 2017, that is, 7.7 percentage points below the projected debt to GDP in the baseline (33.6 percent). 3.4.3 Scenario 3. Options for civil service pay reform 28. The third alternative scenario considers the implications of the reform of civil service wages currently under discussion. Wages are the biggest component of Government expenses (36 percent) costing over half of its own revenues. They cause a significant rigidity in Government spending and exacerbate the economic impact of external shocks. In case of negative revenue shocks, spending cuts tend to be borne by other expenditures while wages tend to be protected and are sometimes allowed growing. 29. The Government, with the help of UNDP, has been developing a plan to reform civil service pay. Under the current proposal, four policy options are identified (Table 3.4), all of which call for a revised pay structure for 23 pay grades, while eliminating disparities within each pay grade.59 Table 3.4 Salary Scale Adjustment Options Options Principle Estimated one-off cost (FBu billion) 1. Maximum salary Salaries within each grade to be raised to the highest 273 adjustment observed for the grade 2. Average salary Salaries of those below the current average to be raised to 17 adjustment the average for the grade 3. Reference salary Salaries to be adjusted to the reference wage for each 87 adjustment grade proposed by the union 4. Combination of Salaries to be adjusted to levels that lie somewhere 18 2&3 between the average (Option 2) and the reference salaries (Option 3) Source: Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). Note: Estimated costs associated with these options are one-off costs of adjusting the current salaries to the respective reference salaries under each option. See Government of Burundi, Ministry of Public Service, Work, and Social Security (2012) for more detail. 30. The reform proposal aims primarily to reduce inequality within the pay structure and not to control short-run growth in the wage bill. All options currently under discussion are going to increase wage bill in the short run, but are expected to generate savings and additional tax revenues in the long run. For example, the rationalization of the pay scale is 59 See Chapter four for detail. 63 expected to bring about more orderly management of wage bill eventually resulting in lower shares in terms of revenue, expenditure, and GDP. 31. Dynamic implication must be carefully studied once the details have been worked out. Apart from the initial costs of adjustment, the different wage structures would have different dynamic implications for the medium-term fiscal outcomes. The dynamic costs will depend on, inter alia, growth of new hiring in the priority sectors, their distributions across the pay scales, new rules regarding bonuses and allowances, as well as expected efficiency gains from the reforms. Below, we attempt a very preliminary estimate of the potential fiscal implications, based on many simplifying assumptions. 60 The analysis must be interpreted with caution, however. In addition to making a number of simplifying assumptions, the analysis does not attempt to incorporate potential efficiency gains from the reform 61 —which are highly uncertain—in estimating the fiscal impact of the different options for the reform. 32. The simulations confirm that Options 1 and 3 are the most expensive, even in the medium term (Figure 3.8). In the first three to five years, the dynamics of the wage bill is dominated by the one-off adjustment costs. Growth of the wages bill would be substantially higher under Option 1 and Option 3 even when the costs of adjustments are evenly spread—over three years in Option 1 and 2 years in Option 3. After the adjustment to the new salary scale is completed, the growth of the wage bill would converge to the steady state growth rate as in other options, which is driven by the growth in civil servants in the priority sector (10 percent of the work force in that sector each year) and the automatic annual salary increase (six percent each year). The wage bill would be permanently higher because the higher wage growth in the first few years would not be compensated by lower growth in the subsequent years. As a result, wage rigidities would worsen, the fiscal deficit would significantly expand, and debt would reach a possibly unsustainable level. More specifically, as a share of GDP, the wage bill would rise from eight percent to 15 percent (under Option 1) and 11 percent (under Option 3), instead of falling to seven percent by 2017 as expected under the baseline. The fiscal deficit would expand to 13 percent (under Option 1) and nine percent of GDP (under Option 3). Debt as a share of GDP would reach 63 percent (under Option 1) and 50 percent (under Option 3). 60 Specifically, we assume that (i) the number of civil service employees will grow by 10 percent per year in the priority sectors (education and health), and 0 percent per year in other sectors; (ii) distribution across the civil servants will remain the same even after the new hiring; (iii) after the initial salary adjustment, the automatic annual salary increase of six percent currently in place is maintained. These are common assumptions across all options. In addition, under Option 1 and Option 3 (the costliest options), we assume the costs of adjustments are evenly spread over three years and two years, respectively. For example, under Option 1, a third of the existing employees are brought under the new salary scale in the first year (t+1), another third in the second year (t+2), and all existing employees in the third year (t+3). Under Option 3, a half of the existing employees are brought under the new salary scale in the first year (t+1), and the remaining half in the second year (t+2). During the adjustment period, it is assumed that there will be no automatic adjustment of salaries but newly hired employees are paid according to the new salary scale. Note that under Option 2, the proposal is that only salaries of those below the initial average will be raised to the average for the grade, while salaries of those above the initial average are assumed to be frozen. See Appendix 2 for more detail. 61 The rationalization of pay scale is expected to bring about improved public sector performance—though better job satisfaction, worker motivation, and attraction and retention of qualified professionals—as well as more orderly management of wage bill—through better control of bonuses and allowances. See Government of Burundi (2012) for more detail. 64 33. Option 2, on the other hand, implies initially higher growth in wages, followed by lower growth, thus neutralizing the impact on fiscal outcomes relative to the baseline. During the adjustment—in which salaries of those below the initial average for each grade will be raised to the average—growth of the wage bill would be higher than the baseline, but in the subsequent years, the growth of the wage bill will be lower because the salaries of those above the initial average for each grade (approximately a half of the initial workforce) are frozen. Only those paid at the reference salaries would have their salaries increased at six percent a year. If the reform under this option can reduce the wage inequality as desired, while keeping the fiscal costs in line with the baseline, it would be an attractive option. However, under the given assumptions wage inequalities would persist for a long time, and freezing the salaries of roughly half the (initially better paid) work force for the time it takes to reduce the inequality may be unrealistic.62 34. Option 4, which is a compromise option between Option 2 and Option 3, not surprisingly, provides an intermediate outcome. With the relatively low adjustment cost, growth in the wage bill would quickly stabilize at the steady state growth rate driven by the rates of new hiring in the priority sector and the automatic annual salary adjustment. The nominal wage bill would only be slightly higher than the baseline projection, with muted impacts on fiscal deficit and debt. The challenge, therefore, would be for the government to strike the right balance between setting the new reference salaries sufficiently high to achieve the desired long- term objectives (reducing inequality within the pay grade, providing incentives for better performance, eliminating the majority of bonuses and allowances as substitute for orderly salary increase, etc.) and keeping the short run adjustment costs sufficiently low to avoid a fiscal blowout and risking macroeconomic instability. 62 See Chapter four for more in-depth analysis of the pay policy. 65 Figure 3.8: Comparisons of Alternative Reform Scenarios: Baseline vs. Options 1 -4 a) Wages and salaries b) Fiscal Balance (% GDP) c) Debt to GDP (%) Source: Authors’ own calculations. 66 Chapter 4 : Public Expenditure Analysis Inconsistencies between data sources seriously affect the analysis. Public expenditure/GDP percent is comparatively modest but current spending increases are driving it up. Increased efficiency is essential for PRSP-II delivery. Salary expenditure is increasingly large and highly executed. Over executions are prevalent except for capital spending. Share of budget for priority areas has gradually increased as has execution. Much better information on external financing is needed. MDG goals are elusive despite pro-poor spending. The quality of MTEF has deteriorated over time. Its suitability in the Burundian context is open to question. 4.1 Data Sources and Data Quality Issues 1. Burundi’s public expenditure management while computerized and – in principle – capturing budget data at every stage of the expenditure chains, in practice, seriously undermined by structural and systemic problems, which in turn limit the possibility to undertake detailed and sophisticated analysis.63 Each Loi de Règlement forwarded to parliament since 2008 present two sets of figures that are inconsistent. The first set of numbers is extracted from the Treasury balances and is presented in TOFE format. The second set of numbers, which presents the numbers by administrative classification, is taken from SIGEFI and appears under the title “Annexe II Classification administrative de l’exécution des dépenses.” Table 4.1 below shows the extent of the discrepancies for 2009, and similar tables for 2010 and 2011 can be found in the annex. Table 4.1 Expenditure chain 2009 2009 Loi de Collectif SIGEFI Loi de Execution ratios Financ budgétai Reglement es re Credit Engageme Liquidatio Ordonna Paiemen Ordonnan Paiement/ Loi de nt n ncement t cement/L LFI Reglement/L FI FI Current 387.0 422.0 423.0 417.0 406.0 389.0 403.0 420.0 100.0 104.0 108.0 Expenditures Personnel 163.0 186.0 186.0 185.1 179.8 178.0 177.0 183.0 109.0 108.0 112.0 Goods and 97.0 102.3 103.5 104.0 103.8 99.0 99.0 107.2 102.0 101.0 109.0 Services Interests 33.0 32.0 33.0 28.0 22.0 11.0 26.8 28.0 34.0 79.0 83.0 Transfers 93.0 100.6 100.7 100.0 100.2 99.0 99.0 101.5 107.0 107.0 109.0 Imprévus 1.0 1.0 63 Budget approval, budget release, expenditure commitment (engagement), invoicing (liquidation), payment order (ordonnancement), and payment (paiement). 67 Capital 93.0 81.0 81.0 58.0 56.0 54.0 54.0 59.0 58.0 57.0 63.0 Expenditures Total (Curr. 482.0 505.0 505.0 475.0 463.0 444.0 457.0 480.0 159.0 161.0 171.0 Exp. + Imprévus + Cap. Exp.) Sources: Loi de Finances Initiale, Collectif Budgétaire, Loi de Règlement 2009, and SIGEFI. The data from SIGEFI was cleaned and ordered according to the BOOST methodology. 2. The inconsistencies among budget data sources are significant, and substantially limit the reliability of the results and implications emerging from the analysis. The data from the Treasury balances show budgetary over executions, and hence violations to the expenditure chain, in all the economic categories of the budget with the exception of capital expenditures. SIGEFI, on the contrary, shows under executions for most ministries, including the priority sectors. As the data taken from the Treasury balances is, a priori, a more accurate reflection of the liquidity situation of the state, these inconsistencies question the reliability of SIGEFI as a reporting system. These problems have already been already highlighted in the 2012 PEFA report (cf. pp. 43-46), 64 which describes this situation as “a major accounting and budgeting problem.” 65 Overall, these discrepancies limit our ability to make inferences and provide robust policy recommendations. At the same time, they call for urgent improvements in the accounting and budgeting procedures, including the full integration of SIGEFI with the Treasury. 3. An additional limitation of SIGEFI is that the system cannot retrieve the allocations voted by the Parliament once a wire transfer is done. This is also a tremendous limitation, as the only comparisons between budget and executions can be made at the Ministry level. In light of these limitations, whenever a table is presented, the text will indicate whether the data comes from the treasury balances or from SIGEFI. In the case where an execution ratio is estimated, the text will clarify whether the ratio is estimated using the original budget, or credits after wire. Annex F provides additional information on SIGEFI and its limitations. 4. Data on donor-financed public expenditures are poorly captured in Government systems. As is the case for many developing countries, most donor-funded expenditures are not implemented following Government’s standard budgetary procedures and are not fully integrated in the public expenditure management system. Information on donor funded expenditures is thus often incomplete and inaccurate, reflecting both weak systems to capture such information, but equally important, low priority accorded by many donor agencies to the provision of high quality information on planned project disbursements in a timely manner. Moreover, most of the donor funded expenditures also do not follow country systems for financial management and procurement. This implies that only a small part of foreign financed activities are captured in the Treasury system. As a result, the figures on donor- funded expenditures included in the Budget Law and in the TOFE presented in the Loi de Règlement are inconsistent with the data provided by the CNCA, which is the agency in charge of registering the projects with the donors. 64 Indicators PI-24 and PI-25 also mention the important differences between the Loi de Finances TOFE and the annex showing the information retrieved at the Ministry level. 65 According to the 2012 PEFA Les évaluateurs émettent cependant les plus expresses réserves sur la portée des résultats de cette analyse’. 68 Box 6 Major Sources of Discrepancies between SIGEFI and TOFE: the view from the Government The budget figures included in the expenditure chain do not match those presented in the TOFE presented in the Loi de Règlement. Since data taken from the Treasury balances is, a priori, a more accurate reflection of the liquidity situation of the state, these inconsistencies raise serious concerns on the reliability of SIGEFI as a budget reporting system. To better illustrate the magnitude of the problem, table F.1 in Annex F shows the differences between the figures approved in the Budget Law (the Loi de Règlement) and the steps of the expenditure chain registered in SIGEFI for the period between 2009 and 2011.66 As par to the PER preparation process, the Task Team in collaboration with the Technical Working Group organized several rounds of consultations with officials from the Ministry of Finance to better understand the reasons behind the discrepancies between these two budget account sources. The consultations helped to identify both systemic problems and challenges specific to a given budget year. The section below describes four major problems, while Annex F provides additional information on year specific challenges: 1. Restes à payer. This was mentioned as the main source of discrepancy between SIGEFI and the Treasury balances by officials from the Ministry of Finance. Given low institutional and implementation capacity, some payments are pending at the end of the year. According to the Burundi regulation, these payments have to be processed within two months of the following year, or they become arrears. If, for some reason, a payment is not processed within that timeframe but is paid afterwards, it will not be possible to identify it with a budget line of SIGEFI, though it will obviously be registered in the Treasury balance. Data from the Treasury balances presented in the Lois de Règlement 2008-2011 show that pending payments are becoming increasingly relevant. At the beginning of 2009, the Burundi treasury had payments pending for FBu 5 billion. This figure passed to FBu 57 billion in 2012. 2. Unforeseen expenditures and intrabudget wires after the approval of the budget. The Burundi Budget laws have a line allocated for “unforeseen expenditures” (i.e., imprévus). Additionally, the Burundi legislation allows the Ministry of Finance to do intrabudget wires. Both the imprévus line and intrabudget reallocations can be used by the Ministry of Finance on a discretionary and as-needed basis. The existence of these two mechanisms give room for confusion at the moment of doing the reconciliation between the data extracted from the Treasury balances and the data presented by SIGEFI. 3. Attestations de prise en consignation. Public officials from the OTBu mentioned the following issue as a potential source of discrepancy: for payments done through the banking system, bank officers sometimes register the account number of the beneficiary incorrectly and the payment is not made effectively. Once the error is assessed and the bank sends the money back to the Government’s account, the Treasury reprocesses the 66 The team decided to show 2011 because there were no revisions to the budget that year. However, the data shows the same patterns for 2009 and 2010 (see Annex F). 69 payment starting from the ordonnancement step even if, from a good accounting perspective, it would be necessary to cancel the entire operation from engagement and start over. The counterparts from the OTBu asserted that the operation is effectively cancelled in the Treasury balances and the problem stays at the level of SIGEFI, which registers the payment twice from ordonnancement onwards. Though this is clearly a tremendously big PFM problem, the extent to which it explains the discrepancies between the data provided by SIGEFI and the numbers extracted Treasury balances has to be limited, as the data from the Treasury is usually higher than the data that comes from SIGEFI. 4. Débits d’office and procedures dérogatoires (special procedures). A ministerial ordonnance issued in January 2009 specifies what payments can be done without following the regular procedures. The last PEFA (2012, indicator PI-20) claims that these procedures were suppressed with the exception of international payments linked with missions that are done via SWIFT and are presumably regularized immediately afterwards. Given the lack of clarity regarding the existence and utilization of special procedures, this cannot be definitely ruled out as a source of discrepancies between the treasury balances, which truly reflect the liquidity situation of the Government, and SIGEFI. 4.2 Domestic Financing 5. Over the past decade, public expenditure have increased from about 24.9 percent of GDP in 2000 to about 45.4 percent of GDP in 2008, and declined again to 29.75 in 2011. Although the level of public expenditures remains relatively modest in comparison to other developing countries, this increase in public expenditures has mainly been driven by current expenditures (Figure 4.2). Figure 4.1: Total expenditures Figure 4.2: Composition of spending 50.0 50.0 40.0 40.0 30.0 30.0 20.0 20.0 10.0 10.0 0.0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 0.0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Current expenditure Capital expenditure Source: IMF data. Source: IMF. 6. To achieve the ambitious development target outlined in the PRSP-II, the Government is committed to increase the allocation of resources to priority sectors, scaling up the provision of public services. Given the conditions of enduring fiscal fragility and the vulnerability to shocks (chapter two), the realization of the PRSP-II agenda will require increases 70 in the efficiency of public expenditures, clear prioritization in a context of hard budget constraints, and sustained efforts to keep arrears under control. 4.2.1 Public Expenditures by Economic Classification 7. Current expenditure represented 78 Figure 4.3: Breakdown by expenditure percent of the Government budget (excluding category (actuals) donors’ project support) over the period 120 2009-2011 on average. Capital expenditures have increased in absolute and relative amounts 100 since 2009 (Figure 4.3). In the context of 80 declining budget support, this reflects 60 government efforts to devote additional 40 resources to productive projects (PER 2010). 20 8. After growing faster than the overall 0 budget envelope, salaries represented the 2009 2010 2011 largest single item of current expenditures budget in 2011. The budget67 increased by 65.7 Capital expenditures and percent in nominal terms between 2008 and net lending Recurrent Expenditure 2011. This increase is explained by two factors: first, an increase in different types of transfers, Source: IMF contributions, and subsidies; second, an increase in personnel expenditures across the Government between 2001 and 2011, which grew faster than GDP, but were expected to decrease slightly in 2012. After reaching the HIPC completion point, the Government has reduced its interest payments considerably since 2009. Goods and services have decreased sharply in the last years due to a big increase in transfers, which are partly driven by the fuel subsidy. The category “others” is mostly TAs provided by donors, though some nonrecurring domestic expenditures are included in that category, notably the 2010 national elections. Table 4.2 Burundi: Actual expenditures, 2001-2012 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Expenditure and net lending 21.11 19.91 26.69 30.22 29.05 29.41 39.80 43.73 40.22 47.05 43.64 36.39 Current expenditure 16.28 15.56 16.69 16.19 16.60 16.91 17.80 18.75 19.74 19.66 19.43 17.27 Salaries 5.50 5.97 6.33 5.82 6.01 7.17 7.77 8.29 8.96 8.61 8.69 7.96 Goods and services 6.07 5.02 5.57 5.32 5.44 4.87 4.82 4.99 4.86 4.02 3.30 2.78 Transfers and subsidies 2.18 2.14 1.86 2.62 2.50 3.01 3.19 3.75 4.69 4.07 4.71 4.47 Interest payments 2.54 2.44 2.93 2.43 2.65 1.86 2.04 1.52 1.23 0.60 0.86 0.93 Domestic 1.31 1.00 1.57 1.39 1.61 0.96 1.24 0.88 0.69 0.54 0.74 0.82 Foreign 1.22 1.43 1.36 1.04 1.04 0.90 0.80 0.63 0.53 0.06 0.12 0.10 Others 0.00 0.00 0.00 2.13 5.49 3.58 8.68 10.82 7.50 15.40 8.62 6.71 Capital expenditure 4.82 4.35 10.01 11.90 6.96 8.92 13.31 14.17 12.97 11.98 15.59 12.41 Domestic resources 2.54 0.83 3.12 3.60 1.61 2.69 1.97 2.43 2.73 2.03 3.52 2.78 67 Including payments to debt principal. 71 External resources 2.28 3.52 6.88 8.30 5.34 6.22 11.34 11.74 8.89 8.79 12.07 9.62 Project loans 1.47 1.81 2.67 4.65 3.71 3.08 1.69 2.01 0.77 1.48 1.10 0.61 Project grants 0.81 1.71 4.21 3.66 1.63 3.15 9.65 9.72 8.12 7.31 10.97 9.01 Net lending -0.53 -0.19 -0.23 -0.26 -0.12 -0.06 -0.02 -0.04 0.00 0.00 0.00 -0.03 Source: IMF. 9. In addition to representing a large share of Figure 4.4: Execution rate (LR/LFI) the budget, personnel expenditures tend to exceed 150 the ceilings fixed in the Loi de Finances. According to the data extracted from the treasury balances and 100 presented under TOFE format in the Lois de Règlement, personnel expenditures and employee 50 benefits, presented over-executions consistently between 2009 and 2011 (figure 4.4). Government 0 officials attribute the over-executions to the imprévus 2009 2010 2011 Recurrent Expenditure budget line, which allows to do intrabudgetary Capital expenditures and… reallocations on an as-needed basis in the middle of Source: LR/LFI. the budget year. This is only a partial explanation, however, as the over-executed amounts tend to be much higher than the figures reported under the “imprévus.” In general, though deviations from the original budget tend to be an indicator of a suboptimal budget preparation, in the case of Burundi over executions of personnel expenditures may be – at least in part – the result of poor HR management system (see chapter five). 10. The execution rate of capital projects is low, though it has improved recently. In line with other fragile states, Burundi presents low execution rates for capital projects. Low capacity Figure 4.5: Capital Expenditure by among public officials and liquidity constraints are source of financing the main reasons behind this. Government officials interviewed for this study also attribute the low 14 execution rates to the procurement process, deemed 12 to be long and cumbersome. Data provided by the IMF indicates that, while capital expenditures 10 Governm executed with the government’s own resources 8 ent's own averaged 2.5 percent of GDP between 2000 and 6 resources 2012, projects funded by donor passed from three to Donors' 4 financing nine percent of GDP reaching a plateau of 12 percent in 2012. This confirms similar trends from 2 previous period (2005-2008), as documented in the 0 2010 PER. It is less clear to what extent externally 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 funded projects are executed. The TOFE presented in the Loi de Règlement presents figures that are too Source: IMF. Data as percent of GDP. low to be realistic. The reason behind that is few projects pass through the Treasury accounts and are not registered in the books of the Government. On the other hand, the data provided by the CNCA tends to include projects that are not included in the budget at all or are registered with more conservative amounts. 72 Box 7 Measuring Budget Execution Rates: a technical note In order to measure the reliability of the budget, execution rates are estimated using the budget approved at the beginning of the year and using the ordonnancement step of the expenditure chain, or the data available in the Loi de Règlement. The purpose of using the original budget and ordonnancement is twofold: first, it allows seeing whether the original budget is a reliable planning tool; second, in principle, it allows seeing the extent to which the State does indeed fulfill its financial commitments. In practice, however, this is difficult to estimate, given the limitations of the SIGEFI mentioned above. The Burundi authorities estimate the execution rate as the ratio between engagement and credits after wire. While this ratio measures the efficacy of releasing credits, in the case of Burundi, where SIGEFI provides unreliable data, it tends to provide high execution rates. Annex F provides all the steps of the expenditure chain including the revised budgets of 2009 and 2012. 11. Over-executions68 are prevalent in the recurrent budget indicating violations of the expenditure chain and noncompliance with the budget ceilings set by Parliament (see Table 4.3). The average budget execution rate for the period 2009-2011 was 109.35 percent for recurrent expenditures and 73.09 percent for Government own-funded capital expenditures. This is the result of poor planning and violations of the expenditure chain. In the case of interest payments, additionally, over-executions are partly driven by exchange rate losses derived from the volatility of the Burundi Franc exchange rate. The fact that the original budget is systematically over-executed suggests (a) the poor quality of the budget as a planning document and (b) recurrent violations to the expenditure chain (credits are ceilings under the Burundi budget system).69 68 In order to measure the reliability of the budget, execution rates are estimated using the budget approved at the beginning of the year and using the ordonnancement step of the expenditure chain, or the data available in the Loi de Reglement TOFE. The purpose of using the original budget and ordonnancement is twofold: first, it allows seeing whether the original budget is a reliable planning tool; second, it allows seeing the extent to which the State does indeed fulfill its financial commitments; however, given the limitations of SIGEFI mentioned at the beginning of this chapter, this needs to be qualified. The Burundi authorities estimate the execution rate as the ratio between engagement and credits after wire. While this ratio measures the efficacy of releasing credits, it tends to provide high execution rates. In any case, Annex X provides all the steps of the expenditure chain including the revised budgets of 2009 and 2012. 69 This does not apply to 2009, when a revised budget increased the ceilings for most of the credits. However, overexecutions still represent a poor budget planning. 73 Table 4.3 TOFE: Budget Execution Rates, 2009-2011 2009 2010 2011 Loi de Loi de Executio Loi de Loi de Executio Loi de Loi de Executio Finance Regleme n rate (in Finance Regleme n rate (in Finance Regleme n rate (in s nt percent) s nt percent) s nt percent) Initiale Initiale Initiale Recurrent 419.85 483.27 115.10 507.75 541.79 106.70 573.02 608.92 106.26 Expenditure Wages 163.43 183.91 112.53 212.41 233.93 110.13 272.44 272.81 100.14 Goods and 97.62 107.23 109.85 98.00 105.72 107.88 98.08 99.20 101.14 services Cont. to Ios 3.60 6.54 181.47 6.25 10.90 174.58 10.14 18.03 177.74 Transfers 43.52 44.47 102.20 33.14 32.65 98.52 32.92 31.52 95.74 Subsidies 45.89 50.53 110.10 88.64 96.99 109.42 87.98 90.18 102.50 Unforeseen 1.42 0.00 0.00 8.12 0.00 0.00 5.00 0.00 0.00 Tax 20.00 44.60 223.02 20.00 15.31 76.55 20.00 36.83 184.13 exemptions Interest 33.70 37.41 110.99 30.23 36.78 121.68 35.58 45.65 128.31 payments Exchange 0.00 0.00 … 0.00 10.81 … 0.00 5.46 … rate loses External 12.60 11.46 90.93 12.47 10.54 84.50 11.37 16.96 149.14 Debt Internal 21.10 25.95 122.97 17.76 15.44 86.94 24.21 23.24 95.96 Debt Funds 10.67 8.57 80.35 10.95 9.49 86.65 10.88 14.71 135.17 Capital 406.28 93.94 23.12 355.31 153.29 43.14 453.15 212.87 46.98 expenditures and net lending Capital 407.28 95.06 23.34 356.31 157.27 44.14 454.15 212.87 46.87 expenditures Own 93.75 59.47 63.43 107.19 76.90 71.74 127.05 106.83 84.09 resources Project 76.79 16.60 21.61 59.99 42.91 71.52 44.89 25.84 57.56 loans Project 236.74 19.00 8.02 189.13 37.46 19.81 282.21 80.20 28.42 grants Pret net du -1.00 -1.12 111.71 -1.00 -3.98 397.61 -1.00 0.00 0.00 tresor Source: Loi de Finance and Loi de Règlement (TOFE) 2009-11. 4.2.2 Public expenditures by administrative/sector classification70 12. On a sectorial basis, the budget tends to be concentrated in ministries linked with basic education, health, and security. However, the lack of a functional classification (see Annex F), in addition to constant changes in the names and composition of the ministries, does not allow a clear cut analysis of what sectors the Government is allocating its money.71 However, between 2008 and 2011, the same four ministries concentrated 55 percent of the budget: The Ministry of Elementary and Secondary Education (under its different denominations), the Ministries of Defense and Security, and the Ministry of Health. 72 (Annex F provides further details on their share of total budget). 70 Sources of data: SIGEFI. The analysis and conclusions presented in this section should be considered with caution, in light of the data irregularities described at the beginning of the chapter. 71 A list with the names of the ministries by year is available in Annex F. 72 This analysis excludes debt payments. If debt payments were included, the Ministry of Finance would also be part of this list. 74 13. The share of the budget Table 4.4 Priority Sectors as Share of Total Budget given to the priority sectors has (1) increased gradually since 2009. The relative budget allocated to the 2009 2010 2011 2012 2013 Priority sectors 41.91 43.44 43.81 45.48 47.92 priority sectors has increased Health/AIDS 7.73 9.78 9.79 8.82 9.78 gradually driven mostly by Education 24.23 25.34 23.80 25.46 27.58 Health/AIDS and Education.73 The Agriculture 3.60 2.67 5.85 5.79 4.75 budget allocated to the ministries Energy and 2.54 2.56 2.21 3.25 4.15 of Agriculture and Health and Mining Mining has increased as well. At Public Works 3.82 3.10 2.16 2.17 1.67 the same time, resources allocated (3) to public works have decreased Security sectors 20.66 17.06 18.39 18.66 19.37 since 2009 despite of the fact that Defense 12.15 10.41 10.78 11.27 11.78 the Ministries of Public Works and Security 8.51 6.64 7.61 7.39 7.59 Transports were merged in 2012. The share of the budget allocated (1) Refers to the Loi de Finances Initiale and includes debt to the security sectors has payments. remained practically constant since 2009, taking up almost 20 percent (3) Includes transports in 2012 and 2013. Sources: Lois de Finances 2009-13 and authors’ own calculations. of the total budget. 14. Though low execution rates were prevalent in the priority sectors between 2009 and 2011, improvements have been reported in 2012. The ministries of Agriculture and Energy and Mining, both of which have received more resources in every budget law since 2009, were among the 5 ministries with lowest execution rates for the period between 2009 and 2012. In the case of Ministry of Agriculture, some improvements in the execution rate have been made in 2010 and 2011, but preliminary data for 2012 suggests much lower performance. The budget of the Health/AIDS sector was also systematically under-executed, with the exception of 2009, when a revision to the budget approved in the middle of the year increased the ceiling for this sector. In contrast, the education sector presents one of the highest execution rates. It should be mentioned that, in general, the ministries linked with the security sector (Defense and Security) present better execution rates than those of the priority sectors.74 Table 4.5 Execution Rates, 2009-11 (billions of FBu) 2009 2010 2011 2012 (prel.) LFI Ord. Exec. LFI Ord. Exec. LFI Ord. Exec. LFI Ord. Exec. rate rate rate rate (in (in (in (in %) %) %) %) Priority sectors 232.3 192.8 293.6 276.3 323.7 222.6 361.5 316.1 Health/AIDS 42.8 43.7 102.1 66.1 61.9 93.7 72.4 56.1 77.5 70.1 63.2 90.2 73 “Education” includes all the ministries in charge of primary, secondary, and tertiary education, as well as research. 74 If SIGEFI were integrated with the Treasury systems, this would mean, in practical terms, that the Government is devoting more resources to security than to the priority sectors. The data extracted from the Treasury balances indicates that overexecutions are prevalent across the Government, particularly personnel payments. As most expenditures of the priority ministries are related to personnel payments, this means, logically, that the data extracted from SIGEFI is wrong. The only conclusion that can be extracted from the table below is that the security sectors tend to register more operations in SIGEFI than the priority ministries. 75 Education 134.3 120.0 89.4 171.2 172.1 100.5 175.9 113.5 64.5 202.3 187.9 92.9 Agriculture 19.9 11.3 56.6 18.0 13.6 75.4 43.2 31.6 73.0 46.0 30.3 65.8 Energy and 14.1 4.6 32.4 17.3 14.0 81.0 16.3 8.2 50.3 25.8 19.3 74.6 Mining Public Works 21.2 13.3 62.7 20.9 14.7 70.1 16.0 13.3 83.8 17.2 15.4 89.3 Security sectors 114.5 114.3 115.3 107.6 135.9 123.3 148.3 136.6 Defense 67.3 66.8 99.2 70.4 65.3 92.8 79.7 74.5 93.6 89.6 85.7 95.7 Security 47.2 47.5 100.7 44.9 42.3 94.3 56.2 48.8 86.7 58.8 50.9 86.6 Source: Loi de Finance and Loi de Règlement (Annexe II Classification administrative de l’exécution des dépenses), 2009-2011. 15. Data provided by SIGEFI indicates that violations to the expenditure chain were prevalent, as several ministries over-executed their budget over the period 2010-2012. Under Burundi law, credits are ceilings, and the steps of the expenditure chain (chaine de la dépense) cannot be overruled. The Ministry of Finance has the prerogative to assign reallocations (virements) up to 10 percent of a given ministry without parliamentary approval. Once the ceiling is set, ministries cannot exceed it. In theory, modifications beyond 10 percent require a revision to the budget law approved by Parliament. In practice, several ministries over-executed their budget. The budget was revised in 2009, which explains the prevalence of over-executions with respect to the original budget law in that year (see Annex F). However, the budget was not revised in either 2010 or 2011 and many ministries presented violations to the expenditure chain (i.e., commitments or payments above the amounts authorized by Parliament). This was especially the case of the Ministry of Interior (260 percent execution rate in 2010), Ministry of Public Services (238 percent execution rate in 2012), and the Second Vice Presidency (163 percent execution rate in 2012) among others. Table 5.6 below shows the top five overspending ministries over the period 2010-2011, while table F2-F6 in Appendix F provide a comprehensive list of the whole expenditure chain, by administrative classification. This calls for a reinforcement and rationalization of the expenditure internal controls, as highlighted in the 2011 PEFA report. Table 4.6 Budget execution rates by Ministries (2010-2012) 2010 LFI Ordonnanc Executi ement on rate (in percent) Ministère de l'Intérieur 4.06 10.55 260.03 Ministère des Droits de la Personne Humaine et du Genre 0.38 0.65 172.14 Deuxième Vice-Présidence 1.07 1.29 120.58 Ministère des Relations Extérieures et de la Coopération Internationale 20.84 24.92 119.58 Ministère de la Fonction Publique, du Travail et de la Sécurité Sociale 4.66 5.55 118.93 2011 LFI Ordonnanc Executi ement on rate (in percent) Ministère des Affaires de la Communauté Est Africaine 1.59 2.40 151.44 Ministère des Relations Extérieures et de la Coopération Internationale 25.20 32.73 129.89 Présidence de la République (y compris le Secrétariat General du 15.44 17.44 112.94 Gouvernement et l'Institution Ombudsman) Deuxième Vice-Présidence 1.06 1.13 106.67 Parlement 12.17 12.29 101.00 76 2012 (prel.) LFI Executi Ordonnanc on rate ement (in percent) Ministère de la Fonction Publique, du Travail et de la Sécurité Sociale 2.65 6.30 238.12 Deuxième Vice-Présidence 1.06 1.72 162.91 Prémière Vice-Présidence 1.68 2.14 127.70 Ministère des Relations Extérieures et de la Coopération Internationale 19.65 22.07 112.31 Vice-ministère des Affaires de la Communauté Est Africaine 1.59 1.76 110.80 Sources: Loi de Finance and Loi de Règlement (Annexe II Classification administrative de l’exécution des dépenses), 2010-2011. 16. The growing level of expenditures paid without prior authorization (sans ordonnancement préalable) remains an issue of concern. Data extracted from the Treasury balances indicates that – excluding debt payments - payments without prior authorization (sans ordonnancement préalable) are becoming increasingly important, passing from one to five percent of budgetary executions between 2009 and 2011. In 2011, the expenditures made without prior authorization were regularized at the end of the fiscal year, though it is unclear whether regularizations were done within the one month period specified by Burundi regulations. 17. Excluding debt payments, the rationale for payments without prior authorization is unclear. Payments without prior authorization (sans ordonnancement préalable) can derive either from ministerial decisions (décisions ministérielles), or from a list (liste ministerielle) of items contained in Ordonnance ministerielle 510/0015 of January 9, 2008. This list reflected Government efforts to regularize and monitor payments without authorization. However, the data extracted from the Treasury balances does not provide details on the budget items that were paid without authorization as a result of a ministerial decision and the team was not able to collect data on that. Regarding the Ordonnance, most of the items that can be paid outside the expenditure chain are neither part of the pro-poor budget, nor are they linked with the productive sector of the economy. Instead, most of these items refer to back office functions and transport allowances for high level politicians, such as parliamentary missions to the interior of the country (see Annex F for further details). 75 Though the Burundi authorities claim that the Ordonnance has not been valid since 2009, data extracted from the treasury balances and presented in the Settlement Laws (Lois de Règlement) of 2008-11 indicate that payments without authorization are still passed through the liste ministérielle subaccount. 18. Debt payments are not registered appropriately in SIGEFI. While in principle all payments should go through regular budgetary and treasury controls, Burundi, similarly to other African countries (notably the member states of the West African Economic and Monetary Union (WAEMU) allow debt payments to be treated outside the expenditure chain. Though debt payments can be done outside the expenditure chain, they should, in principle, be regularized within one month. The numbers presented in the Loi de Règlement are different from those extracted from SIGEFI, which suggests that regularizations are not performed appropriately. Public officials at the Ministry of Finance interviewed for this study were not able to provide explanations regarding these discrepancies. 75 The list of items that can be paid without authorization is available in Annex F.6. 77 19. It is necessary to complete the link between SIGEFI and the Treasury, as well as revising the processes to regularize expenditures. The analyses presented in this section have shown that SIGEFI does not produce reliable and analyzable information (Annex F provides additional details on SIGEFI and its challenges). 4.3 External Financing 4.3.1 Framework of aid coordination 20. Prior to 2007, a situation pertained to aid coordination where ministerial responsibilities overlapped and some skill related conflicts emerged amongst Burundi’s aid partners. Coordination issues previously handled between ministries, chiefly the Ministry of Finance and Economic Development Planning, the Ministry in charge of Planning, the Ministry of Foreign Affairs, and the Ministry of Interior were now to become the operational responsibility of the new Aid Coordination National Committee (Comité National de Coordination de l’Aide / CNCA) and its Permanent Secretariat (SP/CNCA). Additionally, in an effort to further improve the work of the new system, the Government developed a broader framework for partner coordination comprised of the Political Forum (Forum Politique), the Strategic Forum (Forum Stratégique), and Sector Groups (Groupes sectoriels) for sector based guidelines. This newly established architecture is detailed in Annex G. 21. The new aid management platform implies direct data entry from each donor in the database and pivots around the operation of Sector Groups. In collaboration with each Sector Group leader, the Sector Groups are responsible for all strategic planning operations within their respective sectors, which are structured to be concordant with the general and specific orientation of the PRSP. 4.3.2 Analysis of aid allocations and expenditures (2007-2011)76 Analysis per donor 22. Table 4.7 shows per donor engagements and disbursements for the 2010-2011 period other than budget support. 77 In 2010, the completion rate is 110 percent, with a disbursement level of US$462 million against a US$419 million forecast.78 It must be noted that the 110 percent completion rate does not necessarily translate into an increase in absorbing capacity. However, there is a possibility that engagement data is incomplete. 79 There was a decrease in development public aid between 2010 and 2011. Engagements decreased from US$419 million to US$416 million. In terms of volume, disbursements went from US$462 million in 2010 to US$401 million, almost a 15 percent drop. Over both years, disbursement volume is of US$863 million against a US$836 million engagement level, i.e. a 103 percent rate. 76 The analysis in this section relies on data extracted from the CNCA. Uncertainty remains on the quality of the database. However, CNCA is the only source detailing disbursements for projects executed and funded by donors. 77 This analysis excludes budget support for two reasons. First, chapter two includes a whole section dedicated to budget support. Second, CNCA data largely differs from Central Bank (BRB) figures. Generally, Burundi Central Bank data should be more reliable as they indeed correspond to amounts received on BRB Treasury account. 78 This analysis also excludes donors without any specific agreement and whose funding is received through NGOs 79 During meetings with the PER team, CNCA civil servants have remarked that engagement data starts to be fairly reliable from 2010 onwards, although problems still remain with donors who do not necessarily report on all engagements. 78 Table 4.7: Engagement and disbursement per donor (2010-2011) Year 2010 2011 Total 2010-2011 Planned Actual Rate Planned Actual Rate Planned Actual Rate ACBF 1.0 0.8 74% 0.5 0.0 0% 1.6 0.8 49% Germany 6.2 50.1 805% 40.6 47.4 117% 46.8 97.5 208% BADEA 0.0 0.3 7.9 8.0 101% 7.9 8.3 105% African Development Bank 20.6 19.6 95% 4.1 1.0 24% 24.6 20.6 84% World Bank 100.6 68.2 68% 78.9 71.4 90% 179.5 139.6 78% Belgium 42.0 32.2 77% 46.2 45.7 99% 88.2 77.9 88% China 0.3 0.0 0% 0.0 0.0 0.3 0.0 0% European Union Commission 57.4 85.0 148% 61.6 37.5 61% 119.1 122.5 103% Spain 0.4 0.0 0% 0.4 0.0 0% 0.8 0.0 0% United States of America 2.5 40.2 1576% 9.6 52.5 545% 12.2 92.7 761% France 7.5 6.8 91% 9.7 7.5 77% 17.2 14.3 83% Global Alliance Vaccines and 0,8 4,6 609% 0,8 2,5 331% 1,5 7,1 470% Immunization Global fund 18,0 2,1 12% 11,8 9,6 81% 29,8 11,8 39% Japan 12,6 16,1 128% 6,8 3,8 56% 19,4 20,0 103% Luxemburg 0,1 0,1 100% 0,0 0,0 0,1 0,1 100% United Nations 93,1 89,1 96% 88,3 63,3 72% 181,4 152,4 84% Norway 3,0 2,6 86% 0,4 0,1 30% 3,4 2,7 79% Org. of Petroleum Exp. Countries 3,4 0,9 26% 13,2 11,4 86% 16,6 12,3 74% Netherlands 8,4 9,3 112% 4,1 9,9 244% 12,4 19,2 155% United Kingdom 33,7 24,7 73% 20,0 18,8 94% 53,7 43,5 81% Switzerland 8,0 9,7 121% 11,9 11,0 93% 19,9 20,7 104% Overall Total 419,4 462,5 110% 416,9 401,4 96% 836,3 863,9 103% Source: CNCA Database and authors’ own calculations. 23. The disbursement decrease observed in 2011 could be explained by two main factors: First, some important donors were at the end of their planning cycles. For example, this was the case with the World Bank whose engagements went from US$100 million to US$79 million in 2011; and the European Union, whose disbursements went from US$85 million to close to US$37 million in 2011. The second reason may lie with the international recession forcing some countries to adopt austerity budgets which consequently compress engagements for Burundi. 24. Multilateral aid is fairly predictable as donors have multiyear programs with the country. Bilateral donor coverage is variable:80 As Figure 4.6 shows, bilateral donor countries with indicative cooperation programs with Burundi (Germany, Belgium, Japan, etc.) find no difficulty in communicating engagement and disbursement forecasts. On the other hand, aid from countries with no reference cooperation programs directly with the country, headed by the U.S., is totally unpredictable. Such resources are often mobilized in favor of NGOs or through International Organization channels. 80 This excludes donors’ budget support. The data provided by the Central Bank (BRB) on budget support was considerably different from that registered in the CNCA. As the BRB is the original recipient of those funds, it is clearly a more reliable source of information. The fact that donors fail to report budget support is, in itself, a worrisome sign. 79 25. Despite progress, predictability issues Figure 4.6: Mapping key donor’s in terms of aid remain and can continue to engagement and disbursements (2010- compromise planning and budgeting 2011) operations and thus overall aid efficiency. 140 The situation requires the development and improvement of consultation approaches and 120 mutual arrangements between Government and partners at various aid management process 100 levels. 80 26. From a more general perspective, the 60 tracking of aid estimates and execution is 40 plagued by inconsistent data collection and recording systems. There is an urgent need to 20 have a single data repository linking donor- funded projects and expenditure financed with 0 the Government’s own resources. 2010_Commitment 2010-Disbursement 2011_Forecast 2011_Disbursement Source: CNCA Data. 80 27. Further analysis of Figure 4.7: PRSP focus areas, external funding spending around PRSP-II areas 250 indicates that the cluster area 200 for “Human Capital 150 100 Development” has benefited 50 Amount from significant resources from 0 Promot donors allowed the country to Improv ion of obtain encouraging results in ed Human Sustain Govern the health and education ance capital Fightin able sectors. This focus area received and develo g AIDS Econo US$382 million against a pment mic Securit Growth US$302 million forecast over the y and… period, with a 27 percent 2010-Forecasts 81 128 26 184 overshoot reflecting the 2010_Executed 74 200 12 176 predictability challenge in this area. The most significant 2011_Forecasts 41 174 17 184 forecast overshoots are noted 2011_Executed 60 185 18 138 with the United States, Germany and the European Union. For the Source: CNCA Data, other than NGO and budget support. United States of America, there is an overshoot level of close to US$80 million against the forecast in 2010 and 2011, while Germany has only registered a US$50 million overshoot level. The case of the European Union is more nuanced. Overshoots and losses compensate one another over the period with a US$27 million overshoot level in 2010 and a US$24 million loss in 2011, reflecting a US$3 million overshoot during that period. Overall, this focus has played an important role in the encouraging results achieved in the health and educations sectors (CAS FY13-16). 28. Although health and education Figure 4.8: Disbursements by Ministry, 2007-11 are priority sectors, there seems to be (million of US$) little coordination between donors and Government to adequately fund these areas and systematically track resource allocations. The projects linked to the ministries in charge of education and health saw a remarkable increase between 2007 and 2010 (from USS26 to 85 million) just to see a dramatic fall to 69.7 in 2011. The amounts disbursed as part of the projects linked to the sector of education were low throughout the sample period. It is notable that the focus area for “Promoting Sustainable and Equitable Economic Growth” benefited from US$314 million as against a US$368 million forecast. Further, the disbursement level has dropped substantially, from US$176 million in 2010 Source: CNCA Data. 81 to US$138 million in 2011, a 22 percent decrease against the US$184 million originally forecasted in 2010 and 2011. Ministry by ministry allocations graphically displayed below, are dealt with in greater detail in Annex G. 4.3.3 Aid management and reporting: key challenges 29. Reporting on aid planning requires improvements to function as effective budgetization tool: data on donor-funded projects reported in the CNCA database is consistently higher than the figures in the finance law approved by the Parliament. Several factors explain these discrepancies, including different regulations between finance and CNCA,81 and time differences (i.e., CNCA includes projects committed and signed after the finance law is presented to Parliament). With the fusion of the Ministry of Finance and Planning, progress has been registered: in 2012 there was a remarkable reduction in the discrepancies between the amounts recorded in the CNCA database and those presented to Parliament (Table 4.8). This suggests an improvement in aid programming and budgetization consistency. However, about 25 percent of donor-funded projects are still not reflected in the budget, indicating that further efforts are needed to improve reporting on aid flows and allow proper budget planning. Table 4.8: Project recorded in the finance law and CNCA data Year Budget Planned amounts reported Amounts included in the budget law as Law by donors to the CNCA % of amounts recorded in the CNCA database 2010 189 515 36.7 2011 282 513 55.0 2012 411 545 75.4 Note: Data displayed in this table excludes budget supports, grants benefitting NGOs, and those coming from the private sector. Source: Finance Law (Loi des Finances) and CNCA. 30. Data on executed project reveal inconsistencies and are not fully captured in the budget (Loi de Règlement): There are large differences between the execution amounts recorded in the CNCA and those presented to Parliament. Contrary to budget amounts, there has been little improvement for the reporting of donor-funded executed projects. In 2011, only about 16 percent of the amounts executed and registered by donors at the CNCA were actually included in the Settlement Law (Loi de Règlement). While the figures reflect a substantial improvement from the previous year, they indicate that about 84 percent of the donor-funded executed amounts are not reflected in the budget. One of the main reasons behind this gap is that few projects funded by donors pass through the Government’s treasury account, with the majority of disbursements reflected in commercial bank accounts. There is no reporting mechanism in the Treasury for the moment. The Government should launch discussions with the development partners to review together strategies to improve coordination among all the actors involved in the budget preparation process, strengthen the quality of budge data and promote an integrated approach to 81 Burundian regulations require that the finance law presented to Parliament includes exclusively the projects for which there is a financing agreement signed between the Government of Burundi and the donor. However, the CNCA database does not follow this regulation and committed projects without signed financing agreements are included. 82 budgetization. The possibility of integrating donor disbursements in a Treasury subaccount might also be explored as a medium-term goal, subject to substantial improvements in the PFM systems.82 Table 4.9: Executed project recorded in the Settlement Law and CNCA Data Year Settlement Actual values reported Amounts included in the Settlement Law as a Law by donors to the CNCA percent of amounts recorded in the CNCA database 2010 37 569 6.5 2011 80 506 15.8 Note: Data displayed in this table excludes budget supports, grants benefitting NGOs, and those coming from the private sector. Data for 2012 were not available from either source at the time of producing the report. Source: Settlement Law (Loi de Règlement), and CNCA. 31. All together, these challenges seriously undermine aid efficiency in Burundi. Given the discrepancies discussed above, donor-funded development projects are not fully integrated in the budget cycle. Though improvements have been made in terms of projects’ planning, discrepancies in execution data remain significant, revealing aid fragmentation. This in turn undermines the integration between the government priorities and the donors’ programs. A better coordination among donors, CNCA, and the MoF is urgently needed to fully integrate aid into the budget cycle and promote effective spending. 4.3.4 Key recommendations 32. The analysis in the previous sections suggests a series of recommendations for both donors and Government towards mutually responsible aid management. 33. For donors: Aid efficiency concerns constitute a key issue for developing countries such as Burundi. To improve predictability, donors should make an effort and benefit from the new aid management platform advantages to provide forecast and achievement data for their interventions. The coming years correspond to CSLP-II implementation with the will expressed by the Government to promote rapid, sustained and employment creating growth. Such a will shall not be achieved without the support of regular and potential Government partners. This commitment was shown during the donor roundtable talks in Geneva. It would be desirable that donors materialize expressed promises by a stronger support for growth promotion actions in order to fully achieve CSLP-II. Finally, it is important to distinguish between project operating costs and actual investments. 34. For the Government: Reinforcing upstream and downstream existing links among planning tools in order to enhance development aid efficiency remains a necessity. The budget management software in the Ministry of Finance must take into account the pressing necessity of ensuring the networking of the aid management platform, the Central Bank, the payroll and debt management software, etc. Efforts should be concentrated on budget preparation level as well as on development policy monitoring evaluation. This includes enhancing the transparency of the budget preparation process, as well as its execution. Likewise, the Government must implement 82 The difficulties faced by the Burundi authorities to reconcile the treasury balances with SIGEFI (see chapter four) suggest that full integration within the treasury account, while desirable, might not be feasible in the short term. 83 a strategy to reduce and eventually eliminate all violations to the expenditure chain. The Government should also pursue its continuous dialog with its partners on public aid management improvement mechanisms. A particular focus should be placed on aid operations reporting in the central administration. To conclude, in order to develop a public finance management integrated approach, data related to external funding should be integrated in the budget according to current nomenclatures rather than being annexed to it. These are important resources contributing to Government program implementation. 4.4 Selected Non-salary Public Expenditures 4.4.1 Pro-poor priority programs 35. Since 2008, the Government has managed to finance some poverty related priority programs, assuring that 80-95 percent of opened credits 83 come to the payment phase within the budget year. In fact, the quasi mandatory characteristic for poverty reduction and peace consolidation tend to guarantee their execution even in case of cash constraints / shortages. Therefore, the Government has endeavored to allocate available cash receipts to guarantee implementation of the activities of these priority programs,84 ring fencing them with a sense of urgency. Most of these priority programs were either created (e.g., school feeding, irrigation of major regional perimeters, strategic stock of fertilizers, multiplication/distribution of high yielding varieties, etc.) or expanded (e.g., support to repatriation process, allocation to persons affected by civil strife, boarding school allocation, etc.) through donor budget supports to mitigate the impact of international food, fuel, and debt crises. Table 4.10 Ring fenced programs in Burundi’s post-conflict context (billion FBu) Credit Commitment Liquidity Authorization Payment 2008 (revised budget) School feeding and boarding schools 10.99 10.96 10.96 10.96 10.75 Free fees policy (health, education) 13.51 13.49 13.49 13.40 13.41 Scholarship and student travel 9.19 9.10 9.10 9.10 9.10 Support to HIV/AIDS persons 3.69 3.62 3.55 2.09 2.09 Repatriated, displaced, and emergency 5.41 5.29 5.05 4.20 4.19 Food security programs 7.66 7.66 7.66 7.66 7.66 Total 50.45 50.12 49.81 47.40 47.21 2009 (revised budget) School feeding and boarding schools 13.75 13.74 13.74 13.74 13.74 Free fees policy (health, education) 14.72 17.84 17.84 17.84 17.84 Scholarship and student travel 11.28 8.52 8.52 8.52 8.52 Support to HIV/AIDS persons 1.70 1.67 1.67 1.67 1.67 Repatriated, displaced, and emergency 3.67 3.08 2.95 2.95 2.95 Food security programs 6.44 3.03 3.03 2.07 2.07 Total 51.55 47.88 47.75 46.79 46.79 2010 School feeding and boarding schools 15.89 9.88 9.88 9.88 9.88 83 Open credits refer to voted credit that have been adjusted based on the 10 percent limit provisioned in the Loi Organique of 2008. The figures of open credits are extracted from SIGEFI. 84 The PFM reform has contributed to the initiation of some of these priority programs, including school feeding, strategic stock of fertilizers, small scale irrigation schemes. Other programs, i.e., support to HIV/AIDS, repatriated and other internally displaced persons, have been rendered mandatory by the post-conflict context. 84 Free fees policy (health, education) 16.90 16.87 16.87 15.96 15.96 Scholarship and student travel 14.64 12.81 12.81 12.81 12.81 Support to HIV/AIDS persons 2.30 2.19 2.09 2.09 2.10 Repatriated, displaced, and emergency 7.32 7.21 5.68 5.68 5.68 Food security programs 8.91 7.34 7.05 6.92 6.95 Total 65.96 56.30 54.39 53.35 53.40 2011 School feeding and boarding schools 14.76 14.75 14.75 14.74 14.74 Free fees policy (health, education) 16.28 16.28 16.28 16.28 16.23 Scholarship and student travel 11.08 10.72 10.72 10.60 10.49 Support to HIV/AIDS persons 2.55 1.98 1.98 1.57 1.58 Repatriated, displaced, and emergency 5.77 5.55 5.55 5.52 5.52 Food security programs 22.63 21.04 21.04 20.08 20.23 Total 73.08 70.31 70.31 68.79 68.78 2012 (revised budget) School feeding and boarding schools 15.12 15.06 15.06 15.06 15.06 Free fees policy (health, education) 20.70 20.70 20.70 20.70 20.70 Scholarship and student travel 14.29 14.28 14.28 14.23 14.27 Support to HIV/AIDS persons 2.18 1.74 1.47 1.43 1.45 Repatriated, displaced, and emergency 5.69 5.68 5.64 5.63 5.63 Food security programs 21.21 18.18 16.24 16.21 16.32 Total 79.18 75.63 73.38 73.25 73.42 Source: Budget data from SIGEFI. 36. Representing nine to 11 percent of total open credits from 2008 to 2012, these specific priority programs have to compete for unpredictable cash resources. Public expenditures generated by the policy of free primary school and of free health care (for pregnant women and children under five years old) have always been financially covered, the lowest execution rate of 94.5 percent appearing in 2010. Contrastingly, HIV/AIDS programs have shown the lowest execution rates (cash / payment basis) in three out of five examined budgetary years, and food security promotion appeared to be the least executing program in 2009. Other relatively important and well performing priority programs are found in the education sector, in particular school feeding, boarding schools (in French, “les Internats”), and student scholarships. For these three categories, opened credits have regularly reached the payment step at rates close to 100 percent, with the exception of 2009 (where a drastic cut happened in scholarships because of cash shortages) and 2010 (FBb6 billion school feeding credits were sine die cut-off during the 2010 election year). Domestically financed activities in support of both the repatriation process and conflict related vulnerable people have improved in execution terms over the years, but their relative share has diminished steadily, lingering at around seven percent of the 2012 opened credits for the identified priority programs, as Burundi moves away from the emergency economic standpoint. 37. Out of the priority programs identified, allocated credits have significantly increased in favor of direct interventions aimed to promote food security in Burundi. The share of food security promoting programs in total priority programs has almost doubled, averaging 28.9 percent in 2011-2012 from 15.2 percent in 2008. This evolution clearly marks the commitment of the Government to become more active in this critical sector for the great majority of the rural people’s wellbeing. As compared to the record during the PRSP-I (average of 70.0 percent from 2008 to 2010), execution performances (payment basis or authorization basis) have improved, but further efforts are needed to reach performance levels similar to that of 85 free policy and education programs. Irrespective of the basis retained, execution rates remained around 82.6 percent in 2011-2012. Probably at the center of this erratic performance, we have, inter alia, ineffective budget preparation, difficulties in always succeeding to procure capital goods within the budget year, and uncertainty in the timeline of financial aid disbursement. 4.4.2 Areas aside from pro-poor 38. Some well executing expenditure areas have low pro-poor linkage and tensions can emerge regarding their continuing allocation levels in conditions of cash scarcity. Examples are external, official missions (96.3 percent in 2011-12) and special commissions (94 percent most recently). Table 4.11 Three expenditures with no prior authorization and official rents (billion FBu) Credit Commitment Liquidity Authorization Payment 2008 (revised budget) Mission outside Burundi 5.77 5.73 5.76 5.67 5.72 Mission inside Burundi 0.88 0.88 0.88 0.87 0.82 Commission 0.66 0.66 0.66 0.66 0.65 Estate rent (office and lodging) 14.86 14.78 14.78 14.78 14.72 Ministry of Transport 6.19 6.15 6.15 6.15 6.09 2009 (revised budget) Mission outside Burundi 6.11 6.05 5.97 5.90 5.93 Mission inside Burundi 0.97 0.96 0.96 0.96 0.94 Commission 3.07 3.06 3.06 3.06 2.95 Estate rent (office and lodging) 13.86 13.62 13.60 13.60 13.54 Ministry of Transport 4.54 4.53 4.53 4.52 4.50 2010 Mission outside Burundi 5.38 5.21 5.20 5.09 5.11 Mission inside Burundi 1.04 1.03 1.03 1.03 1.01 Commission 12.10 11.67 11.67 11.63 11.46 Estate rent (office and lodging) 15.59 15.53 15.53 15.47 15.44 Ministry of Transport 5.60 5.57 5.57 5.53 5.51 2011 Mission outside Burundi 6.09 5.85 5.85 5.76 5.75 Mission inside Burundi 1.03 1.03 1.03 1.02 1.01 Commission 3.87 3.83 3.83 3.82 3.64 Estate rent (office and lodging) 17.28 16.42 16.42 16.34 16.31 Ministry of Transport 5.40 5.39 5.39 5.35 5.33 2012 (revised budget) Mission outside Burundi 7.04 6.94 6.93 6.90 6.91 Mission inside Burundi 1.25 1.24 1.24 1.24 1.24 Commission 5.18 4.94 4.94 4.94 4.92 Estate rent (office and lodging) 18.73 18.47 18.47 18.45 18.46 Ministry of Transport 7.32 7.30 7.30 7.30 7.29 Source: Budget data from SIGEFI. 39. Another important source of tension in budget management has originated in the rent of buildings and houses assumed by the State. In 2010, yet another special technical commission was appointed to make proposals in order to solve durably the issue of the explosion of official rents in Burundi’s budget. In principle, the Ministry of Transport is the only agency in 86 charge of renting real estate. However, as the table above shows, other agencies also allocate money to rental. 4.4.3 Domestically financed capital expenditures 40. Despite execution improvements over the years, capital expenditures under domestic budget financing have suffered the most from shortages in disposable public funds. Starting from a low average of 67.2 percent in 2005-2006, the commitment of opened credits reached 78.9 percent in 2007-2010 during the PRSP-I and 85.6 percent in 2011-2012 during the PRSP-II. But, the Government still ends up with 14.4 percent of budgetary credits for capital expenditures that are deemed ineligible for the commitment phase in the spending chain, including even during budget revisions (2008, 2009 and 2012). In terms of authorization and payment, in fact the execution rates of capital expenditures show a lot less progress, as only 76.9 percent of opened credits were authorized in 2011-2012, against 72.9 percent in 2007-2010 and 65.8 percent in 2005-2006, similarly for payment based execution rates. More efforts should be devoted to ring fence capital expenditures, given the context of meager domestic resources allocated to such expenditures and low execution rates of externally financed capital goods. Table 4.12 Burundi’s capital spending in electricity, road, and state transport (billion FBu) Credit Commitment Liquidity Authorization Payment 2008 (revised budget) Electricity 3.77 2.39 2.40 2.33 2.33 Road 5.16 5.06 5.06 5.03 3.79 State vehicle / transport 1.58 1.58 1.58 1.58 1.58 2009 (revised budget) Electricity 3.66 2.24 2.24 1.98 1.99 Road 5.60 4.80 4.80 4.80 4.20 State vehicle / transport 1.83 1.83 1.83 1.83 1.83 2010 Electricity 7.29 5.69 5.69 5.15 5.10 Road 7.49 6.01 6.01 6.01 5.50 State vehicle / transport 0.54 0.18 0.18 0.18 0.18 2011 Electricity 12.46 11.41 11.41 5.12 5.03 Road 5.91 4.70 4.70 4.70 4.56 State vehicle / transport 0.44 0.35 0.35 0.35 0.35 2012 (revised budget) Electricity 17.41 12.87 12.87 12.87 13.26 Road 4.34 4.34 4.34 4.34 4.34 State vehicle / transport 0.24 0.24 0.24 0.24 0.24 Source: Budget data from SIGEFI. 41. In the critical energy sector, credits have significantly increased, reaching FBu 17.41 billion in 2012 (revised budget) and FBu 12.46 billion in 2011, up from FBu 3.77 billion in 2008 (revised budget). With an average of 73.6 percent of opened credits that could be committed from 2008-2012, the Government is probably facing issues: (i) to ensure quality (in terms of economic and financial analyses) in programmed electricity projects; (ii) to align such projects with sector priorities; (iii) to always successfully procure power generation equipment for various reasons of capacity, procedures, and narrow field of bidders; and (iv) to secure financial 87 resources for programmed power investments due to tight cash management constraints. An unusually high performance in the commitment based execution rate occurred in 2011 (91.6 percent of the FBu12.46 billion were committed). A clear understanding of the underlying reasons of this performance, which is excellent especially for this sector so used to poor execution rates, could help address the issues face by the Government down the road. Efforts to increase the production of electricity are further undermined by drastic resource cuts at both authorization and payment stages. Domestically financed projects aimed to increase electricity production have executed dismally from 2008-2012, averaging 60.3 percent (authorization basis) and 60.6 percent (payment basis). Moving in to the PRSP-II period performance has picked up, however, and the Government has been able to almost quadruple the level of opened credits that translate into budgetary payments for electricity promotion interventions (i.e., the 10.4 MW dam of Mpanda, microhydropower dams, etc.), moving from an average of FBu 3.61 billion in 2008-2011 to FBu 13.26 billion in 2012. 42. For roads, the Government has managed to maintain satisfactory execution performance, but credit allocations in this sector have remained limited.85 As far as road rehabilitation and the State’s financial contributions to ongoing donor-funded road projects are concerned, opened credits have averaged FBu 5.7 billion from 2008-2012, with a peak of FBu 7.49 billion achieved during the 2010 budgetary year. In terms of authorization and payment of these opened credits, yearly averages of FBu 4.97 billion and FBu 4.48 billion were processed by the authorities. Authorization based execution rates declined from 97.4 percent in 2008 to 79.5 percent in 2011, before increasing to 100 percent in 2012. Payment based execution rates showed slight variability from 2008-2011 at around 73 to 78 percent, but greatly increased in 2012 (100 percent execution rate). 4.4.4. Miscellaneous expenditures 43. The long tradition of Burundi to financially contribute to a myriad of membership organizations has raised concerns in the tight budgetary environment. The number of such memberships has increased regularly (35 in 2008 to 52 in 2012), costing the equivalent of FBu 7.58 billion on average from 2008 to 20129 (with a peak of FBu 11.93 billion in 2010). In 2011, a particularly stressing situation happened because of an off-expenditure chain expenditure that the Government had to honor (contribution to EAC), causing (budgetary) credit overruns of FBu 7.76 billion that were authorized and paid. Table 4.13 Prioritized expenditures with unclear pro-poor impacts (FBu billion) Credit Commitment Liquidity Authorization Payment 2008 (revised budget) Contribution to organizations 2.94 2.93 2.93 2.86 2.85 EAC 1.08 1.08 1.08 1.08 1.08 Participation to capital 0.10 0.10 0.10 0.10 0.10 increase 2009 (revised budget) Contribution to organizations 6.50 6.46 6.46 6.46 6.47 EAC 4.23 4.23 4.23 4.23 4.23 85 This excludes the activities of the “Fonds Routier National.” 88 Participation to capital 0.91 0.00 0.00 0.00 0.00 increase 2010 Contribution to organizations 11.93 11.73 11.73 11.70 11.70 EAC 8.00 8.00 8.00 8.00 8.00 Participation to capital 1.52 1.40 1.40 1.40 1.40 increase 2011 Contribution to organizations 10.83 18.81 18.81 18.59 18.59 EAC 7.46 15.46 15.46 15.46 15.46 Participation to capital 6.23 4.09 4.09 4.09 4.15 increase 2012 (revised budget) Contribution to organizations 5.72 5.64 5.64 5.32 5.17 EAC 1.47 1.47 1.47 1.47 1.47 Participation to capital 2.82 2.01 2.01 2.01 2.03 increase Source: Budget data from SIGEFI. 44. Burundi’s participation as a shareholder to the capital / stock increases of the organizations belonged to require also further attention by the authorities, though it should be noted execution of this expenditure has declined significantly to 71.2 percent (authorization) and 72.1 percent (payment) in last three years. 4.5 Pro-poor Budget Spending 45. Data quality seriously undermines the scope of the analysis. Though the pro-poor budget complies with the Joint Guidelines written by the World Bank and the Government of Burundi, pro-poor spending has little continuity across time: the budget lines that are identified as pro-poor change every year, making comparison difficult 86 (a detailed examination of the evolution of pro-poor spending is contained in Annex H). 46. The limited analysis that is feasible – given data constraints - shows increases in budgetary share and generally high execution rates for pro-poor spending. But, notwithstanding progress in social sectors, Burundi is still lagging behind the MDG targets. 86 The government is in the process of changing again the classification of pro-poor spending with the introduction of the functional classification defined in the new chart of accounts. It is expected that this change will reinforce the stability and predictability and will enhance, therefore, the evaluation of pro-poor spending. 89 At the ministry level, the analysis shows Figure 4.9: Where does the pro-poor budget go? that the pro-poor budget is concentrated in the priority sectors of health, 400,000,000,000 education, agriculture, energy and 350,000,000,000 mining, and public works, as well as the 300,000,000,000 19.6% 20.3% social services. This indicates that, at 250,000,000,000 least at the allocation level, the pro-poor 200,000,000,000 10.0% budget is consistent with the poverty 150,000,000,000 19.9% reduction strategy (PRSP-II). The 100,000,000,000 14.2% 79.7% 80.4% 90.0% Ministry of Basic and Secondary 80.1% 50,000,000,000 85.7% Education87 managed the highest amount (FBu) of the pro-poor funds (average of 46 0 2008 2009 2010 2011 2012 percent). This is in line with the Current expenditure Capital authorities’ policy goal of expanding access to basic education. Moreover, the Source: Budget data from SIGEFI Ministry of Superior Education and Research was the second highest recipient of the pro-poor budget, receiving on average 12.1 percent of the funds. Adding these two ministries, 60 percent of the pro-poor budget was allocated to education. The third largest manager of pro-poor spending is the Ministry of Health and Fight Against AIDS,88 which received, on average, 12.5 percent of the pro-poor budget. Although it has the highest concentration of pro-poor budget lines, the Ministry of Agriculture handles only 5.5 percent on average of the total pro-poor budget. Table 4.14 Pro-poor budget distribution by administrative classification of spending 2008 2009 2010 2011 2012 Average General Services 9.2% 7.4% 8.7% 15.8% 10.5% 10.3% Social Services 72.3% 76.2% 79.3% 69.5% 72.1% 73.9% Economic Services 18.5% 16.5% 11.9% 14.6% 17.4% 15.8% Source: Budget data from SIGEFI. Table 4.15 Pro-poor budget allocation in priority sectors Priority sectors 2008 2009 2010 2011 2012 Average Health/AIDS 9.4% 10.5% 12.9% 15.9% 13.5% 12.5% Education 59.7% 62.7% 63.2% 51.7% 56.8% 58.8% Agriculture 7.3% 5.3% 3.0% 6.3% 5.9% 5.6% Energy and Mining 1.2% 3.6% 1.5% 2.6% 5.6% 2.9% Public Works 8.2% 5.7% 5.4% 3.9% 4.1% 5.5% Grand Total 85.8% 87.7% 86.0% 80.5% 85.9% 85.2% Source: Bank staff own calculation with data from SIGEFI. 87 The name of the ministry has changed over time. 88 Before 2011, this ministry was divided in two ministries. We added the two ministries to make this comparison. 90 47. Among the priority sectors, we observe that Health/AIDS, primary Figure 4.10: Execution rate of the pro-poor education and higher education have the budget highest executions rates, with an average 120% (Committed / Initial credit) of 92 percent, 93 percent and 90 percent, 100% respectively. The execution rate for the Ministry of Agriculture was 74 percent in 80% average, while the Ministry of Energy and 60% Mining registered the lowest execution rate 40% with an average of 56 percent. Capital 20% expenditure has 20 percent of the pro-poor expenditure. Within this classification, the 0% 2008 2009 2010 2011 2012 largest expenditure is on fixed assets with Current expenditure Capital expenditure an average of 12 percent of total pro-poor Total expenditure budget. In line with the rest of the budget, capital expenditures present an execution Source: Budget data from SIGEFI. rate of 69 percent for the period between 2008 and 2012. 48. The fact that the performance of pro-poor spending is not translating more identifiably into realization of MDG goals will give the Government pause for thought. According to the 2012 Joint Staff Advisory Note on the Second Poverty Reduction Strategy Paper (PRSP-II), while the effective implementation of the PRSP-II will help to converge toward the MDG’s, very few of them will be met by the end of 2015. The document estimates that poverty, extreme poverty, and hunger rates are unlikely to be halved within the next four years in comparison to the 1990 benchmark. Moreover, none of the health and social protection MDG’s will be met, despite notable progress observed in some important areas such as HIV/AIDS prevalence. 4.6 Experience with MTEF: a stock taking exercise Box 8 The Medium-Term Expenditure Framework: an overview of the process The MTEF involves a top-down process of defining a resource envelope that is consistent with macroeconomic stability and explicit strategic priorities, while providing bottom-up estimates of the current and medium term costs of existing and new programs. Iterative decision making helps to resolve key tradeoffs and match costs with available resources, thereby encouraging strategic allocation of resources in line with government priorities. The output of a MTEF process is often a multiple year rolling plan, integrating capital and recurrent expenditures, and comprehensively including all expenditure whether funded by the Government or by the donors. In conclusion, a MTEF aims to integrate policy with resource allocations decision in a multiple year context to better reflect the effects of decision on public finances, and generally lead to improved policy decisions. 49. Medium-Term Expenditure Frameworks (MTEFs) were introduced to Burundi in the context of targeting budget formulation and execution to results. A first attempt to 91 introduce program budgeting at the sectorial level (education, health, justice) took place in the second half of the 2000s, to allow line ministries to better link sector policies with budget allocations. The 2008 PEMFAR praised these early attempts but mentioned the following as preconditions for Burundi to implement a program based budget successfully (World Bank, 2008): • Implementation of sound budgetary operating procedures, as provided for in the draft organic law and evidenced by the assessments provided by internal audit and inspection bodies (including Inspection Générale de l’Etat, IGE); • A significantly higher degree of consolidation of public expenditure into the budget than was currently the case at that time (i.e., a marked increase in on-budget expenditure); • A reasonable degree of financial discipline, as evidenced by the credibility of the budget (comparison of budget estimates and outturns, both in aggregate and by sector); • Reasonably accurate recording and reporting of financial and physical data, as evidenced by the regular production of consistent Government accounts and the opinions of external auditors. 50. A series of diagnostic work (PEMFAR 2008, PEFA 2009, 2009, World Bank 2011) recommended the Government to progressively begin the introduction of a medium-term expenditure framework to improve budget planning and credibility, observing that the absence of an MTEF prohibits the optimization of actions in the medium term and timely planning of resource availability. The case was also made to use the MTEF instrument to identify and reduce low efficiency expenditures. With the support from the donor community (including TA from the World Bank), in 2010 the Government adopted and institutionalized the process to elaborate a central MTEF as a framework for budget programming. 89 This has laid the foundation for improved in budget planning (as captured by the PEFA score on “budget preparation” change from D (2009) to B (2011). 51. Burundi produced its first central MTEF in 2010 (to cover the fiscal exercises of 2011-13) with assistance from the World Bank. Since then, it has produced two additional documents which are approved by the Council of Ministries and submitted to the Parliament as part of the lettre de cadrage macroéconomique. The lettre de cadrage highlights the assumptions of the macroeconomic model and provides some macroeconomic context. It then presents the medium-term budgetary forecasts in two formats: (i) the first format is the TOFE; (ii) the second one distributes the budget envelope in three components: resources allocated to ministries, (envelopes alloués aux ministères), common charges (charges communes), and arbitrage provisions (provisions pour arbitrage). In principle, arbitrage provisions are “leftovers” of the overall budget envelop to be allocated among ministries. An annex to the document specifies the amounts allocated to each ministry in terms of recurrent and capital expenditures. 52. The methodology and the format in which the data is presented have significantly changed every year since 2010, making comparisons over time difficult. As the analysis will 89 According to the ERSG documents, “a central MTEF (in contrast to sector MTEF) is a tool developed by central ministries such as the Ministry of Finance and the Ministry of Planning (now combined into a single ministry: Ministry of Finance and Economic Development Planning), which indicates the global envelop of resources expected based on macroeconomic projections and provides expenditures ceiling for each institutions financed through the budget. 92 show, the changes in the methodology are one of the reasons behind the discrepancies within the document. The main changes in format are: • The 2010 MTEF (covering fiscal years 2011-13) does not include the arbitrage provisions, which were included in the 2011 exercise. • In the annex breaking down the envelopes by Ministry, the 2011 MTEF (covering fiscal years 2012-14) distributed capital expenditures by external and internal resources. This feature, which is quite relevant for Burundi as most of the investment budget is financed through donors, disappeared in the next exercise. • The TOFE of the 2012 MTEF (covering fiscal years 2013-15) presents the data in GFS 2001 format. As will be discussed below, the 2012 MTEF is the one with the largest discrepancies between the TOFE presentation and the one breaking down the envelopes by ministries, arbitrage provisions, and common charges. During the validation mission, members of the TOFE team recognized informally and off the record that they were not comfortable using GFS 2001 despite receiving technical assistance and trainings in the past. As Burundi is going through a profound reform trying to move from GFS 1986 to 2001, this may be an issue in other areas of the financial management sector. Box 9 Practicing the Process of Engagement: the joint validation of the MTEF assessment 1. In the context of the PER validation process, the Bank team met with Government officials to understand the reason beyond the data issues reported above. Discussions with members of the MTEF team suggest that the change in the methodology might be a leading factor explaining the inconsistencies. The first exercise produced one single envelope which had to be distributed among ministries. 2. Starting in 2011, a new methodology was introduced to add a simulation component and provide more flexibility at the moment of allocating the envelopes at the ministry level. Three simulations per ministry are done simultaneously and then an EXCEL sheet picks up the simulation with the highest value so long as it stays within the budget envelope. Arbitrage provisions are then calculated as a residual. The reason behind the adoption of this methodology is to provide policymakers with some room to allocate the medium term projections for CSLP-II axes, which are the pillars around which the Government structures its policy actions. This method to compute the simulation creates a large Excel file which is hard to manage: the size and complexity of this file often results in data entry errors. 3. Regarding the differences with the TOFE, it is worth observing that the MTEF 2012-2014 and the MTEF 2013-2015 are the first attempts to produce the TOFE in GFS 2001 format. Given the low capacities within the MTEF Team, it is likely that the process to put the data in TOFE format originated many data entry errors. 4. It is clear that the MTEF file needs an audit to review the simulation formulation and, possibly, a simplification that better aligns the needs and capacities of the public officials in charge of using the file for MTEF formulation purposes. Also, training on GFS 2001 might be needed for the MTEF team, as this will reduce the risk of inconsistencies with the TOFE and proper replication of its format. 93 53. The quality of the MTEF in Burundi has deteriorated over time, raising concerns on the suitability of the tool’s use. The results from the analysis suggest two major problems: (i) there are significant inconsistencies within the MTEF document; and (ii) there are inconsistencies between the MTEF and the estimates shown in the budget framework letter (lettre de cadrage). 90 Figure 4.11 below illustrates the first case, showing the magnitude of differences between the amount presented as total by ministries and the amounts obtained adding the ministries one by one. While being negligible Figure 4.11: Inconsistencies within the in 2011, they became quite considerable over MTEF document time, raising concerns about the capacity of the 0.7% Burundi authorities to adopt MTEF. Interestingly, (Percentage difference) 0.6% for the last MTEF, the differences decrease as the 0.5% 0.4% time horizon expands. 0.3% 0.2% Y1 0.1% 54. Technical discussions with Government 0.0% Y2 officials helped to identify the major factors -0.1% Y3 accounting for these inconsistencies. In large part, they seem to derive from the changes in the methodology used to produce the MTEF and the complexity of the Excel based formulas used to produce these documents91. To some extent, this Note: the figures shows inconsistencies between the total by ministries presented in main table and the deterioration is also the byproduct of high total from adding each individual ministry. turnover in the MTEF team and the failure to Source: Bank staff own calculation with data from retain critical staff once trained (training was MTEF. provided under EMSP). 55. The budget approved by Parliament is quite different from the MTEF estimates, and this is also true for budget execution rates across ministries. This suggests that MTEF is not yet used directly as a budget planning tool, and Burundi has not yet reached a stage where MTEFs materially impact the expenditures. This is not surprising and reflects a larger trend in African fragile states which have found difficult to develop MTEF when the macroeconomic framework is inherently unstable and sources of financing (including external financing) are volatile and unpredictable. 90 Annex G provide additional background data and details on this analysis. 91 The Excel file is set up in such a way that three simulations per ministry are done simultaneously and then a sheet picks up the simulation with the highest value so long as it stays within the budget envelope. Arbitrage provisions are then calculated as a residual. The reason behind the adoption of this methodology is to provide policymakers with some room to allocate the medium-term projections for CSLP axes, which are the pillars around which the Government structures its policy actions. While this argument makes sense intuitively, creating these simulations in an Excel environment becomes extremely complex and makes data entry errors both more likely and harder to trace. 94 56. Utilization of MTEF for the 2013 Figure 4.12: Percentage difference between budget displays deterioration. A expenditures estimates and approved budget93 comparison of the envelope produced by the MTEF with the budget shows that the estimates for the first MTEF (2011-2013) have been generally close in total to the budget approved by the Parliament (0.4 percent difference in 2011). However, the differences have increased over time, reaching up to 18.3 percent difference for the current budget year. 92 Figure 4.12 illustrates the magnitude of the problem. Source: MTEF’s and approved budget data. 57. A comparison of the MFEF and Figure 4.13: MTEF Estimates vs. Approved the budget by expenditure categories Budget (current expenditure) shows that the largest differences are for 10% capital expenditures. This is not surprising, (Percentage difference) 5% and to some extent reflects the difficulties of 0% forecasting donor’s contributions. -5% Interestingly, the differences are smaller in Y1 -10% the MTEF for 2011-2013. Burundi’s -15% Y2 reliance on donors’ support suggests the Y3 utility of separate forecasting to reduce deviations and improve the credibility of the MTEF as planning tool. Source: TOFE. 92 Ideally, a second phase of the exercise would compare the envelopes with the budget execution data as well. However, this part of the exercise was not performed for two reasons: (i) at the aggregate level, the execution figures for external financing included in the Loi de Règlement, are unreliable; and (ii), the execution data at the ministry level extracted from SIGEFI is unreliable. It is not possible to extract execution data at the ministry level from any other source other than ministries themselves. 93 Expenditures excluding debt servicing: For the MTEF, the figures shown correspond to the manual addition of the ministries plus charges communes, plus arbitrage provisions. For the budget, the figures are the total expenditures including dépenses des fonds and tax exemptions. Though the numbers do not match logically (MTEF does not forecast either dépense des fonds or exemptions), this comparison is shown for purpose of completeness. 95 58. Technical discussions with Government officials helped to identify the major factors Figure 4.14: MTEF Estimates vs. Approved accounting for these inconsistencies: First, the Budget (capital expenditure) budget process is political and considerations 0% (Percentage difference) other than macrofinancial stability are taken into -10% account during budget preparation phase. Second, -20% the budget is overly complicated and includes -30% Y1 several budget lines which do not a have -40% straightforward equivalence in the MTEF. 94 -50% Y2 Third, funds provided by donors remain Y3 unpredictable and their disbursement difficult to estimate as part of the budget preparation process. In all events the inconsistent and fitful application of the framework is a considerable concern. Source: TOFE. 59. The implications of non-application require to be addressed head on; MTEF must have demonstrable efficacy if investment in its installation is to have any proper return. At minimum, a full independent audit of the Excel file used to produce the MTEF estimates is needed to correct internal inconsistencies. Differentiation between domestic and external financings should also be reintroduced as good practice to improve the quality and credibility of the MTEF. More broadly, it needs to be determined if even a concentrated technical assistance package for effective operationalization would build the necessary levels of competence in the system, given Burundi’s current institutional environment. A legitimate alternative to be considered - in collaboration with the Government - might be to move toward sectoral MTEFs in targeted ministries where preconditions (including a clearly defined sector strategy) are in place and are likely to yield better results. 94 For instance, the Burundi budget includes a budget line called “dépense des fonds,” which is used to fund the construction of roads across the country, among other expenditures. The MTEF does not produce an estimate for this and other budget lines such as the education fund. 96 Box 10 Does the Walk Match the Talk? A reflection on MTEFs experiences across developing countries In recent years, the donor community has emphasized the need for African and low-income countries to have a Medium-Term Budgetary Framework (MTBF) and a Medium-Term Expenditure Framework (MTEF) as a planning tool conducive to macroeconomic and budgetary stability. Despite these recommendations, and growing levels of financial resource and technical assistance to support the elaboration of both MTBFs and MTEFs, the record has been uneven. Leinert (2007, 56) made the case that, in general, the MTEF exercise is “largely mechanical, without any impact on the size of the annual budget or allocation between spending programs,” and as a result “the MTEFs prepared in African countries have only a marginal influence on macroeconomic stability and allocative efficiency, and probably no influence on the efficiency of spending policies.” This is consistent with the observation by Brumby and Dorotinsky (2010) that MTEFs have often become a means to “present an unrealistic budget, rather than to provide a dynamic tool” and that at the end it “never changed budget behaviors: MTEF unit were busy developing new processes, but annual budget still made in the traditional way.” Specifically, the following constraints to produce high quality MTBFs and MTEFs in Africa can be identified: • African countries are under pressures by donors, who advocate the adoption of MTEFs, building on medium-term budget frameworks (MTBFs). • It is highly challenging to develop MTEFs when the macroeconomic framework is inherently unstable, a result of narrowly based economies, high dependence on exports and imports whose prices fluctuate widely, and, in several cases, political and institutional instability. In view of this, it is perhaps not surprisingly that no African country has produced an effective, operational MTEF, i.e., one in which the Y+2 estimates made in Y+0 are systematically used as the starting point in Y+1 for preparing the annual budget for Y+2. • Lack of capacity at the Ministry of Finance and the line ministries, which means that both the MTBF and the MTEF will have a low quality by definition. Many African countries have made efforts to introduce a medium-term perspective to budgeting, but very few have reached a stage where MTEFs materially impact the expenditures (inclusive of donor-financed spending) of annual budgets adopted by parliaments. Most of these factors are present in Burundi to a large degree. And in fact, during the data and analysis validation process, technical counterparts in the MTEF highlighted those as recurrent problems during the preparation of the lettre de cadrage and the MTEF. Source, Liener (2007); Brumby and Dorotinsky (2010). 97 Chapter 5: Confronting Competing Pressures for Reform: a case study of the wage bill and pay policy Numerous structural, process, and capacity weaknesses in the public service constrain performance. Although the size of the wage bill is within the normal range for the region, there is little fiscal space for expanding it. The Government has recognized the need for comprehensive reform and has adopted a National Program for Administrative Reform (PNRA), endorsed by the Council of Ministers. In the short term, a more pragmatic approach to improve performance may be to stimulate a targeted, incremental reform of key functions within Government departments and agencies. Immediate steps should also be considered to strengthen human resource management functions. 5.1 The Scope of the Problem 1. There are currently no coherent policies that guide pay increases, the size of the public workforce, or job promotions in Burundi. The current policy on annual pay increases has not been adjusted since 2006, which means that the real wages of new entrants is significantly less due to inflation. At the same time, there is no centralized authority that evaluates the optimal size of the public payroll, which has increased by nearly 20 percent since 2007. Furthermore, job promotions are solely dependent upon educational attainment and years of service, without any consideration of performance. 2. Although the size of the wage bill is within the normal range for the region, there is little fiscal space for expanding it. The wage bill accounts for nearly half of the recurrent budget, and has increased in recent years. As part of the peace deal, many soldiers were added to the payroll. Concurrently, the Government has added teachers, nurses, and doctors to boost the prospect of achieving the Millennium Development Goals (MDGs). Wage expenditures, consequently, have outgrown GDP, and if this trend continues, drastic measures may have to be taken to reduce wage bill growth. 3. The Government has recognized the need for reform and has adopted a National Program for Administrative Reform (PNRA),95 which has been endorsed by the Council of Ministers. The PNRA calls for rationalizing the civil service, promoting greater accountability, and strengthening dialogue between the Government and unions. A pay reform roadmap has also been recently developed and discussed with Government officials, which contains a proposed timeline to implement a simplified and unified pay structure. 95 Government of Burundi (2012c). 98 4. Experience from other low-income countries, however, has shown that ambitious civil service reforms that seek to transform civil service management have often failed to match expectations. In addition, across the board pay harmonization exercises are typically met with strong resistance from unions and politicians, and do not necessarily improve performance. Given the difficulties in attaining comprehensive civil service reform, several countries have experimented with an asymmetric approach. This approach establishes more modest goals and adopts an incremental approach to tackle pay and performance problems, both in reform measures and in application to government agencies (with scope for expansion over time). Although every country has adopted a unique approach to this reform, all asymmetric reforms have linked higher pay to the introduction of meritocratic human resource management (Nunberg and Taliercio, 2012). 5.2 Recent Developments in Government Wages and Employment 5. Burundi’s general government wage bill96 is slightly above the 2011 Sub-Saharan Africa average of 7.4 percent of GDP, but within a normal range (see Figure 5.1). 97 It further decreased in 2012 to eight percent of GDP and the latest IMF projections have it gradually decreasing to seven percent in 2017. The true size of the wage bill, however, is likely to be slightly higher since the official statistics do not capture the full size of the public sector workforce. Figure 5.1. Sub-Saharan Africa Wage Bills, 2011 (percent of GDP) 20 18 16 14 12 Average: 7.4 10 8 6 4 2 0 Burkina Faso Sierra Leone Côte d'Ivoire Seychelles Benin Madagascar Mauritius Senegal Comoros Malawi Mozambique Burundi Uganda Nigeria Kenya Rwanda Eritrea Central African Rep. Guinea Zambia Ghana Angola Botswana Liberia Namibia Equatorial Guinea Congo, Rep. of Ethiopia Tanzania Congo, Dem. Rep. of South Africa Togo Lesotho Chad Cameroon Guinea-Bissau Gabon Gambia, The Cape Verde Niger Mali São Tomé & Príncipe Swaziland Zimbabwe Source: IMF (2012). 96 Throughout this paper, the wage bill is defined in terms of general Government wage expenses. Pensions are not considered. 97 This finding, however, should be interpreted with caution since the regional wage bill average does not necessarily imply good practice, only the existing situation. Also, differences in accounting and reporting standards may make the figures not strictly comparable across countries (World Bank, 2010). 99 6. Within Burundi, the size of the wage bill has been volatile in recent years (see Table 5. 1). It reached a high of nine percent of GDP in 2009, before falling back to eight percent of GDP in 2012. This volatility has been driven by changes in the size of the civilian workforce. At the end of the civil war, many security personnel were added to the wage bill, which was necessary to maintain the peace and stabilize the country. Although the conflict officially ended years ago, the security sector remains a significant element of the wage bill (see Box 11). Wage expenditures on the military and national police, which collectively account for a third of total expenditures, have remained relatively constant since 2008.98 Table 5.1 Size of the Wage Bill (percent of GDP) 2007 2008 2009 2010 2011 2012 Compensation of employees 7.8 8.3 9.0 8.6 8.7 8.0 Civilian 4.6 5.7 6.3 6.5 6.1 5.7 Military 1.9 1.5 1.6 1.3 1.5 1.4 National Police 1.2 1.0 1.0 0.9 1.0 0.9 Source: IMF (2012). Box 11 Managing the Security Sector Wage Bill in a Post-Conflict Environment Since the Arusha Agreement signed in August 2000, Burundi has experienced both great strides and major setbacks on the road to lasting peace. Following the adoption of a new constitution in February 2005, and presidential and parliamentary elections in August 2005, a ceasefire accord was reached in 2006 with the last rebel movement, the Front National pour la Libé ration (FNL), but sporadic fighting continued. In 2008 and 2009, FNL renewed its commitment to negotiated peace, and was reintegrated into the political system. Disarmament, demobilization, and reintegration efforts have continued over recent years. Burundi's second democratic elections in 2010 represented a milestone in its political transition despite the boycott decided by a coalition of opposition parties who alleged fraud and vote rigging. The largest scope for creating fiscal space in Burundi is through reallocation of expenditure from non-priority to priority sectors as well as through improvement in the spending efficiency. Specifically, the ongoing process of demobilization/downsizing of former ex-combatants (now members of security forces) is viewed as the main element of this strategy. The Government should only choose a pace of demobilization/downsizing, however, that offsets the marginal benefit of increased pro-poor and growth enhancing expenditures with the marginal cost of the greater threat to security. In other words, the rate needs to be adjusted to the private sector capacity to absorb demobilized workers into productive activities. Source: World Bank (2012). 98 Military salaries, which accounted for 46.1 percent of total wage bill in 2001, fell to 17.5 percent in 2012. 100 7. Since 2007, the wage bill has grown faster than GDP, though the gap narrowed in 2012 (see Figure 5.2). From 2007 to 2009, real growth of the wage bill had increased by 16 percent more than real GDP. The gap narrowed recently, however, as the Government instituted a hiring freeze in 2012 for all sectors except education, health, and justice to restrict the growth of the wage bill. 8. In terms of total recurrent Figure 5.2: Wage Bill vs. GDP real growth expenditure, the wage bill accounts for (2007=1.0) almost half the budget (see Table 5.2). In 1.50 2012, it reached 46.1 percent of the recurrent 1.40 budget, its highest level in the last six years, which means that there are fewer resources 1.30 available for other expenditure items, such as 1.20 textbooks, medical facilities, and medicines. This level, however, is still within the normal 1.10 range for comparative countries in the 1.00 region.99 0.90 2007 2008 2009 2010 2011 2012* wage bill Source: IMF (2012); *estimate. Table 5.2 Size of wage bill in recurrent budget (percent) 2007 2008 2009 2010 2011 2012 Compensation of employees 43.6 44.2 45.4 43.8 44.7 46.1 Civilian 25.9 30.5 32.1 33.0 31.6 32.9 Military 10.7 8.2 8.1 6.5 7.9 8.2 National Police 7.0 5.5 5.2 4.4 5.2 5.0 Source: IMF (2012). 9. The growth in the size of wage bill expenditures has mostly been driven by expanding the size of the public sector workforce, as remuneration policies have remained largely unchanged since 2006 (see Table 5.3). From 2007 to 2012, staffing increased by nearly 20 percent and the total public workforce expanded by 20,000 employees. Most of the growth has occurred in the education sector as the number of students increased dramatically after primary school fees were abolished in September 2005. Currently, teachers comprise nearly half of the workforce, while the military and police comprise approximately one third of the workforce. The remaining non-teaching civil servants are mostly comprised of health workers. Although the size of the military has not changed significantly since 2007, the size of the police force has fallen by 2,000 individuals. 99 The recurrent budget is used since donors provide the majority of financing for the development budget. 101 Table 5.3 Public Sector Workforce 2007 2010 2011 2012 Teachers 37,750 49,733 55,467 61,789 t Non-Teachers 21,079* 9,896 9,181 9,786 Police 18,626 17,788 16,704 16,721 Military 28,108 29,237 28,357 28,001 Contract -- 9,432 9,129 8,790 Magistrates -- -- 1,351 1,436 Total 105,563 116,086 118,838 125,087 * contains contract employees, magistrates, and others; t contains health workers Source: World Bank (2008); Garnett, Nganou, and Brown (2012) 10. In Table 5.3, staff levels listed, do not represent the full size of the public sector. Some public employees in health, education, and other select ministries are paid directly from resources generated in their sector, and thus their salaries are not reflected in the official payroll. In addition, the wage bill does not include transfers to semiautonomous institutions, such as the University of Burundi and hospitals, which are often used to pay salaries (World Bank, 2008). 11. A comprehensive census of the public payroll in April 2008 was successful in eliminating nearly 2,000 ghost workers from the payroll. It also identified, however, legitimate staff that were working in the public administration, but who were not formally on the payroll. Many of these individuals were new recruits who have yet obtained or presented the required certifications for their employment (World Bank, 2008). 5.3 Complexities and Inequities of Public Pay 12. Public pay structures provide nearly 3,000 distinct pay possibilities. There are six classification schemes in total, one each for civil servants, judges, military, police, contractors, and autonomous institutions (see Table 5.4). These schemes, however, only cover base salaries, while allowances and bonuses, often significant contributors to total compensation, are separate. In the general civil servant scheme, where the majority of employees are located, there are 2,700 distinct base salary rates. The grid is composed of three categories: direction (high), collaboration (middle), and execution (low). Each category, in turn contains eight sequential grades (1 low, 8 high). In practice, however, an additional grade (9) was introduced in 2007 to placate teacher unions. Furthermore, there are 100 distinct levels within each grade (1 low, 100 high), though few individuals reach the higher levels before progressing to the next grade. Table 5.4 Classification Schemes in Burundi, 2012 Classification Scheme Employees Pay Levels General Civil Servants 76,035 2,700 Magistrates 740 14 Military 27,619 26 Police 16,256 20 Contractors 8,691 N/A Autonomous Agencies unknown N/A Source: Government of Burundi (2012), data from June, 2012. 102 13. Base salaries are determined solely by educational certificates, not by job responsibilities. Each of the 2,700 distinct salary rates assigns an index value from which the base salary is determined. The index value is then multiplied by a fixed amount of 154.1 FBu, which has been in place since 2006.100 Table 5.5 illustrates the different possible index values and monthly base salaries for the lowest and highest position in each category. Within each category, there are wide variations in possible pay, although actual salaries tend to be much more compressed. Table 5.5 Civil Servant Classification – Pay Structure, 2012 (FBu) Category Grade Level Index Base Salary I 1 348 53,627 Direction IX 100 11,651 1,795,419 I 1 178 27,430 Collaboration IX 100 8,646 1,332,349 I 1 119 18,337 Execution IX 100 6,086 937,852 Source: Government of Burundi (2012), monthly data from June, 2012. 14. Magistrates, the military, and police have their own unique pay classification systems. For magistrates, there is a 14 point grade scale,101 the military has a 26 point grade scale, while the police has a three category system with 20 different pay possibilities. Table 5.6 presents the extreme base salary possibilities in each category. Similar to the civil servant pay grid, there is significant differences between base salaries at the lowest level and at the highest level. Contract workers and employees of autonomous agencies do not adhere to any pay grid. Table 5.5 Magistrates, Military, and Police pay grids, 2012 Magistrates Military Police Base Base Base Grade Grade Grade Salary Salary Salary 2* 288,768 1 650,850 Head of Police 504,486 14 33,396 26 19,252 Recruit 3,785 * Grade 1 for Magistrates is currently not used. Source: Government of Burundi (2012), monthly data from June, 2012. 15. The structure of the public workforce is concentrated in the collaboration category (see Table 5.7). Nearly 87 percent of public employees are found within this category, while less than 5 percent of staff are found in the direction category. Similarly, the majority of staff is contained within the first four grades of each category. More than 54 percent of staff is classified in grade two, and approximately 91 percent of staff is classified within the first four grades. Thus, the majority of staff is concentrated in the lower positions of the pay grid. Table5.6 Structure of the Workforce, 2012 Execution Staff Collaboration Staff Direction Staff Percent EI 2,696 AI 2,285 CI 1,222 9.98 EII 980 AII 31,981 CII 916 54.48 EIII 728 AIII 9,729 CIII 240 17.20 100 From 2001 to 2006, 134 FBu was used as the index value. 101 In practice, however, there are only 13 grades as grade 1 is currently not used. 103 EIV 683 AIV 4,682 CIV 182 8.92 EV 304 AV 2,225 CV 113 4.25 EVI 76 AVI 1,431 CVI 60 2.52 EVII 26 AVII 886 CVII 30 1.51 EVIII 25 AVIII 544 CVIII 6 0.92 EIX 14 AIX 119 -- -- 0.21 Total 5,532 Total 53,882 Total 2,769 -- Percent 8.90 Percent 86.65 Percent 4.45 100.00 Source: Government of Burundi (2012), data from June, 2012. 5.4 Inequities in Wages 16. A major source of job dissatisfaction in the public sector is inequitable pay, which undermines work motivation. Substantial inequities exist in the remunerations of different occupational groups. Teachers and some health workers receive significantly more pay than other civil servants with similar qualifications. For example, a newly recruited teacher is paid 50 percent more than another new civil servant recruit with equivalent education, though this cannot be justified by market pay levels (Garnett, Nganou, and Brown, 2012).102 17. The overall average monthly salary in 2012 was 150,000 FBu (Government of Burundi, 2012). Non-teachers (mostly health workers) and magistrates earned, on average, significantly more than the average salary, while the military and contract workers earned significantly less (see Table 5.8). Average salaries increased across the board from 2010 to 2012, and in some cases, at substantial rates. The average salary of non-teachers, police and the military, grew by more than a 50 percent during this period, while the average salary of magistrates increased by 21 percent from 2011 to 2012. Table 5.7 Staffing and Salaries, 2010-2012 (FBu) Average pay per Average pay per Average pay per Growth Category month 2010 month 2011 month 2012 2010-2012 Teachers 150,651 174,409 163,293 8 Non-teachers 159,318 288,908 278,905 75 Police 86,061 111,495 148,890 73 Military 77,386 82,702 115,732 50 Contract 66,322 67,959 70,132 6 Magistrates -- 288,298 349,975 21 Source: Garnett, Nganou, and Brown (2012). 18. Although public salaries are high, on average, compared to per capita income, they are low in absolute terms. The average monthly salary of 150,000 FBu translates to approximately 96 U.S. dollars, while monthly consumption per capita in 2011 was estimated by 102 Primary teachers are upgraded from the execution to the collaboration category in order to make this pay difference possible within the pay classification grid (Garnett, Nganou, and Brown, 2012). 104 the national statistical agency to be 83,248 FBu. Given that the average household contains five individuals (according to the latest census), the average wage level is relatively low (Government of Burundi, 2012). 19. To reform the pay scale, it will be necessary to have sufficiently high remuneration to attract and retain certain professions (see Table 5.9). In particular, the wages for doctors and IT specialists are too low, which has led to shortages in these occupations. In contrast, many international NGOs pay doctors up to ten times more than the average government compensation (Garnett, Nganou, and Brown, 2012). Table 5.8: Monthly Remuneration for Select Jobs, 2008 (FBu) Entry Level Job Category Bonuses Allowances Total Base Salary Doctor 55,000 27,500 85,000 167,000 Engineer 50,000 17,000 59,700 127,000 Secondary teacher 59,000 7,000 39,700 106,000 Nurse 32,000 16,000 57,200 105,000 Economist 44,000 0 59,700 104,000 Administrator 44,000 0 59,700 104,000 Primary teacher 38,000 7,000 39,700 85,000 Source: Garnett, Nganou, and Brown (2012). 20. Autonomous agencies have significantly higher pay (see Table 5.10). Public employees in these institutions earn, on average, significantly more than public workers, especially in the two higher categories. Although the organic finance law (LOFP) and article 136 of the October 18, 2011 decree, provides for the full integration of autonomous institutions into the general government by January 1, 2017, it is recommended that this policy be reversed (Government of Burundi, 2012). Certain autonomous agencies, such as the Burundi Revenue Administration (OBR) have had high organizational performance in recent years in part due to higher pay flexibility, which has allowed it attract and retain highly skilled professionals. Table 5.9 Comparison between General Statute and Autonomous Agencies, 2012 (FBu) CATEGORIES General Statute Autonomous Difference (%) Agencies Execution 115,371 143,867 +24% Collaboration 131,675 298,850 +127% Direction 249,092 989,402 +297% Source: Government of Burundi (2012), average monthly data from June, 2012. 21. Political appointments to the civil service earn significantly more than the average public sector worker (see Table 5.11). Ministers, chiefs of staff, policy advisors, and permanent secretaries earn six to 13 times more than the average civil servant. There is, however, a high turnover of these positions as they are outside of the normal civil service statue. Table 5.10 Monthly Compensation of Political Appointments, 2012 (FBu) 105 Category Salary Chiefs of Staff in President’s Office 1,972,200 Deputy Chiefs of Staff in President’s Office 1,947,000 Ministers and Secretary General 1,770,400 Chiefs of Staff in the Vice President’s Office 1,696,000 Deputy Chiefs of Staff in Vice President’s Office 1,669,600 Key advisors to the President 1,669,600 Key advisors Vice Presidents 1,669,600 Advisors to the President, Vice President, and Secretary General 893,100 Permanent Secretaries of Ministries and Secretary General 893,100 Source: Government of Burundi (2012), monthly data from June, 2012. 5.4.1 The proliferation of bonuses and allowances 22. In addition to the base salaries, a variety of bonuses and allowances constitute a significant component of the compensation package. In some cases, these benefits are higher than the base salary, and some of them have the extra advantage of being exempt from taxation (up to 65 percent of the base salary; Income Tax Law, 2013). Bonuses and allowances have been used with increasing frequency in recent years as a means to “institutionalize” wage increases since the base salary is fixed at the 2006 nominal rate. They are awarded for housing, transport, motivation, risk, and variety of other factors, but none are linked to performance (see Annex D for a list of the main bonuses). In principle, the judiciary, the army, and the national police have their own bonus system, while ad hoc bonuses can be created at any time by autonomous agencies. 23. The distribution of benefits is uneven across ministries (see Figure 5.3). On average, benefits account for twenty-five percent of compensation across ministries, but several ministries, such as the Ministry of Health, the Presidency, and the Ministry of Justice, receive a greater share of benefits than base salaries in their total compensation. 106 Figure 5.3: Compensation by Ministry, (percent; 2012) Health Presidency Justice Vice Presidency 2 Vice Presidency 1 Higher Education Public Service Foreign Affairs Interior Commerce Finance East African Community Good Governance Development Youth Primary Education Energy Communications Transport Agriculture Public Security National Solidarity Water Defense Parliament 0 20 40 60 80 100 Base Salary % Bonus % Family Allowance % Source: Government of Burundi (2012). 24. The multiplicity of bonuses and allowances compensates for the rigidities in the pay structure, but also weakens the established hierarchy. In addition, they make the pay system less transparent, and provide scope for authorities to add new benefits at their discretion, which generates additional pay inequities across the public sector. In order to comprehensively reform the pay structure, the across the board benefits should be integrated into a revised classification system, while the remaining ones should be consolidated and redesigned to reward performance. 5.5 Governance and Institutional Arrangements 5.5.1 Personnel management is not strategic 25. Institutional arrangements constrain the application of effective human resource management. Currently, this responsibility belongs to a directorate within the Ministry of Public Service, Work and Social Security, which used to have ministry status before it was merged with the Ministry of Work and Social Security by a presidential decree. By granting the public service less authority and stature, the decree has weakened the government’s human resource management function. 107 26. The Directorate of Public Service does not have trained staff to administer its functions. Garnett, Nganou, and Brown (2012) found that there is no evidence that it is able to carry out any of the first three (of five) functions that are defined in the decree: (i) assist with the development and implementation of government policy regarding the management and financing of human resources; (ii) ensure that human resources are allocated strategically and efficiently; (iii) coordinate human resource management across government ministries; (iv) co-manage the centralized payroll database with the Ministry of Finance; and (v) submit appeals to the Administrative Court on behalf of public servants. Currently, the Directorate’s main activity is the administration of personnel transactions (recruitment, promotion, payment) for decisions that already been taken in other Government agencies (Garnett, Nganou, and Brown, 2012). 103 If personnel decisions continue to be fragmented across Government, strategic management of the public service will not be possible. 27. None of staff in the Directorate have received professional training in human resource management. Garnett, Nganou, and Brown (2012) speculate that this situation is also probable for the human resource departments in the line ministries. Since it is unlikely that sufficient expertise can be attracted from the private sector, human resource management skills will have to be developed through on-the-job-experience. To this end, Garnett, Nganou, and Brown (2012) recommend that staff from both the Directorate and line ministries undertake a customized training course in human resource management, with additional on the job training provided. 28. In the current system, recruitment occurs on an ad-hoc basis. Although there is a National Recruitment Commission, there are no established personnel levels set for ministries, and thus, the recruiting is not done in accordance with an official organic framework. According to the information available at the time this study was drafted, the recruiting of additional staff can be done only in the ministries of primary and upper education and health. For all the other ministries, recruiting is only authorized to replace departing staff. Furthermore, many appointments to the public service continue to be based on political and ethnic criteria, rather than merit. The Burundi authorities explained this situation as part of the will to rebalance the social interactions inherited from the conflict period. While these are legitimate political concerns, to build a high quality public administration able to support the implementation of an economic development agenda, the Commission should be reestablished as the central agency responsible for executing government recruitment under a transparent and competitive process based on merit (Garnett, Nganou, and Brown, 2012). 5.5.2 Career progression is not tied to performance 29. Career progression is solely dependent upon seniority, irrespective of job function. All new entrants to the public service begin in either grade I or II. As promotion to the next grade is automatic after the six years of “very good” appraisal ratings, an entrant can rise to grade VII after 36 years, while a teacher which began in grade II can rise to grade VIII during the same period. Very few employees pass through the full one hundred levels before accessing the next grade. The best employees are not rewarded by this system, and the least productive employees 103 Its authority has also been undermined by a new remuneration department that was established within the Ministry of Finance. 108 are not hindered in their progress. Career progression is thus completely based upon length of service, without any consideration of performance or capability, and the higher grades are solely used as a means for salary progression. Since there are no limits on the upper number of positions, this passive system of career management is unlikely to encourage motivation for high performance. 30. The current performance appraisal system does not adequately evaluate performance. Public servants are appraised only once a year by their immediate supervisor. This evaluation does not require, however, any discussion between the employee and the supervisor, though the employee must sign the completed appraisal form. The majority of employees are rated to be “very good” as most supervisors seek to avoid damaging their employees prospects of salary progression. 104 In addition, the current criteria are highly subjective and serve little purpose other than a providing a predictable salary increase (Garnett, Nganou, and Brown, 2012). While the development of an objective and respected performance appraisal system should be viewed as a long-term process, institutionalizing a more regular engagement between staff and management could engender significant performance improvements. 31. There are no resources available for training in the public service. Although the Public Service Law entitles training to every public servant at least once every five years, this practice has not been followed. Training is completely financed by donor agencies, and restricted to certain priority areas. While it is unlikely that the public sector will be able to increase its skill set by hiring from the private sector, internal training is not likely to generate improvements in performance without a reformed career development pathway (Garnett, Nganou, and Brown, 2012). 5.6 Management of Pay Policy 32. There is no coherent policy that governs pay increases, no effective limit on salary progression, and pay is not tied to job responsibilities. Base salaries continue to be set at the 2006 level of 154.1 FBu per index value, which means that they have not kept pace with inflation. Consequently, new entrants into the public service receive a significantly lower real wage than entrants from previous years ago. Salary progression has instead occurred through automatic promotions, proliferation of both bonuses and allowances, and concessions awarded to striking unions. 33. Nominal wage increases occur automatically across the majority of the public service (see Table 5.12). In the performance appraisal system, public servants are rated either “elite” (four levels; eight percent increase); “very good” (three levels; six percent increase); or “good” (two levels; four percent increase) each year. Thus, the typical employee in the general statute is promoted three levels each year, corresponding to an average annual increase of six percent in their base salary. For magistrates, there are two additional requirements: have at least three years in one grade and have been rated “very good” at least once during this time. A percentage increase is also added to base salary for each year of service according to the 104 If a public employee receives two consecutive “insufficient” ratings or one “mediocre” rating, they may be dismissed, though this rarely occurs in practice. This rating may also be appealed to the Administrative Court within fifteen days of its receipt. 109 notation: “good,” two percent; “very good,” four percent; “elite,” six percent. For the military and police, promotions are only possible if there is a vacancy available. From this point of view, these systems are much more consistent with standard practice, since there is a direct relationship between rank and job responsibility. Contract workers receive annual increase of three percent per year, while there is no system governing pay increases for autonomous agencies. Table 5.11 Annual Wage Increases and Promotions, 2012 Type Annual Wage Increases Promotions General Statute 3 levels per year grade every 6 years annual increase 2% "very grade if rated very good at Magistrates good", 4% "elite" 6% least once in last 3 years Military annual increase 2% "very only possible if vacancy at Police good", 4% "elite" 6% particular grade Contract 3 % per year N/A Autonomous Agencies N/A N/A Source: Government of Burundi (2012). 34. A formal mechanism to review annual wage increases does not exist. Seven years have passed since the latest revision to the base salary index value. In the medium-to-long run, annual (non-promotion) pay increases should correspond to annual revisions to the index value. Countries have adopted different approaches, ranging from salary review commissions, index linked salary adjustments, cost of living adjustments, market benchmarking pay adjustments, and collective bargaining agreements. Currently, strikes by powerful groups have been used to extract pay increases, and since settlements are reached with one group at a time, the annual budget process is completely disrupted. 35. Official payroll data has been moved to an improved ICT database with World Bank support, which has the potential to improve payroll management. The Government has recently established a joint center (Ministry of Public Service, Work, and Social Security, and the Ministry of Finance) for payroll processing and career management, and the database was officially launched on March 29, 2013. The development of the database entailed the manual imputation of more than 80,000 paper records and is a significant achievement. It assigns a unique identification number and automatically calculates the base salary, bonuses, allowances, and years to retirement for each public sector employee. The new system is expected to be used by the end of this calendar year, but it is not yet clear whether it will be used to assist with the strategic management of human resources, or simply as a digital record system. 5.7 Realistic and Feasible Reform Options: a proposal 36. The Government of Burundi has recognized the need to reform its wage bill management practices by adopting the PNRA. The PNRA calls for a return to management principles of fairness, transparency, and performance for public employees, and documents many of the challenges discussed in this paper. It also recognizes that reforms must be properly sequenced and prioritized as there are too many issues to tackle at once. Consequently, a road map with a prioritized set of reforms to harmonize the salary structure has been endorsed by the Ministry of Public Service, Work, and Social Security (Government of Burundi, 2012b). The road map is the result of several workshops between the Ministry, UNDP, and public unions. In particular, it calls for a simplified and more transparent pay classification structure, in which the 110 one hundred levels associated with each grade are eliminated, and for the majority of bonuses and allowances to be integrated into the base salary. 37. The integration of benefits into the base salary has generally been accepted by the government and unions, but it will be difficult to resolve wage inequality without increasing the overall size of the wage bill to an unsustainable level. If the one hundred levels are eliminated, the range of salaries within a grade will have to be normalized, which by default means that some employees will have a pay increase, while others will have a pay decrease (if the reform is to be cost neutral). It is not likely however, that the reform will be accepted by individuals who earn a salary above the grade average if they are forced to accept less pay (even a freeze on wages of these individuals was not acceptable to some unions). Furthermore, in the lead up to the elections scheduled for 2015, the Government will likely seek to minimize tensions with the unions. There will also likely be additional pressures for recruitment in the education and health sectors as the Government tries to meet its MDGs and support its free primary education policy. 38. Given that a cost neutral pay redistribution will likely encounter strong resistance, the Government of Burundi (2012) report suggested four adjustment scenarios, each entailing a substantial fiscal cost (see Table 5.13). Scenario Two has the least adjustment cost, where public employees earning below the average in a particular grade would see their salaries increased to the average, and those earning above the average would not be affected. This scenario, however, would require an additional annual cost of 0.50 percent of GDP, which would extract additional resources from the budget for the same level of service delivery. At the same time, it is not clear that any of these scenarios would garner both political and trade union support. Table 5.12 Adjustment Scenarios, fiscal cost per year Adjustment Billion FBu % of GDP Scenario 1: Max Salary in each Grade 273 7.66 Scenario 2: Average Salary in each Grade, those 17 0.48 above average not affected Scenario 3: Donor proposed revised salary scale 87 2.44 Scenario 4: Combination of 2 & 3 18 0.50 Source: Government of Burundi (2012). 39. Even if one of the above scenarios is approved, there is no guarantee that it will actually lead to an improvement to the performance of the civil service. It will not have solved the fundamental problem of “equal pay for equal work” as individuals earning more than the average will continue to do so, and it will not permit higher wages to be paid to attract and retain both health professionals and IT specialists. At the same time, the deeper institutional deficiencies will remain. 40. Simplification of the pay structure should thus be pursued as a medium-term objective, with the full support of the unions, to find a fiscally sustainable solution that constrains aggregate wage bill growth. As part of this process, the Government should take a strategic decision regarding what type of public sector management it wants. A first step would be to more directly connect the proposed PNRA strategy to its second poverty reduction strategy 111 plan. It is also recommended that any new structure promote long term career based incentives, such as opportunities for long-term enhancement of reward and status, competitive promotions, and deferred compensation to improve incentives for high performance (see Hasnain and Manning, forthcoming). In practical terms, this would entail attaching job responsibilities, such as decision making and the difficulty of tasks, to each grade, which would boost performance by providing higher pay to those taking on more difficult tasks. The new structure would also allow new entrants to the public service to enter the new pay grid based on their qualifications, and would progress based on their skill or merit. Managers would be located at a higher level based upon their increased responsibilities. 41. In the short term, a pragmatic approach to improve the performance of existing staff may be to stimulate a targeted, incremental pay reform for key government departments and agencies. 105 Previous asymmetric civil service reforms have already been tested in Tanzania, Mozambique, Zambia, Bolivia, Afghanistan, and Cambodia. This approach, though departing from conventional approaches of public administration reforms by establishing a more complex and less homogeneous civil service, follows recent global trends by the OECD and others in this area (see Box 12). Hasnain and Manning (forthcoming) explain that this approach should prioritize “transformations and changes that would enhance the legitimacy of the center rather than undermine it, and which both exploit and nurture the appetite for reform that does exist within certain areas of the government administration.” Examples of this approach include diversified pay arrangements at the agency level, annual performance targets, and moves toward position based (as opposed to seniority based) systems that encourage lateral entry (Hasnain and Manning, forthcoming). Box 12 Pay Differentiation and Asymmetric Pay: defining the concepts Over the past fifteen years, a majority of OECD countries have moved away from the traditional pay model of uniform pay for jobs in similar grades and pay increases based largely on seniority. The changes introduced include performance related pay, some delegation of authority over pay setting to the entity level, and differentiation of pay scales between agencies and sectors. Middle-income countries, and to some extent low-income countries, possibly drawing on the OECD example, are also implementing these changes and have sought the World Bank’s advice to help them in the process. Differentiated pay setting can be indefinite (allowances for hardship or remote positions) or temporary, and it is in essence some combination of three key features, or design elements, which can be present in varying degrees: 1. Individualization, or performance related pay: Enabling pay to differ for civil servants doing the same job by linking a portion of the civil servants’ pay to individual or group effort or performance with performance targets set either at the individual level and/or the group level. 105 In a comprehensive review of 110 studies, Hasnain, Manning, and Pierskalla (2012) find that performance related pay had a positive effect in more than half of the studies. They further state that the results were concentrated in situations where the outputs or outcome were readily observable, such as education, healthcare, and revenue collection. There have been few studies, however that have examined the impact of the performance related pay in low-income countries. 112 2. Differentiation: Pay differences within and across government ministries, departments, and agencies for civil servants doing the same job, based on the need to attract and retain qualified staff for those jobs, or to persuade staff to accept new working arrangements. 3. Delegation: Transferring authority over pay setting and human resource management from a central civil service agency to ministries, agencies, or departments. The transfer may entail transferring from the central to the entity level (or a mixture): pay negotiations (if any); setting the overall wage bill; and design of pay scales. At root, flexibility is defined in terms of what it is not — it is not adherence to the caricatured ideal of how pay is determined in a Weberian bureaucracy, traditionally characterized by entry level hiring and secure tenure combined with internal promotions with wage increases based on a rigid formula driven by seniority (no individualization), uniform pay schedules across all ministries and departments (no differentiation), and strong control of pay levels by a central bureaucracy (no delegation). Individualization has been a significant trend in many OECD countries, and in some sectors in developing countries. Extracted from Hasnain and Manning (forthcoming). 42. An asymmetric pay reform would also avoid an egalitarian approach to pay reform, and instead shift the focus to performance of the civil service. Typically, as described by Olowu (2010, p. 639), “politicians pursue egalitarian goals at the expense of quality civil service. [There] has been a tendency for the salaries of people in the lower cadres of the civil and public service to be higher than those paid to counterparts in the private sector while the salaries of those with expertise stagnate.” As a result, there is a shortage of highly skilled workers who leave for the private sector or international organizations. It would also avoid a wholesale, potentially painful, reform that may alienate both civil servants and the population ahead of the 2015 elections. 43. In Burundi, the Government should consider pursing an asymmetric pay reform in one of its priority sectors, such as education, health, energy, or agriculture. For example, it could offer more autonomy to certain ministries, offer a larger compensation packages for doctors in exchange for performance agreements, and pilot a new performance evaluation system. Financing arrangements would have to be considered, though the cost would be significant less than a whole of government approach. Furthermore, the reform could build upon previous experiences of similar countries that have pursued this approach under difficult conditions. 44. Simultaneously, the Government should consider reestablishing the Directorate of Public Service as its own ministry. The PNRA calls for the mandate of the Directorate within the Ministry of Public Service, Work, and Social Security to be reinforced, and given both the capacity and authority to coordinate this function. Granting it ministry status would engender greater political clout and more authority to perform its reform agenda. 113 45. One of the first tasks of such a Ministry should be to establish an agreed organization structure and conduct a preliminary analysis of staffing needs for all ministries, departments, and agencies. This analysis could reinforce establishment controls by permitting hiring only when there is an existing vacancy, and direct each additional staffing decision according to the greatest need. At the same time, it would place a limit on the currently unlimited number of higher level positions in the general civil service statute. Career advancement would thus depend upon the existence of a vacancy, and candidates could thus be chosen through a competitive, merit based system. 46. At the same time, immediate measures could be taken for the National Recruitment Commission to be strengthened to perform its proper role as defined in the 2006 decree. It should also ensure that it is leading the recruitment for the public sector through a depoliticized, merit based process. 47. Furthermore, the Ministry of Finance should establish a clear overall budget constraint for the compensation of public employees. It may also consider establishing sectoral ceilings for priority sectors participating in an asymmetric pay reform. Any new proposed measures could thus be evaluated against existing budget constraints. 5.8 Concluding Remarks 48. Reforming the public administration in Burundi is not an end in itself, but a means to generate quicker developmental progress through a more professionalized civil service (Government of Burundi, 2012c). It is also a means to restore the frayed linkages between citizens and the institutions of the government by improving service delivery outcomes. 49. The public administration in Burundi is characterized by weak productivity. The analysis has shown that pay structures, recruitment, promotions, and human resource management do not currently provide sufficient incentives for strong performance, either at an agency or individual level. Although the Government has developed a public administration reform strategy that calls for comprehensive reform, it is not likely that a major transformation in the performance of the civil service can be achieved in the short or medium term. The proposed pay reform roadmap would likely entail a high fiscal adjustment, and again not necessarily engender any performance improvements. Furthermore, this reform may meet strong resistance from both unions and politicians ahead of the 2015 elections. 50. A targeted asymmetric civil service reform may be an incremental way forward. Experience from countries at similar levels of development has shown that modest, but tangible results can be achieved with such a reform in severely constrained capacity environments. This reform could be targeted to several of the Government’s key priority sectors. 51. Immediate steps should also be considered to strengthen human resource management functions. Although capacity constraints will limit scope for improvement in the short term, the Government should consider reestablishing the Ministry of Public Service as a standalone agency. This reform would send a signal that the Government is committed to strengthening the public administration. 114 Concluding Remarks 1. As elucidated in the “Rationale and Approach” of the PER, the report has been geared specifically to increase the instrumental value of its product to the deepening of the current developmental discourse on Burundi. The analysis in successive chapters therefore should be measured in terms of its utility to the Government and its partners in isolating critical policy zones for concerted examination and thereafter, it is hoped, helping to sculpt appropriate and workable governmental responses. 2. To anchor the next steps of the process, a launch event is envisaged bringing together directly involved/interested parties - at the appropriate senior level indicated by the subject matter – to consider the PER findings/assessment and the future direction of their practical application. 3. To enhance the tractability of the thematic areas already identified for the policy debate, it is proposed that supplementary, targeted policy notes be elaborated. Such notes at present would be foreseen on: • Domestic Revenue Mobilization. Enabling/hindering factors in the current environment. Appropriate targets for the OBR, in aggregate and distributed across sources. Need for further legislative/regulatory adjustment. • Drivers of Fragility. Which would be the prudent risk management strategies for adoption? • Experience with MTEF. Is the system fit for purpose in its current application in Burundi. What is the extent of the capacity building/TA required to bring the system to a suitable level of functionality? • Reforms for Wage and Salary Management. How to manage the tradeoff between the objective of fiscal sustainability and the need to address capacity constraints in the civil service? The feasibility of an asymmetric approach to the compensation regime. Considering the trade-offs/interplay between short-term incentives based on measurable performance and longer term career based performance incentives. 4. The Policy Notes will serve as discussion documents for the convening of follow-up technical meetings aimed at setting out road maps for resultant action in respective policy areas. 5. In collaboration with the CMU and the government, the team will also adapt the policy notes for public distribution in local language (including Kirundi). Radio shows will be organized to inform on and broaden the discussion further and a web outreach and marketing package will also be designed and targeted. 115 References Africa Research Institute (2013) “For State and Citizen: Reforming revenue administration in Burundi” publié le 15 October 2013 < http://www.africaresearchinstitute.org/publications/policy-voices/taxation-in-burundi/> Andrews, Matt. 2013. “Make Way for the Hybrids.” IMF PFM Blog, April 3 2013, retrieved on April 4 2013. http://blog-pfm.imf.org/pfmblog/2013/04/make-way-for-the-hybrids.html.> Africa Development Bank. 2010. Coffee Production in Africa and the Global Market Situation. Commodity Market Brief, Vol.1, No. 2, July, 2010. Benedek, Dora et al. 2012. Foreign Aid and Revenue: Still a Crowding Out Effect?, Working Paper 12/186, Washington, DC: International Monetary Fund. Brumby and Dorotinsky. 2010. “MTEFs-Key Technical and Institutional Challenges”. Presentation at the PREM Learning Week. Washington DC: World Bank. Celasun, O., X. Debrun, and J.D. Ostry. 2007. Primary Surplus Behavior and Risks to Fiscal Sustainability in Emerging Market Countries: A “Fan-Chart” Approach. IMF Staff Papers Vol. 53, No. 3. Economic Intelligence Unit. 2013. Burundi Country Report: 1st Quarter 2013. London. United Kingdom. Economic Intelligence Unit. 2008. Burundi Country Profile 2008. London. United Kingdom. Garnett, Henry, Jean-Pascal N. Nganou, and Kevin Brown (2012), “Improving the Performance of the Public Administration in Burundi Through Improving Incentives for Management and Staff While Managing a Sustainable Wage Bill,” unpublished report, the World Bank. Government of Burundi (2012), “Pour une Politique Salariale Equitable dans l’Administration Publique Burundaise,” Ministry of Public Service, Work, and Social Security. Government of Burundi (2012b), “Plan d’Action pour la Mise en Place de la Politique Salariale du Secteur Public,” Ministry of Public Service, Work, and Social Security. Government of Burundi (2012c), “Programme National de Réforme de l’Administration – Réformer l’Administration Burundaise: 35 Actions, Simples ou Plus Complexes, Pour Changer l’Etat”, Ministry of Public Service, Work, and Social Security. ___________________. 2012. Loi de Règlement, 2008, Bujumbura: Government of Burundi. ___________________. 2012. Loi de Reglement, 2009, Bujumbura: Government of Burundi. 116 ___________________. 2012. Loi de Reglement, 2010, Bujumbura: Government of Burundi. ___________________. 2012. Loi de Reglement, 2011, Bujumbura: Government of Burundi. ___________________. 2007. Loi portant fixation du Budget Général de la République du Burundi pour l’exercice 2008, Bujumbura: Government of Burundi. ___________________. 2008. Loi portant fixation du Budget Général de la République du Burundi pour l’exercice 2009, Bujumbura: Government of Burundi. ___________________. 2008. Loi portant fixation du Budget Général révisé de la République du Burundi pour l’exercice 2008, Bujumbura: Government of Burundi. ___________________. 2009. Loi portant fixation du Budget Général de la République du Burundi pour l’exercice 2010, Bujumbura: Government of Burundi. ___________________. 2010. Loi portant fixation du Budget Général de la République du Burundi pour l’exercice 2011, Bujumbura: Government of Burundi. ___________________. 2011. Loi portant fixation du Budget Général de la République du Burundi pour l’exercice 2012, Bujumbura: Government of Burundi. ___________________. 2012. Loi portant fixation du Budget Général révisé de la République du Burundi pour l’exercice 2012, Bujumbura: Government of Burundi. ___________________. 2012. Loi portant fixation du Budget Général de la République du Burundi pour l’exercice 2013, Bujumbura: Government of Burundi. Franco-Rodriguez, Susana and Oliver Morrisey. 1998. “Aid and the Public Sector in Pakistan: Evidence with Endogenous Aid.” World Development, Vol. 26, No. 7, pp. 1241-1250. Helis, Jean-Luc. 2010. “Burundi: What Lessons Has the IMF Learned from its Technical Assistance?”, IMF Public Management Blog, retrieved on December 20th 2012.http://blog-pfm.imf.org/pfmblog/2010/03/burundi-what-lessons-has-the-imf-learned- from-its-technical-assistance-.html. Hemming, Richard and Murray Petrie. 2000. A Framework for Assessing Fiscal Vulnerability. IMF Working Paper. International Monetary Fund. 2009. Uganda and Rwanda: Selected Issues, Country Report 09/36, Washington, DC: International Monetary Fund. ________________________. 2010. Burundi: Poverty Reduction Strategy Paper—Second Implementation Report, Country Report 10/312, Washington, DC: International Monetary Fund. ________________________. 2011. Assessing Reserve Adequacy, Washington, DC: International Monetary Fund. 117 ________________________. 2011. Burundi: Poverty Reduction Strategy Paper—Progress Report, Country Report 11/53, Washington, DC: International Monetary Fund. ________________________. 2011. Burundi: Fifth Review Under the Three-Year Arrangement Under the Extended Credit Facility and Request for Extension of the Arrangement—Staff Report; Staff Supplement; Press Release on the Executive Board Discussion, Country Report 11/104, Washington, DC: International Monetary Fund. ________________________. 2011. Burundi: Sixth Review Under the Three-Year Arrangement Under the Extended Credit Facility and Request for Extension of the Arrangement—Staff Report; Staff Supplement; Press Release on the Executive Board Discussion, Country Report 11/104, Washington, DC: International Monetary Fund. ________________________. 2011. Revenue Mobilization in Developing Countries, Washington, DC: International Monetary Fund. ________________________. 2012. Burundi: 2012 Article IV Consultation and First Review Under the Three-Year Arrangement Under the Extended Credit Facility and Request for Modification of Performance Criteria—Staff Report; Public Information Notice and Press Release on the Executive Board Discussion; and Statement by the Executive Director for Burundi, Country Report 12/226, Washington, DC: International Monetary Fund. ________________________. 2012. Burundi: Poverty Reduction Strategy Paper II, Country Report 12/224, Washington, DC: International Monetary Fund. ________________________. 2012. Burundi: Seventh Review Under the Three-Year Arrangement Under the Extended Credit Facility and Request for a New Three-Year Arrangement Under the Extended Credit Facility—Staff Report; Staff Supplement; Press Release on the Executive Board Discussion; and Statement by the Executive Director for Burundi, Country Report 12/28, Washington, DC: International Monetary Fund. ________________________. 2013. Burundi: Second Review under the Extended Credit Facility—Staff Report; Staff Supplement; Press Release on the Executive Board Discussion, Country Report 13/64, Washington, DC: International Monetary Fund. International Finance Corporation. 2012. Burundi: Summary Analysis and Proposals for Reforming the Regime for Investment Incentives, Mimeo, September 2012. International Labor Organization. 2012. Rapport Final: Pour une Politique Salariale Equitable dans l’Administration Publique Burundaise. Mimeo, November 2012. International Monetary Fund. 2013. Burundi: Second Review Under the Extended Credit Facility—Staff Report; Staff Supplement; Press Release on the Executive Board Discussion, IMF Country Report No. 13/64, March 2013. International Monetary Fund. 2012. Global Risks, Vulnerabilities, and Policy Challenges Facing Low-Income Countries. Mimeo, October 2012. 118 International Monetary Fund. 2009. The Implications of the Global Financial Crisis for Low- Income Countries—An Update. Mimeo, September 2009. International Monetary Fund. 2008. Food and Fuel Prices—Recent Developments, Macroeconomic Impact, and Policy Responses. Mimeo, June 2008. International Monetary Fund. 2006. Burundi: Selected Issues and Statistical Appendix, IMF Country Report No. 06/307, August 2006. Joint IMF-World Bank. 2013. Burundi Debt Sustainability Analysis. Prepared by the Staff of IMF and the IDA. January 2013. Kimonyo, J.P., and D. Ntiranyibagira. 2007. Reform of the Coffee Sector in Burundi: Prospects for Participation, Prosperity and Peace. International Alert, USAID. May 2007. Kojo, N. 2010. Diamonds Are Not Forever: Botswana Medium-term Fiscal Sustainability. World Bank Policy Research Working Paper 5480, November 2010. Law of Finance (LOF). 2012. Revised General Budget of the Republic of Burundi for the 2012 Budget Year, Law No. 1/18 of June 29, 2012, Bujumbura: The Republic of Burundi. Law of Finance (LOF). 2011. General Budget of the Republic of Burundi for the 2011 Budget Year, Law No. 1/13 of December 31, 2010, Bujumbura: The Republic of Burundi. Law of Finance (LOF). 2010. General Budget of the Republic of Burundi for the 2010 Budget Year, Law No. 1/31 of December 31, 2009, Bujumbura: The Republic of Burundi. Law of Finance (LOF). 2009. Revised General Budget of the Republic of Burundi for the 2009 Budget Year, Law No. 1/21 of September 8, 2009, Bujumbura: The Republic of Burundi. Law of Finance (LOF). 2008. Revised General Budget of the Republic of Burundi for the 2008 Budget Year, Law No. 1/19 of August 22, 2008, Bujumbura: The Republic of Burundi. Lienert, Ian. 2003. A Comparison between Two Public Expenditure Management Systems in Africa, Working Paper 03/2, Washington, DC: International Monetary Fund. Leinert, Ian. 2007. “Inherited budget practices and medium-term expenditure frameworks in Francophone and Anglophone Africa”. Are We Asking the Right Questions? Embedding a Medium-Term Perspective in Budgeting. Proceeds of the 4th Annual CABRI Seminar, Berlin, GTZ. Lonti, Zsuzsanna. and Matt Woods (2008), “Towards Government at a Glance: Identification of Core Data and Issues related to Public Sector Efficiency,” OECD Working Papers on Public Governance, No. 7, OECD Publishing. Lim, Christian and Rugwabiza, Leonard. 2009. A shock Analysis of Burundi’s Economy. Africa Development Bank, East Africa Department. 119 Moussa, Yaya. 2004. Public Expenditure Management in Francophone Africa: A Cross-Country Analysis, Working Paper 04/42, Washington, DC: International Monetary Fund. Ministry of Agriculture and Livestock (Minagrie). 2013. National Agrarian Survey of Burundi for the 2011-2012 Agricultural Campaign. Provisory Report of April 2013, Bujumbura: The Republic of Burundi. Ministry of Agriculture and Livestock (Minagrie). 2012 (B). Assessment of Harvests, Food Supplies, and Nutritional Stance, Report No. 2012-B, Bujumbura: The Republic of Burundi. Ministry of Agriculture and Livestock (Minagrie). 2012 (A). Assessment of Harvests, Food Supplies, and Nutritional Stance, Report No. 2012-A, Bujumbura: The Republic of Burundi. Ministry of Agriculture and Livestock (Minagrie). 2011 (B). Assessment of Harvests, Food Supplies, and Nutritional Stance, Report No. 2011-B, Bujumbura: The Republic of Burundi. Ministry of Agriculture and Livestock (Minagrie). 2011 (A). Assessment of Harvests, Food Supplies, and Nutritional Stance, Report No. 2011-A, Bujumbura: The Republic of Burundi. Ministry of Agriculture and Livestock (Minagrie). 2010 (B). Assessment of Harvests, Food Supplies, and Nutritional Stance, Report No. 2010-B, Bujumbura: The Republic of Burundi. Ministry of Agriculture and Livestock (Minagrie). 2010 (A). Assessment of Harvests, Food Supplies, and Nutritional Stance, Report No. 2010-A, Bujumbura: The Republic of Burundi. Nillesen, E. and P. Verwimp. 2010. Grievance, Commodity Prices and Rainfall: A Village-level Analysis of Rebel Recruitment in Burundi. Household in Conflict Network (HiCN) Hasnain, Zahid, and Nick Manning (forthcoming), “Pay Flexibility and Pay Policy Reform in the Public Sector,” the World Bank. Nguenang, Jean Pierre. 2012. “Burundi – Success with the New Mechanism for Managing External Resources”, IMF Public Management Blog, retrieved on December 20th 2012. http://blog-pfm.imf.org/pfmblog/2012/06/burundi-success-with-the-new-mechanism-for- managing-external-resources.html Nunberg, Barbara, and Robert Taliercio (2012), “Sabotaging Civil Service Reform in Aid- Dependent Countries: Are Donors to Blame?” World Development, 40(10): 1970–1981. Office Burundais de Recettes (2013), « Rapport provisoire de collecte des recettes du mois de Mai 2013 (Internal communication », Bujumbura : Office Burundais de Recettes. 120 Olowu, Dele (2010), “Civil Service Pay Reforms in Africa,” International Review of Administrative Sciences, 76(4): 632–652. Painchaud, F. and T. Stučka, 2011. Stress testing in the Debt Sustainability Framework (DSF) for Low-Income Countries, Knowledge Brief, Economic Policy and Debt Department, World Bank. Pack, Howard and Janet Rothenberg Pack. 1990. “Is Foreign Aid Fungible? The Case of Indonesia.” The Economic Journal, Vol. 100, No. 399, pp. 188-194. PEFA. 2012. Evaluation PEFA 2011: Burundi, Brussels: European Union. PEFA 2010. Evaluation PEFA 2009: Burundi, Brussels: European Union. Ramakrishnan, Subramaniam. 1998. “Budgeting and Financial Management in Sub-Saharan Africa: Key Policy and Institutional Issues.” Development Discussion Paper No. 622, Cambridge, Massachusetts, Harvard Institute for International Development. Republique du Burundi, Ministere de la Fonction Publique (2012). Pour une Politiques Salariale Equitable Dans L’administration Burundaise, Rapport Provisoire. Bujumbura: Burundi. Republique du Burundi, Ministère Des Finance (2012), Burundi PAGE - Rapport d’Achèvement du Project, Bujumbura: Burundi. Republique du Burundi, Ministère Des Finance (2011), Burundi PAGE - Rapport d’Achèvement du Project, Bujumbura: Burundi. Schaechter, A. et al. 2011. A Toolkit to Assessing Fiscal Vulnerabilities and Risks in Advanced Economies. IMF Working Paper WP/12/11. Sekwat, Alex. 1997. “Public Budgeting Deficiencies in Sub-Saharan Africa: A Review.” Journal of Public Budgeting, Accounting & Financial Management, Vol. 9, pp. 143-62 USAID. 2010. Burundi Coffee Industry Value Chain Analysis. USAID/ COMPETE East Africa Trade Hub, June 2010. World Bank. 2012. Country Assistance Strategy for the Republic of Burundi for the Period FY13–16, Washington DC: the World Bank. World Bank. 2011. Burundi Country Economic Memorandum, Washington DC: the World Bank. World Bank. 2011. Burundi Country Economic Memorandum, Washington DC: the World Bank. World Bank. 2008. Republic of Burundi Public Expenditure Management and Financial Accountability Review (PEMFAR), Washington DC: the World Bank. World Bank (2012), “Fiscal Challenges, Security and Growth Issues in Burundi,” Burundi Policy Note Series. 121 World Bank (2010), “Serbia: Right-Sizing the Wage Bill,” Report No. 54056, available at http://www- wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2010/11/04/00033495 5_20101104023105/Rendered/PDF/540560ESW01gra11121201010BOX353774B.pdf World Bank (2008), “Republic of Burundi: Public Expenditure Management and Financial Accountability Review (PEMFAR)”, available at http://ddp- ext.worldbank.org/EdStats/BDIper08.pdf World Bank. 2012. Country Assistance Strategy for the Republic of Burundi for the Period FY13-16, Report No. 72334-BI, Washington DC: the World Bank ______, (2011a) Public Expenditure Review: Improving the Efficiency of Public Investment, Washington DC: the World Bank ______, (2011b) Burundi Country Economic Memorandum, Washington DC: the World Bank. ______, (2008) Country Assistance Strategy for the Republic of Burundi for the Period FY09-12 ______,2008. Republic of Burundi - Public Expenditure Management and Financial Accountability Review (PEMFAR) : Improving Allocative Efficiency and Governance of Public Expenditure and Investing in Public Capital to Accelerate Growth and Reduce Poverty, Report No. 42160-BI, Washington, DC: the World Bank. World Bank. 2011. Program Document for a Proposed Grant in the Amount of SDR 22.1 Million (US$35 Million equivalent) to the Republic of Burundi for a Fifth Economic Reform Support Grant (ERSG V), Report No. 61132-BI, Washington, DC: the World Bank. __________. 2012. Program Document for a Proposed Grant in the Amount of SDR 16.5 Million (US$25 Million equivalent) to the Republic of Burundi for a Sixth Economic Reform Support Grant (ERSG V), Report No. 70931-BI, Washington, DC: the World Bank. 122 Annex A : GoB-donor PER Technical Working Group – TOR Context 1. The Government of Burundi faces a narrow fiscal space that reduces considerably the options for domestic resources to contribute adequately to the financing of the PRSP-II’s program. In the medium-long term, the promotion of an enabling business environment coupled with critical economic infrastructure, as well as the reform in land management and entitlement, are expected to foster the incentives of private investors - especially from abroad (FDI) - to develop investment projects that widen employment opportunities and create positive effects on fiscal revenues. 2. In the short term, however, Burundi’s economy remains fragile and exposed to important sources of shocks and volatility. In particular, external financing remains critical to the development program of Burundi, and its uncertain prospects pose serious difficulties to the achievement of expected results. Overall, this volatile internal and external economic and financial context has increasingly put the Government in a difficult position where the execution rhythm of voted budgets (often prepared in very difficult conditions requiring time and negotiation) within planned timelines and resource envelopes (as detailed in line ministries’ commitment plans) is constantly undermined by delays in the disbursement of the external financial resources intended to cover annual fiscal deficits. In addition, the lengthy and burdensome procedures governing the chain of expenditure, alongside inter alia the inability of the information system to manage properly budgetary commitment and legal commitment, require simplification and computerization effective enough to improve controls and reduce human errors. 3. It is therefore essential to get a better understanding of the budget process, its vulnerability to various shocks and their fiscal implications, with the ultimate goal of enabling the government to better manage future risks and mitigate their impact for the national economy and the welfare of society as a whole. The Burundi Public Expenditure Review (PER) aims to do precisely that, as part of the broader World Bank support to strengthening public finance management (PFM) and economic governance reform in Burundi. More precisely, the objective of the PER is to provide the Government of Burundi with a robust and systematic analysis of the challenges and opportunities to improve the fiscal space under conditions of structural vulnerability, thereby strengthening the country’s resilience to critical shock. Mandate and objectives 4. The authorities are eager to contribute actively to the elaboration of the new Public Expenditure Revue (PER) of Burundi. The Government contributions will take the form of data sharing, policy orientations, and feedback on preliminary results. During the consultation process, the Permanent Secretary of the Ministry of Finance and Economic Development Planning (MFEDP) reiterated the strong interest of the Government for the Burundi PER. Moreover, she expressed her satisfaction and agreement to lead a joint Government-Donor “Technical Working Group”, hoping that this group will help to improve the quality of the final PER report and strengthen its overall policy relevance. 123 5. The PER Technical Working Group (PER-TWG) is designed to act as an essential instrument for the study. More precisely, through regular interactions and exchanges with the World Bank team, the PER-TWG is expected to (i) identify critical issues and provide realistic solutions that improve effectively resilience in public finance management; (ii) promote government ownership of the findings from the analysis; and (iii) maximize the prospects of implementation of any action plan and policy recommendation resulting from the analytical work. In addition, it would help the PER team to ensure that the foreseen ‘action plan’ based on the study complements ongoing efforts to consolidate public finance management in Burundi like the new Strategy for the Public Finance Management. 6. The PER-TWG will also closely collaborate with the “Cellule d’Appui” in charge of monitoring implementation of PFM reforms. In case needs be, the Cellule d’Appui will provide the technical working group with those operational cadres that are capable of affording specific contributions to the analytical work. Through such a direct collaboration, the PER-TWG will effectively involve the “Cadre de Partenariat” in the elaboration process of the PER, offering to the governmental authorities and the broader donor community the opportunity to be informed of progress made in the study and provide comments and feedback on the key findings emerging from the analysis. Structure and Composition 7. The structure and composition of the PER-TWG is expected to the Bank team access complementary analytical work and contextual information to respond to the objectives of the PER study, thus proposing solutions to the critical issues facing budget managers in Burundi. 8. It has been agreed that the Minister of Finance and Economic Development Planning will formalize the designation of these high level local staff through an ordinance. The Second Vice- President will be informed of that designation. Since the PRSP unit is greatly involved in the monitoring of the PRSP-II program, the authorities think that a representative from this unit should be appointed in the PTWG as a contributor. In addition, the Working Group will include representatives from key government agencies and departments interested to provide guidance and inputs for the analysis. 9. The PTWG will operate through both informal exchanges of views and data and periodic formal meetings chaired by the Permanent Secretary of the MFEDP. Bilateral discussions and consultations will constitute the first channel through which Bank Task Team - under the guidance and supervision of the CMU - will discuss with Government authorities the key hypotheses of the study, seek comments and feedback on draft reports and preliminary analysis, as well as discuss the political and technical feasibility of possible reform solutions. As agreed with the Permanent Secretary, at least 1 meeting per month will be organized, to take stock of progress achieved in the different areas covered by the PER, including brainstorming on possible orientations and adjustments in the substance and possibilities to share any relevant technical tools developed by PER contributors. If deemed necessary, videoconferences will also be organized to ensure close collaboration between World Bank office and PTWG. 10. The PER-TWG will be composed of a Chair plus representatives from the Government acting as focal points’ for each of the key subject areas covered in the study. Also, one member 124 from the World Bank and one representative from the donor community will join the group acting as focal point for the PER team. Membership and responsibilities 11. The core members of the PER-TWG (and their respective responsibilities) will be as it follows: i. Mrs. Immaculée Bigirimana, Permanent Secretary: MFEDP. She will act as Chair of the PER-TWG. She has experience in the area of public expenditures as per her former functions in the Ministry as Director General of Budget and Public Accounting (Directrice Générale du Budget et de la Compabilité Publique - DGBCP). She will provide guidance on the operational issues that could be analyzed in detail or not. She will secure that the Minister of Finance and Economic Development Planning remains informed during the whole PER process. Under her leadership, we expect that the PTWG will be contributing to the extent of their competencies and time constrains She will report directly to the Minister. ii. Mrs. Aude Toyi, Director of the Office of Big Taxpayers: OBR-MFEDP. She will act as focal point for the analysis of revenues. Specifically, she will provide feedback and inputs, and, if needed, paragraph write-ups, on issues pertaining to the implementation of tax reforms. She can easily provide any data we would need on fiscal and non-fiscal revenues as she can connect the PTWG and Bank team to relevant services within the OBR. On fiscal space issues, her sharp understanding of the need to address fiscal loopholes and rationalize potential tax contributing sectors (e.g., mining, etc.) make her a potentially active contributors to the PTWG. She is often Acting General Commissioner and will report to the OBR General Commissioner. iii. Mrs. Béatrice Samandari, Director of Budget: MFEDP. Recommended by the Permanent Secretary of the MFEDP, she will act as focal point for the analysis of public expenditure. More precisely, this contributing authority will link the PTWG and Bank team to both policy and implementation aspects of the chain of expenditures, especially putting in perspective the new reform aiming at promoting the “Contrôleur des Engagements des Dépenses (CED)”, which is set to be progressively implemented and simultaneously with a new information management system using a new budget nomenclature. She will work closely with Mrs. Léa Ntabarushimana, the person in charge of a department that monitors and manages the “classification fonctionnelle”. iv. Mrs. Rose Kankindi, Director of Payroll: Ministry of Civil Service (MCS). She will act as ‘focal point’ for the analysis of wage bill / public expenditure management. On the impact of the pay policy on public finance management, she will provide feedback and inputs in collaboration with Mr. Balthazar Barutwanayo, Administrator of the Information & Technology Payroll Center. These two contributors will help find out or update any data required to carry out deep analyses on wage bill issues. In particular, they will help the Bank team shape better its proposals on feasible alternatives to manage better civil servants (e.g., career development, reduction in numbers, recruitment of rate skills, etc.) and improve their overall work conditions. They will report to the Minister of Civil Service and to the Permanent Secretary of the MFEDP. 125 v. Mr. Emile Nimpaye, Economist and local expert on aid flows’ management and data: National Committee for Aid Coordination (CNCA). He will act as ‘focal point’ for the analysis on aid-dependency / aid effectiveness. The political economy of aid coordination in Burundi will probably need to be understood better beyond the numbers that often appear difficult to interpret for outsiders. Delays in aid disbursement (budget support and investment) undermine the credibility / feasibility of the development program of the Government on a daily basis. The conditions underlying the disbursement of already scanty aid resources need to be properly understood from both Government and Development Partners (DPs) sides. Among other issues, Mr. Emile Nimpaye could help us clarify these issues as the Bank team endeavors to elaborate on the role of external financing unpredictability and disbursement delays in the fragility / vulnerability that affect negatively the normal execution of voted budgets. He will report to Mr. Pamphile Muderega, Permanent Secretary of the CNCA. vi. Mr. Aurélien Serge Beko and Mr. Gerard Muringa: they will both act as focal points for the PER team while also representing/covering the donor perspective on key subject areas of the study. 126 Annex B : Arrears and pending payment Differentiating Between ‘Arrears’ and ‘Pending Payments’ 1. Since 2008, the Loi de Réglèment presents the balance of the different components of the treasury account (balance cumulée des comptes). This allows seeing the payments pending at the end of every year by different categories. Table 1 shows the movements of the account “providers of goods and services” (fournisseurs) for 2008. The number shown under balance d’entrée – credit (FBU 304 million) represent the payments to providers of goods and services pending from 2007. The numbers presented under mouvements – debit and mouvements – credit, represent, respectively, the payments to and claims from providers of goods and services. Mouvements registered under année courante (current year) represent claims and payments done for operations budgeted in 2008. Mouvements registered under année précedente (previous year) correspond to payments of claims dating from 2007 or before. The numbers registered under solde fin de période represent the balances that will be carried over to 2009: in practical terms, this means that, in 2009, the Burundi Government will have to pay FBU 3.068 billion (the government got claims for 103 and paid 100) to providers for goods and services received in 2008, plus FBU 115 million for goods and services received in 2007 or even before (the government had a stock of pending payments for 304 million and paid 188). Table B.1 . Balance of the account Fournisseur, 2008 2008 Balance d'Entrée Mouvements Solde fin de période Débit Crédit Débit Crédit Débit Crédit 401.Fournisseurs Année Courante 0 0 100,403,066,953 103,471,292,210 0 3,068,225,257 Année antérieure 0 304,125,295 188,533,654 0 0 115,591,641 Source: Loi de Réglèment 2008 2. The Burundi legislation sets a deadline of 60 days to cover payments; otherwise, the balance will become an arrear. This means that, in practice, the FBU 115 million registered under Solde fin de période – credit is an arrear: this number is the remaining balance of the accounts that were unpaid in 2007 (304 million) of which a portion (188 million) were paid in 2008. It is less clear whether the 3.068 billion registered under année courante are arrears: they might be claims made, for instance, on December 31th 2008 that will be disposed regularly within the 60 days deadline, or they might be claims made since January 1st 2008. It is impossible to determine the extent to which a number registered as pending payment for the current year is an arrear. 3. Inferences on the accumulation of arrears can be made on the basis of the data presented in the Lois de Réglèment. Table 2 shows the balance of the account Fournisseurs for 2009: the number presented under Balance d’entrée – credit for année antérieure (FBU 3.183 billion) is the addition of the end of year balances (solde fin de période) for 2008 127 (année courante) and before (année antérieures). The data shows that the Burundi state made payments for FBU 1.950 billion (mouvements – debit). Therefore, the FBU 1.233 billion presented as an end of year balance (solde fin de période) must logically be an arrear. The reason why that amount is an arrear is the following: The 1.233 billion is the pending balance of a combination of two stocks of pending payments: the first stock is the 115 million registered as année antérieure in the 2008 balance, and dates, at the very least, from December 31st 2007; the second stock is the 3.068 billion registered as année courante in the 2008 balance and dates, at the very least, from December 31st 2008. Even if we assume that the 115 million pending from 2007 were paid during 2009, the remaining balance at the end of the year (1.233 billion) must date, at the very least, from, December 31st 2008. The days passed between December 31st 2008 and December 31st 2009 is 365, which is larger than 60, the number of days provided by the Burundi law to deal with payments (figure 1 presents a simplified figure of this explanation). Table B.2 Balance of the account Fournisseur, 2009 2009 Balance d'Entrée Mouvements Solde fin de période Débit Crédit Débit Crédit Débit Crédit 401.Fournisseurs Année Courante 0 0 101,111,013,978 104,065,604,245 0 2,954,590,267 Année antérieure 0 3,183,816,898 1,950,476,041 0 0 1,233,340,857 Source: Loi de Réglèment 2009 128 Figure B.1: Simplified explanation of treasury balance movements Payments pending at the Payments made during Payments pending at the beginning of 2009 2009 end of 2009 3,183 million 1,950 million 1,233 million= 3,183 – 1,950 Composed of: Composed of: 115 million 3,068 An unknown At least pending million amount 1,233 from pending dating from million 12/31/2007 from 12/31/2007 pending 12/31/2008 (hopefully 0) from 12/31/2008 4. It is important to distinguish between payments pending and arrears. Pending payments is the addition of two balances: unsettled operations of the current year plus unsettled operations from previous years. Arrears can only be inferred based on the remaining balances of claims made in previous years. Differentiating between arrears and pending payments is important as the implications of each one of them on the Burundi treasury are different. 5. The data provided in the Lois de Réglèment allow us to do two things: 1. Assess, based on the explanation provided in this section, whether the government is cumulating arrears. 2. Start a policy dialogue on the extent to which pending payments put strain on the Burundi Treasury at the beginning of the year. This section will complement the discussion on the arrival of budget support. 129 6. The PFM systems of Burundi do not allow assessing whether the Government is accumulating arrears directly. The answer to this question can only be inferred on the basis of the treasury balances and will be a back of the envelope calculation subject to discussion, but useful to start a fruitful policy dialogue aimed to improve the management of the Treasury and, therefore, increase the standard of living of the people of Burundi. 7. Table 3 shows the movements of the different accounts of the treasury balance for the period 2008-2012. For each year, column Balance d’entrée Crédit represents the stock of pending payments to be paid in year n. Mouvements represents the claims (crédits) and payments (débits) made in year n. Last, column Soldes fin de période Crédits represents the stock of pending payments for claims made either during year n (année courante), or in year n-1 or even before (années antérieures). For year n, années antérieures represents the summation of année courante and années antérieures of year n-1, as explained in the previous section. The source of the data is the Lois de Réglèment 2008-2011 approved by the Burundi Legislative Power. 8. The table contains red, green, and yellow cells. • Red cells represent values where the summation of the balances of the current year (année courante) plus the balances of the previous years (années antérieures) presented in column solde de fin de période in the Loi de Règlement of year n-1 does not correspond 100% with the values of Balance d’entrée Crédits presented in the Loi de Règlement of year n. Most of the differences are small, but are quite considerable in a few cases, particularly those found for providers of goods and services (Fournisserus) and salaries and pensions (Salaries – Pensions et Indemnités) for the 2010-2011 period. At the request of the Burundi authorities, we kept the values presented in the Lois de Réglèment of year n instead of the manual computations. • Green cells represent errors that were found in the original data but were corrected after consultation with the Burundi authorities. This problem concerns only one account: providers of goods and services (fournisseurs) in year 2011. The mistake was that there was a value different from 0 in column Mouvements - Crédits, which does not make sense from an accounting perspective: the accounts for previous years (années antériures) cannot have “new” claims made during a fiscal exercise. The Burundian authorities confirmed the mistake and confirmed that the stock at the end of the year (Soldes fin de période Crédits) was correct. After removing the amount of Mouvements – Crédits in the two sides of the balance, the real value of Mouvements – Débits was obtained by difference. • Yellow cells represent accounting errors similar to those of green cells, with the only difference that the Burundi authorities were unable to provide guidance or clarifications. As a result, the team decided to keep the values as they appear in the Loi de Réglèment. 9. The table shows two totals at the bottom: • “Payments pending, BOY” is the summation of the total payments pending at the beginning of year n. For the reasons explained in the section above, it is not clear how much of these pending payments are arrears: a claim may be waiting for 130 payment since January 1st of year n-1, or since December 31st of year n-1. Our assumption is that all of these claims are within the 60 days legal deadline to be paid and cannot be considered arrears. This also means that our estimates of arrears accumulation are actually conservative and may be even considered the lower bound. In any case, these are claims that the Treasury has to pay at the beginning of the year and may put strain on the Treasury as will be explained below. • “Stock of arrears, EOY” is the summation of the stocks of pending payments registered under années antérieures at the end of year n (Solde fin de période Crédit). For reasons explained above, at the end of year n, a claim registered under années antérieures must logically be an arrear, as it has been waiting for payment at least since December 31st of year n-1. 131 Table B.3 Pending payments: stocks and flows, 2008 - 2012 2012 2008 2009 2010 2011 (est.) Solde Solde Solde Solde Balance fin de Balance fin de Balance fin de Balance fin de Balance d'Entrée Mouvements période d'Entrée Mouvements période d'Entrée Mouvements période d'Entrée Mouvements période d'Entrée Crédit Débit Crédit Crédit Crédit Débit Crédit Crédit Crédit Débit Crédit Crédit Crédit Débit Crédit Crédit Crédit 401.Fournisseurs Année Courante 0.0 100,403.1 103,471.3 3,068.2 0.0 101,111.0 104,065.6 2,954.6 0.0 87,675.2 91,644.8 3,969.5 0.0 84,429.1 87,932.6 3,503.5 0.0 Années antérieures 304.1 188.5 0.0 115.6 3,183.8 1,950.5 0.0 1,233.3 4,268.4 569.2 0.0 3,699.1 7,668.7 5,073.3 0.0 2,595.4 6,098.8 402. Fournisseurs Immobilisation Année Courante 353.0 48,170.1 48,232.2 415.1 0.0 41,073.0 43,120.9 2,047.9 0.0 66,350.0 69,128.8 2,778.8 0.0 77,925.3 92,425.5 14,500.2 0.0 Années antérieures 0.0 0.0 0.0 419.6 0.0 0.0 419.6 2,467.5 1,964.5 0.0 503.0 3,281.9 1,288.4 0.0 1,993.5 16,493.6 408.Autres créanciers Année courante 0.0 28,449.5 28,510.0 60.5 0.0 49,239.6 49,246.2 6.6 0.0 23,679.8 23,718.0 38.1 0.0 28,787.8 29,304.0 516.2 0.0 Années antérieures 0.0 0.0 0.0 60.5 0.0 0.0 60.5 67.1 0.0 0.0 67.1 105.3 0.0 0.0 105.3 621.5 421. Salaires -Pensions et indemnités Année courante 0.0 138,143.9 138,792.7 648.8 0.0 161,257.7 161,257.7 0.0 0.0 198,977.0 198,977.0 0.0 0.0 294,894.9 319,836.4 24,941.5 0.0 Années antérieures 14.6 14.6 0.0 0.0 648.8 560.9 0.0 88.0 880.0 806.9 0.0 73.1 73.1 1,275.7 2,405.3 1,202.6 26,144.1 4221. Mutuelle de la Fonction Publique Année Courante 0.0 9,574.1 9,640.8 66.8 0.0 10,240.9 10,517.6 276.7 0.0 9,328.8 9,779.6 450.8 0.0 9,928.8 10,134.6 205.8 0.0 Années antérieures 49.2 49.1 0.0 0.1 66.9 0.0 0.0 66.9 343.6 3.5 0.0 340.1 0.0 0.0 0.0 0.0 0.0 4222. INSS Année Courante 0.0 168.8 184.6 15.8 0.0 172.7 188.6 15.9 0.0 2,429.1 3,092.4 663.3 0.0 8,532.2 9,192.8 660.7 0.0 Années antérieures 63.5 63.5 0.0 0.0 15.8 0.0 0.0 15.8 31.7 0.0 0.0 31.7 695.1 0.0 0.0 695.0 1,355.7 4228. Autres organismes Année Courante 0.0 4,749.9 4,750.1 0.2 0.0 5,684.3 6,375.9 691.6 0.0 10,429.8 14,269.8 3,840.1 23,064.1 23,168.3 104.2 0.0 Années antérieures 131.2 0.0 0.0 131.2 131.3 258.7 255.5 128.1 819.7 0.0 0.0 819.7 4,659.8 0.0 0.0 4,659.8 4,764.0 441. Projets de dévéloppement Année Courante 0.0 8,151.0 8,621.8 470.8 0.0 7,956.3 8,236.9 280.6 0.0 6,326.5 6,365.8 39.3 0.0 10,173.9 10,221.3 47.4 0.0 Années antérieures 53.3 5.0 0.0 48.3 519.2 1.3 0.0 517.8 798.4 0.0 0.0 798.4 837.7 20.1 0.0 817.6 865.0 442. Subventions publics autonomes Année Courante 0.0 18,037.8 18,566.8 529.0 0.0 24,764.1 24,877.4 113.3 0.0 29,432.4 30,350.8 918.3 0.0 30,557.2 30,715.7 158.5 0.0 Années antérieures 60.7 14.6 0.0 46.1 575.0 257.6 0.0 317.5 430.8 6.4 0.0 424.4 1,342.7 910.6 0.0 432.1 590.6 443. Admnistrations personnalisées Année Courante 0.0 26,866.5 26,885.9 19.3 0.0 30,692.5 30,842.1 149.7 76,592.0 77,393.1 801.1 0.0 77,889.5 78,867.3 977.8 0.0 Années antérieures 62.4 52.6 0.0 9.8 29.1 229.6 200.5 0.0 149.7 149.7 0.0 0.0 801.1 208.9 0.0 592.3 1,570.1 Payments pending, BOY 738.9 5,650.1 10,257.0 19,465.3 58,503.4 Stock of arrears , EOY 351.0 2,847.5 6,756.7 13,093.5 Outflow: payments to claims of previous years 387.9 3,258.6 3,500.3 8,777.1 132 Payments pending and the Treasury 10. The data from the Treasury Balances allows us to assess the leftovers to be paid at the beginning of every year. As explained in the previous sections, in order not to accumulate arrears, the government must dispose of all claims within a delay of 60 days. Looking at monthly data on revenue collected by the OBR and on the donors’ budget support received at the BRB, this section will make the case that payments pending put strain on the Burundi treasury that prevents the Government to execute the budget parsimoniously, calling for a better cash management and more realistic budgets. 11. Table B.4 presents a calculation on the financing needs of the Burundi Government in the first two months of the year. The table shows the money collected by the OBR106 and the budget supports received at the BRB during January and February of 2011 and 2012;107 the payments pending divided in the pending balance of operations budgeted in the previous year and arrears; an estimate of the payments to be made during the first two months of the year,108 and 4 balances depending on the decisions made by the Burundi authorities. 12. The first balance assumes that the government clears all the pending payments and executes the budget parsimoniously (i.e., 100% of our estimated values go through the chaîne de la dépense). The second balance assumes that the government only clears the balance of the operations from the previous years, does not clear the arrears, and executes the budget parsimoniously. These two scenarios are unrealistic, as capital projects take a long time to be implemented and Burundi’s execution rate of capital expenditure has been traditionally low. A more realistic assumption is that the government postpones all capital investments and still tries to clear the pending payments from last year. The balances for this assumption are shown at the bottom of the table. The last two scenarios assume that the government executes the budget parsimoniously but does not clear the payments pending. 106 The 2012 budget expected the following extraordinary revenues: reimbursement for the AMISOM mission, liquidation and privatization of state-owned companies, and recovery of assets. None of these were received in January or February 2012. 107 Ideally, this analysis should have been done with more years of data, but the OBR only had data for 2011 and 2012. In any case, the opinion of the authors is that this area deserves closer investigation and follow-up. 108 For the sake of simplicity, the payment of debt principal was excluded from the calculations. The implication is that debt is financed regularly and does not represent an additional source of strain on the Treasury. 133 Table B.4 Income, Expenditures, and Balances of the Burundi Government during the first two months of 2011 and 2012, millions of FBu 2011 2012 Income received during first two months of the year 96,172.6 131,931.9 Revenue collected by OBR 64,756.1 85,282.2 Budget support received at the BRB 31,416.5 46,649.7 Payments pending from previous year 19,465.3 58,503.4 Balance of operations budgeted in the previous year 12,708.6 45,409.9 Arrears 6,756.7 13,093.5 Payments to be made during first two months of the year 1/ 110,699.0 118,007.3 Current expenditure 89,523.6 97,094.7 Salaries 45,406.3 46,611.5 Goods and services 16,347.2 16,317.5 Contributions to International Organizations 1,690.8 627.0 Transfers to households 5,486.5 5,653.6 Subsidies 14,662.6 20,683.8 Interest payments 5,930.2 7,201.3 Capital expenditures 21,175.3 20,912.6 (I) Balance assuming that Government clears all payments pending and executes budget at 100% -33,991.7 -44,578.8 (II) Balance assuming that Government clears payments pending excluding arrears and executes budget at 100% -27,235.0 -31,485.3 (III) Balance assuming that Government clears all payments pending and postpones capital expenditure (rest of the budget, executed at 100%) -12,816.3 -23,666.2 (IV) Balance assuming that Government clears payments pending excluding arrears and postpones capital expenditure (rest of the budget, executed at 100%) -6,059.6 -10,572.7 (V) Balance assuming that Government DOES NOT clear pending payments and executes budget parsimoniously 4,938.9 72,428.0 (VI) Balance assuming that Government DOES NOT clear pending payments and postpones capital expenditures 26,114.3 93,340.6 1/ Assumption: values in the budget divided by 6 Sources: OBR, BRB, Loi de Règlement 2011, Lois de Finances 2011-2012, and authors’ own calculations. 134 Annex C : Regression Results underlying FSA 1. Given the short time series available for Burundi, we keep the estimation to the simplest possible forms. The following regression variables are used: • GDP is the real GDP in local currency at 2005 prices • AG_GDP is the agricultural value added in local currency at 2005 prices • REVENUE is the central government revenue excluding grants in local currency at 2005 prices • P_COFFEE is the world price of Arabica coffee in constant US cents per kilogram • RAIN is the deviation of annual precipitation from the long run average, normalized to vary from 0 to 100 • WAR is the 1993–2000 civil conflict dummy. 2. The unit root test results for the variables are reported in Table A1. The models and regression results are reported in Table A2–A3. Coffee shocks 3. The impact of coffee price shocks on government revenue is econometrically estimated using a simple reduced-form time-series regression. World coffee prices and civil war are treated as exogenous variables. Adding their lagged values did not significantly improve the estimation. The results imply: ∆Ln (REVENUE t ) = −0.564 + 0.105 Ln (P _ COFFEE t ) − 0.029WAR t (− 1.32 ) (1.42 ) (− 0.49 ) The impact of shocks to coffee prices on growth of revenue is given by: ∂∆Ln (REVENUE ) = 0.105 ∂Ln (P _ COFFEE ) 135 4. The stochastic coffee price shock to be applied to the baseline projection of growth in revenue is calibrated based on the historical data in 1980–2012, and assumed to be normally distributed with mean zero and standard deviation defined by the historical data: ε tP _ COFFEE ~ N (0 , 0.38) The growth of revenues under the stochastic scenario is defined as: SS t = baseline REVENUE growth ratet + 0.105 * ε tP _ COFFEE . Rainfall shocks 5. The impact of rainfall shocks on the agricultural GDP is estimated using a simple reduced-form time series regression. Annual rainfall and civil war are treated as purely exogenous variables. We tried several models, including models using REVNEUE and GDP as a dependent variables, respectively. The model using AG_GDP as a dependent variable was chosen as it offered the best fit with the data. Including lagged values of rainfall and civil war did not significantly improve the estimation. The estimation results imply: ∆Ln( AG _ GDPt ) = −0.032 + 0.013Ln(RAIN t ) − 0.025WARt . (− 1.16) (1.60) (− 1.05) The impact of shocks to rainfall on growth rate of agricultural GDP is given by: ∂Ln( AG _ GDP ) = 0.013 . ∂Ln(RAIN ) 6. We assume that a 1 percent shock to agricultural GDP will lead to 0.50 percent shock in GDP—somewhat higher than the share of agriculture in GDP (0.34) but lower than the total elasticity implied by the IMF’s (share of agriculture in GDP + 0.71). 7. The stochastic rainfall shock to be applied to the baseline projection of GDP growth is calibrated based on the historical rainfall data (1980−2008), and is assumed to be normally distributed with mean zero and standard deviation 0.266: ε tRAIN ~ N (0 , 0.266 ) The growth of real GDP under the stochastic scenario is defined as: SS t = baseline GDP growth ratet + 0.50 * 0.013 * ε tRAIN . 136 Table C.1 Unit Root Tests: Augmented Dickey-Fuller and Phillips-Perron Tests ADF PP Level First difference Level First diff Variables: With constant No constant With trend With constant No constant With trend With trend With trend Log of GDP -1.55 0.965 -2.179 -1.412 -1.202 -1.372 -1.433 -4.14* Log of AG_GDP -1.527 -0.045 -3.335* -1.937 -1.933* -1.499 -1.919 -5.912* Log of REVENUE -2.124 1.003 -2.042 -4.477* -4.405* -4.397* -2.492 -7.723* Log of P_COFFEE -2.062 -0.18 -1.89 -4.146* -4.222* -4.236* -2.308* -5.744* Log of RAIN -1.466 -0.91 -1.718 -2.866* -2.971* -2.761 -4.579* -13.082* Note: Optimal lag lengths for tests for each variable were chosen based on Schwarz Information Criterion for the ADF tests as follows: LN_GDP (3), LN_AG_GDP (4), LN_REVENUE (1), LN_P_COFFEE (1), LN_RAINFALL (2). An asterisk (*) indicates the rejection of the unit root null hypothesis at 90 percent critical level or better. Table C.2 Regression Results for Coffee Price Shocks . regress dlnrevenue lnp_coffee war Source | SS df MS Number of obs = 32 -------------+------------------------------ F( 2, 29) = 1.23 Model | .056297984 2 .028148992 Prob > F = 0.3081 Residual | .665610861 29 .022952099 R-squared = 0.0780 -------------+------------------------------ Adj R-squared = 0.0144 Total | .721908845 31 .023287382 Root MSE = .1515 ------------------------------------------------------------------------------ dlnrevenue | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- lnp_coffee | .1053907 .0740651 1.42 0.165 -.0460895 .2568709 war | -.0292527 .0599567 -0.49 0.629 -.1518779 .0933726 _cons | -.5639143 .4200392 -1.34 0.190 -1.422991 .2951622 ------------------------------------------------------------------------------ 137 Table C.3 Regression Results for Rainfall Shocks . regress dlngdp lnrain war Source | SS df MS Number of obs = 27 -------------+------------------------------ F( 2, 24) = 3.70 Model | .01452525 2 .007262625 Prob > F = 0.0397 Residual | .047073279 24 .001961387 R-squared = 0.2358 -------------+------------------------------ Adj R-squared = 0.1721 Total | .061598529 26 .002369174 Root MSE = .04429 ------------------------------------------------------------------------------- lngdp | Coef. Std. Err. t P>|t| [95% Conf. Interval] --------------+---------------------------------------------------------------- lnrain | .0049515 .0062728 0.79 0.438 -.007995 .017898 war | -.0492275 .0182825 -2.69 0.013 -.0869607 -.0114942 _cons | .0177835 .020971 0.85 0.405 -.0254986 .0610655 ------------------------------------------------------------------------------- 138 Annex D : Wage Bill Management study – background data Policy Scenario Analysis: Reform of civil service wages 1. To estimate the impact of civil service wage reform under alternative options, we use the survey collected in January 2012—slightly older than the June 2012 survey underlying the study by the Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). The distributions of workforce across pay grades are similar between the two surveys, however. Compare Table D1 below with Table 7 in Chapter 5. The sample mean salaries within each pay grade also seem to match up reasonably well. This gives us some confidence that the simulation based on the data at hand will give us reasonably close estimates of the impacts of the proposed reforms. Table D.1 Distribution of the workforce across the pay grade Source: Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). 2. The four reform options under discussions are shown below in the text with their respective proposed salary scale. 139 Table D.2 Option 1 proposed wage scale Source: Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). Table D.3 Option 2 proposed wage scale Source: Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). Table D.4. Option 3 proposed wage scale Source: Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). 140 Table D.5 Option 4 proposed wage scale Source: Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). 3. The reference salaries for SC1 (the maximum within each grade) and SC2 (the average within each grade) are estimated based on the Jan 2012 survey. For the former, the 99-percentile salary in each grade is used instead of the maximum to avoid the extreme values. The reference salaries for each pay grade under SC3 and SC4 are taken directly from the study by the Government of Burundi, Ministry of Public Service, Work, and Social Security (2012). 4. In estimating the fiscal impact, the following assumptions are made under all options. (i) The number of civil service employees will grow by 10 percent per year in the priority sectors (education and health), and 0 percent per year in other sectors; (ii) Distribution across the civil servants will remain the same, even after new hiring; (iii)After the initial salary adjustment (at t+1), the annual automatic salary increase of 6 percent is maintained (as assumed in the Government of Burundi, Ministry of Public Service, Work, and Social Security, 2012). 5. In addition, the following assumptions are made about the adjustment process. • Under Option 1 and Option 3, we assume the costs of adjustments are evenly spread over 3 years and 2 years, respectively. For example, under Option 1, a third of the existing employees are brought under the new salary scale in the first year (t+1), another third in the second year (t+2), and all existing employees in the third year (t+3). Under Option 3, a half of the existing employees are brought under the new salary scale in the first year (t+1), and the remaining half in the second year (t+2). • During the adjustment period, there will be no automatic adjustment of salaries. • Newly hired employees are paid according to the new salary scale. 141 Let: TWBt Total wage bill GRt Growth in wage bill N tP Number of employees, Priority sector P NP N t Number of employees, Non priority sector, NP P Wt Weighted average wage, P Wt NP Weighted average wage, NP π Automatic wage growth per year g Net hiring (growth of employee) in P α Ratio of employees in NP relative to P, initial period β Ratio of weighted average wage in NP relative to P, initial period λ1 Ratio of weighted average wage under the new scale relative to the existing, P λ2 Ratio of weighted average wage under the new scale relative to the existing, NP Option 1.Maximum salary adjustment 6. Under this option, salaries within each grade is raised to the highest observed for the grade. Given the assumptions above, the total wage bill and its growth rate under this option are therefore obtained as follows. Table D.6 Maximum salary adjustment Year Total wage bill under Option 1 Growth rate of wage bill under Option 1 2012 TWBt = N tPWt P + N tNPWt NP = N tPWt P (1 + αβ ) 2013 1 P P TWBt +1 [2 + λ1(1 + 3g ) + (αβ )(λ2 + 2)] − 1 TWB t +1 = N t Wt [2 + λ1 (1 + 3 g ) + (αβ )(λ 2 + 2 )] −1 = 3 TWBt 3(1 + αβ ) 2014 TWBt + 2 = 1 P P 3 [ ( ) N t Wt 1 + λ1 2 + 6 g + 3 g 2 + (αβ )(2λ2 + 1) ] TWBt + 2 −1 = [ ( ) 1 + λ1 2 + 6 g + 3g 2 + (αβ )(2λ 2 + 1) −1 ] TWBt +1 [2 + λ1 (1 + 3g ) + (αβ )(λ2 + 2)] 2015 [ TWB t + 3 = N tPWt P λ1 (1 + g )3 + αβλ 2 ] TWB t +3 −1= [ 3 λ1 (1 + g )3 + αβλ 2 −1 ] TWB t + 2 [ ( ) 1 + λ1 2 + 6 g + 3 g 2 + (αβ )(2λ 2 + 1) ] 2016 [ TWB t + 4 = (1 + π )N tPWt P λ1 (1 + g )4 + αβλ 2 ] TWBt + 4 −1 = [ (1 + π ) λ1 (1 + g )4 + αβλ 2 − 1 ] TWBt +3 [ λ1 (1 + g )3 + αβλ 2 ] 2017 2 [ 5 TWB t + 5 = (1 + π ) N tPWt P λ1 (1 + g ) + αβλ 2 ] TWBt +5 −1 = [ (1 + π ) λ1 (1 + g )5 + αβλ 2 − 1 ] TWBt + 4 [ λ1 (1 + g )4 + αβλ 2 ] 142 Option 2. Average salary adjustment 7. Under this option, salaries of those below the current average within each grade are raised to the average. Salaries of those above the current average are assumed to frozen during the adjustment period (which lasts, presumably, until the inequality within the grade is diminished). The adjustment will be completed in the first year (t+1). 8. Given these and other assumptions described above, the total wage bill and its growth rate under this option are therefore obtained as follows. Table D.7 Average salary adjustment Year Total wage bill under Option 2 Growth rate of the wage bill under Option 2 201 TWBt = N tPWt P + N tNPWt NP = N tPWt P (1 + αβ ) 2 201 3 TWB t +1 = ∑ max [w i k t +1 ] , wti + gN tP Wt P Simulated where i = 1L N , N = N tP + N tNP wtk+1 = the reference salary for pay grade k wti = employee i in pay grade k’s salary at time t 201 4 TWBt + 2 = ∑ max [w i k t +1 (1 + π ) , wti ]+ g (1 + g )(1 + π )N tPWt P Simulated 201 TWBt + 3 = ∑ max [w k ( t +1 1 + π 2 ) ] , wti + g (1 + g )2 (1 + π )2 N tPWt P Simulated 5 i 201 6 TWBt + 4 = ∑ max[w i k t +1 (1 + π )3 , wti ]+ g (1 + g )3 (1 + π )3 N tPWt P Simulated 201 TWB t + 5 = ∑ max [w k t +1 (1 + π )4 , wti ]+ g (1 + g )4 (1 + π )4 N tPWt P Simulated 7 i Option 3: Reference salary adjustment 9. Under this option, salaries are adjusted to the reference scale proposed by the unions, as shown in Table B4. The costs of adjustments are spread over two years, as described above. Given the other assumptions also described above, the total wage bill and its growth rate under this option are therefore obtained as follows. 143 Table D.8 Reference salary adjustment Year Total wage bill under Option 3 Growth rate of the wage bill under Option 3 2012 TWBt = N tPWt P + N tNPWt NP = N tPWt P (1 + αβ ) 2013 1 P P TWBt +1 [1 + λ1 (1 + 2 g ) + (αβ )(λ 2 + 1)] TWBt +1 = N t Wt [1 + λ1 (1 + 2 g ) + (αβ )(λ2 + 1)] = −1 2 TWBt 2(1 + αβ ) 2014 [ TWB t + 2 = N tPWt P λ1 (1 + g )2 + αβλ 2 ] TWBt + 2 −1 = [ 2 λ1 (1 + g )2 + αβλ2 ] −1 TWBt +1 [1 + λ1 (1 + 2 g ) + (αβ )(λ2 + 1)] 2015 [ TWB t + 3 = (1 + π )N tPWt P λ1 (1 + g )3 + αβλ 2 ] TWBt + 3 −1 = [ (1 + π ) λ1 (1 + g )3 + αβλ 2 − 1 ] TWBt + 2 [ λ1 (1 + g )2 + αβλ 2 ] 2016 [ TWB t + 4 = (1 + π )2 N tPWt P λ1 (1 + g )4 + αβλ 2 ] TWBt + 4 −1= [ (1 + π ) λ1 (1 + g )4 + αβλ 2 − 1 ] TWBt +3 [ λ1 (1 + g )3 + αβλ 2 ] 2017 [ TWB t +5 = (1 + π )3 N tPWtP λ1 (1 + g )5 + αβλ 2 ] TWBt +5 −1 = [ (1 + π ) λ1 (1 + g )5 + αβλ 2 − 1 ] TWBt + 4 [ λ1 (1 + g )4 + αβλ 2 ] Option 4: Combination of Option 2 and 3 10. Under this option, salaries are adjusted to the reference levels that lie somewhere between the average (Option 2) and the reference salaries proposed by the union (Option 3), as shown in Table B5. The adjustment will be completed in the first year (t+1). Given all the assumptions described in the outset, the total wage bill and its growth rate under this option are therefore obtained as follows. Table D.9 Combined adjustment Year Total wage bill under Option 4 Growth rate of the wage bill under Option 4 2012 TWBt = N tpWt P + N tNPWt NP = N tPWt P (1 + αβ ) 2013 +1 + N t Wt +1 = [λ1 (1 + g ) + (αβλ 2 )]N t Wt TWBt +1 = (1 + g )N tPWt P NP NP P P TWBt +1 [λ (1 + g ) + (αβλ2 )] − 1 −1 = 1 TWBt (1 + αβ ) 2014 [ ] TWBt + 2 = λ1 (1 + g )2 (1 + π ) + αβλ 2 (1 + π ) N tPWt P TWBt + 2 [ λ (1 + g )2 (1 + π ) + αβλ2 (1 + π ) −1 = 1 −1 ] TWBt +1 [λ1(1 + g ) + αβλ2 ] 2015 [ ] TWBt + 3 = λ1 (1 + g )3 (1 + π )2 + αβλ 2 (1 + π )2 N tPWt P TWBt + 3 [ λ (1 + g )3 (1 + π )2 + αβλ 2 (1 + π )2 −1 = 1 −1 ] TWBt + 2 [ λ1 (1 + g )2 (1 + π ) + αβλ 2 (1 + π ) ] 2016 [ 4 3 TWBt + 4 = λ1 (1 + g ) (1 + π ) + αβλ 2 (1 + π ) 3 ] N tPWt P TWBt + 4 −1 = 1 [ 4 3 λ (1 + g ) (1 + π ) + αβλ 2 (1 + π ) ] −1 3 TWBt + 3 [ ] λ1 (1 + g )3 (1 + π )2 + αβλ 2 (1 + π )2 2017 [ ] TWBt + 5 = λ1 (1 + g )5 (1 + π )4 + αβλ 2 (1 + π )4 N tPWt P TWBt + 5 −1 = [λ (1 + g ) (1 + π ) + αβλ (1 + π ) ] − 1 1 5 4 2 4 TWBt + 4 [λ (1 + g ) (1 + π ) + αβλ (1 + π ) ] 1 4 3 2 3 144 Table D.10 Civil Service Census April 2008 Category Number Staff paid 58,493 Staffed counted 64,719 Of which, retained 64,553 Of which, not retained 166 Staff paid and counted 56,492 Of which, primary and secondary education 41,523 Public health 5,523 Justice 2,871 National education 1,796 Agriculture 1,720 Staff paid but not counted 1,801 Of which, primary and secondary education 1,123 Health 244 Staff counted and not paid 7,861 Source: World Bank (2008) 145 Table D.11 Public Staffing and Salaries, January 2012 Gross pay/month Average gross Profession Category Grade Numbers FBu (thousands) pay/month FBu I 8 924,000 115,500 II 32,542 4,296,780,596 132,038 III 11,532 1,723,226,760 149,430 IV 5,637 1,078,132,620 191,260 Collaboration V 2,604 626,376,576 240,544 VI 1,704 487,567,224 286,131 VII 1,067 349,553,468 327,604 VIII 603 243,694,008 404,136 IX 143 64,413,063 450,441 I 4 658,000 164,500 II 1,767 355,149,330 200,990 III 848 216,686,896 255,527 IV 360 112,150,080 311,528 Direction V 177 63,042,975 356,175 Teachers VI 95 39,765,955 418,589 VII 80 37,937,040 474,213 VIII 14 7,413,000 529,500 IX 0 0 I 5 447,000 89,400 II 214 17,769,062 83,033 III 455 52,103,870 114,514 IV 564 72,617,256 128,754 Execution V 185 29,291,050 158,330 VI 15 6,187,995 412,533 VII 20 4,632,000 231,600 VIII 27 7,915,995 293,185 IX 16 4,998,000 312,375 Indeterminate 1,103 190,270,809 172,503 Total Teachers 61,789 10,089,711,177 163,293 I 2,199 612,604,017 278,583 II 558 151,327,926 271,197 III 302 110,098,932 364,566 IV 297 109,747,143 369,519 Collaboration V 145 46,599,955 321,379 VI 100 46,845,000 468,450 VII 32 14,972,000 467,875 VIII 4 1,755,000 438,750 Non- IX 1 301,000 301,000 Teachers I 1,145 419,440,980 366,324 II 129 36,401,994 282,186 III 94 30,731,044 326,926 IV 96 33,549,024 349,469 Direction V 54 22,348,008 413,852 VI 13 5,646,004 434,308 VII 3 2,685,000 895,000 VIII 0 0 IX 0 0 146 I 2,701 640,147,804 237,004 II 793 179,375,014 226,198 III 290 89,210,960 307,624 IV 166 48,072,936 289,596 Execution V 130 34,855,990 268,123 VI 63 20,872,026 331,302 VII 8 2,540,000 317,500 VIII 1 228,000 228,000 IX 0 0 Indeterminate 462 69,009,864 149,372 Total Non-Teachers 9,786 2,729,364,330 278,905 Magistrates 1,436 502,564,100 349,975 Governors 16 3,862,000 241,375 Other Pensioners 4,038 215,641,314 53,403 Contract Employees 8,790 616,460,280 70,132 Total Other 14,280 1,338,527,694 169,031 Agents 12,519 1,519,168,131 121,349 Brigadiers 2,456 516,044,896 210,116 Police Officers 1,154 405,696,778 351,557 Civilians 592 48,677,200 82,225 Total Police 16,721 2,489,589,690 148,890 Officer 2,025 456,846,075 225,603 Senior Officer 5,045 812,966,435 161,143 Military Troupes 19,923 1,895,434,374 95,138 Civilians 1,008 75,363,120 74,765 Total Military 28,001 3,240,611,732 115,732 Total 130,577 19,887,804,623 152,307 Source: Garnett, Nganou, and Brown (2012); January, 2012 data, monthly remuneration 147 Table D.12 Salaries by Ministry, 2012 Ministry Staff Average Salary (FBu) Health 6,122 315,506 Good Governance 12 315,128 Higher Education 90 179,704 Army 26,914 177,501 Communications 21 173,195 Justice 1,193 155,550 Foreign Affairs 121 154,558 Commerce 114 152,111 Police 15,990 145,487 Public Service 255 144,982 Energy 88 144,328 Development 77 130,806 Water 235 130,497 National Solidarity 107 128,322 Finance 328 127,712 East African 49 125,648 Youth 103 124,111 Agriculture 512 121,666 Education 54,360 119,373 Transport 87 118,666 Presidency 8 117,344 Interior 361 110,833 Source: Government of Burundi (2012) 155 Table D.13 List of Bonuses and Allowances Responsibility. This allowance is currently paid to managers in a select few ministries, such as Justice, and Primary and Secondary Education. A new ordinance is being prepared which will provide allowances of 30,000 FBu, 25,000 FBu, 15,000 FBu, and 7,000 FBu for Directors- General, Directors, Heads of Service and Secretaries respectively in all ministries; Efficiency. This allowance is paid to engineers in the Ministry of Planning and all staffs in the Ministry of Finance. It is paid in cash outside the payroll; Retention. This allowance is paid to public servants in the “Direction” and “Collaboration” categories in the Ministry of Primary and Secondary Education, and to non-magistrates in the Ministry of Justice. The amounts range from 7,000 FBu to 20,000 FBu per month; Risk. This allowance, introduced in 2008, is paid to all staffs in the Ministry of Public Health at the rate of 50 per cent of base salary; Loyalty. This allowance provides one month’s tax free salary for all staffs in the Ministry of Primary and Secondary Education after 15 years of service. Three months of tax free salary is provided after 30 years of service; Representation. This allowance is currently paid to Directors-General at the rate of 10,000 FBU/month. There are plans to increase the rate to 25,000 FBu/month; Transport. This allowance is paid at a standard rate equivalent to 9,680 FBu/month for every public servant; Housing. This allowance is paid in all ministries at the rates of 50,000 FBu/month, 30,000 FBu/month, and 20,000 FBu/month for staffs in the Direction, Collaboration and Execution categories respectively; Cash. This allowance is paid to public servants who are responsible for handling cash; Unsocial hours. This allowance, which was introduced this year under the overtime provision in the 2006 Law, is paid at the rates of 25,000 FBu/month, 17,500 FBu/month, and 11,500 FBu/month to employees in the Direction, Collaboration and Execution categories in the Ministry of Public Health. Source: Garnett, Nganou, and Brown (2008) 156 Annex E : Budget classification and data quality issues The use of BOOST database 1. The budget analysis presented in Chapter 3 draws on data obtained from the version 1.0 of the Burundi BOOST database. A BOOST database is an Excel file collecting expenditure data from the treasury systems at the budget line level. Once the data is collected, the members of the BOOST team categorize each line by economic and administrative categories and, if available, by functional and program classifications. In countries following the French public accounting system, the BOOST team tries to collect all the steps of the expenditure chain (chaine de la depense). For other countries, the Team collects original budget estimates, modifications, and actual payments. 2. In the case of Burundi, the BOOST Team extracted data from SIGEFI for all the budget lines (regular budget, pro-poor, PPTE) of all the administrative units of the central government for the period 2008-11. Validation with official sources 3. The data obtained matches the figures approved by the Burundi National Assembly in the Lois de Reglement of the sample period. However, and contrary to other BOOST databases, the team was not able to reconcile its figures with those published by the IMF in the latest Article IV fiscal operations table. Data quality 4. An additional limitation of the data concerns the quality of the data recorded in SIGEFI itself. Counterparts in the Ministry of Finance mentioned to the BOOST Team that some commitments are still done manually and hence not recorded in SIGEFI. What this means in practice is that not all the expenditure operations are recorded in SIGEFI. The analytical implications of this problem for the PER are that our estimates on execution rates may be biased downwards. It goes without saying, the reporting and fiduciary risks of having commitments done manually could be tremendous, but that goes beyond the scope of this document. Budget classification 5. A final limitation concerns the budget classification itself. The Burundi authorities have changed the economic classification three times since 2005: the first time took place in 2005, when the new Plan Budgetaire et Comptable de l’Etat was adopted. This nomenclature covers the period 2008-09. The second time the Burundi authorities changed the economic classification was in 2008, when the 2005 Plan Comptable was modified. As the 2008 modification was implemented in 2010, our sample period has two different economic classifications: one for 157 2008-09, and one for 2010-11. The BOOST Team created a set of bridge tables trying to reconcile the two classifications, available upon request. 6. Though the Burundi Plan Comptable of 2005 mentions the existence of a functional classification, the reality is that SIGEFI is unable to produce reports linking each budget line with a given function. It is expected that the adoption of a new Plan Comptable in 2014, as well as the introduction of a new financial management system will solve, once and for all, the issue of budget classification. Outstanding challenges with Budget classification in Burundi 7. In 2005, through a Presidential Decree (100/168, issued on December 31st, 2004), the Burundi Government adopted a new Plan Comptable. One of the objectives of this document was to introduce the GFS 2001 Manual of public accounting to Burundi’s public financial management systems, as well as a functional classification. This document was modified in 2008, with the introduction of a new Plan Comptable, which allowed a better treatment of arrears, and the introduction of new codes for an improved debt management. 8. In 2010, the Government introduced a completely new chart of accounts aiming to harmonize the plan comptable and the nomenclature budgétaire. This new chart of accounts was used for the first time in the 2013 budget and was endorsed by the IMF. The changes in budget and accounting classifications have taken place at the same time as the ministerial reorganization. Despite this progress, several challenges remain in the budget classification system, and namely: • The functional classification is not operational. Though the 2005 Plan Comptable introduced the notion of functional classification to Burundi’s public accounting system, in practice the budget codes do not include a function identifier. Even though the 2010 Plan Comptable reinforces the calls for the introduction of a functional classification, the 2013 budget only incorporates the new administrative and economic classifications. The lack of a functional classification has two implications: first, without a functional classification, it is not possible to assess effectively how much is allocated to a given sector. The second effect is specific to Burundi. The new accounting of pro-poor expenditures relies on a functional classification.1 The SIGEFI functionalities to incorporate the functional classification must be activated immediately. • The budget codes do not identify the CSLP axis. Though most government documents refer to the CSLP as the framework to bring prosperity to Burundi and increase the standard of living of the Burundi people, the reality is that it is not possible to say how much is spent on each axis. Budget codes do not include a CSLP identifier. Though the 2013 MTEF tries to identify expenditures by CSLP axis, officials in the Ministry of Finance recognized that the identification of ministries (ADMIN1 level) and directions (ADMIN2 level) was done manually. This obviously leaves room for improvisation, human error, or even mismanagements, and reduces accountability. This problem is enhanced by the fact that the CNCA does rely on the CSLP axis to classify projects funded or financed by donors. The bottom line is that it is possible to know how much each donor devotes to every CSLP axis, 158 but not how much Government does. The fact that expenditures were identified by CSLP axis for the first time in the 2013 MTEF is an improvement, but steps should be done to incorporate a CSLP dimension to the budget codes. • There are no bridge tables between the 2005 Plan Comptable, the 2008 modification, and the 2010 Plan Comptable. Though the changes in the chart of accounts are necessary to improve expenditure recording and execution, they have not been complemented by subsequent efforts to create a set of ‘bridge tables’ linking the different charts of accounts. This is likely to result in a loss of institutional memory, while also making difficult to compare budget data over time. For the purposes of the PER analysis, the team – in collaboration with the Government - built a bridge table, linking the codes of the economic classification for the 2005 and 2008 versions of the chart of accounts. Sources of discrepancies between SIGEFI and TOFE - problems specific to a given budget year • In 2009, the Ministry of Finance authorized a FBu 3.7 billion payment to ONATEL (the State-owned phone company) outside of the chaine de la dépense. The purpose of the operation was to compensate ONATEL in advance as the company was facing liquidity constraints during that year. • In 2010, transfers in SIGEFI were higher than planned in the original budget because of the establishment of the Electoral Authority (Commission Electorale Nationale Indépendante – CENI) and the OBR. In addition, an unforeseen payment of FBU 3.7 billion was made to the East African Community. Though a collectif budgetaire should have been approved in principle to take these operations into account, the Government decided not to do it. • In 2011, the Government of Burundi made a payment in advance of FBu 8 billion to the EAC outside of the chaîne de la dépense. • In 2011, personnel payments fall dramatically from liquidation onwards. Public officials in the Ministry of Finance attribute this to the introduction of the staff listing, aimed to have a better control on payroll payments. The listing introduces a clear distinction between the different types of salaries (i.e., base salary and allowances) received by staff members and the withholdings (taxes, contributions to trade unions, and re-payments of cash advances) on each one of them. As the payroll listing is done at the end of the month, the staff in charge of payroll in the Ministry of Public Functionning (Ministère de la Fonctin Publique) sends a list with the net payments (i.e., after withholdings) to finance for processing. Once the listing is received in the Ministry of Finance, the SIGEFI technicians process the payments and enter it into the systems on a net basis (i.e., after withholdings). It is unclear the extent to which this will be a problem in the future and what the Burundi authorities are doing to fix it. Preliminary data from SIGEFI for 2012 show a much lower liquidated amount for salary payments with respect to engagements, but not the extent of 2011. 159 Table E.1 Personnel payments, 2012 Collectif LFI budgétaire Crédit Engagement Liquidation Ordonnancement Paiement Personne l payments 279.7 288.7 286.6 282.2 254.5 253.7 253.8 Sources : Loi de Finances Initiale 2012, Collectif Budgétaire 2012, and SIGEFI 160 Annex F : Public expenditure Analysis- background tables Table F.1 Comparison of budget executions by Source (billions of FBu) 2009 Loi de Collectif SIGEFI Loi de Execution ratios Finances budgétaire Reglement Loi de Engagemen Ordonnanc Ordonnanc Paiement/L Reglement/ Credit t Liquidation ement Paiement ement/LFI FI LFI Current Expenditures 387.8 422.2 423.7 417.8 406.0 389.8 403.4 421.0 100.5 104.0 108.6 Personnel 163.4 186.7 186.2 185.1 179.8 178.6 177.3 183.9 109.3 108.5 112.5 Goods and Services 97.6 102.3 103.5 104.1 103.8 99.7 99.5 107.2 102.1 102.0 109.8 Interests 33.7 32.6 33.2 28.3 22.3 11.5 26.8 28.3 34.1 79.6 83.9 Transfers 93.0 100.6 100.7 100.4 100.2 100.0 99.7 101.5 107.5 107.2 109.2 Imprévus 1.4 1.4 Capital Expenditures 93.8 81.6 81.7 58.1 57.0 54.9 54.2 59.5 58.5 57.8 63.4 482.9 Total (Curr. Exp. + Imprévus + Cap. Exp.) 505.3 505.3 475.9 463.0 444.6 457.6 480.4 159.1 161.9 172.0 2010 Loi de SIGEFI Loi de Execution ratios Finances Reglement Loi de Engageme Ordonnanc Ordonnanc Paiement/L Reglement/ Credit nt Liquidation ement Paiement ement/LFI FI LFI Current Expenditures 468.7 488.2 496.4 480.7 464.1 472.4 506.2 99.0 100.8 108.0 Personnel 212.4 209.5 233.8 220.8 219.6 219.3 233.9 103.4 103.2 110.1 Goods and Services 98.0 105.9 100.4 100.1 95.5 95.3 105.7 97.4 97.3 107.9 Interests 30.2 27.7 27.7 26.0 17.0 26.0 26.0 56.1 86.2 85.9 Transfers 128.0 145.0 134.6 133.8 132.0 131.8 140.5 103.1 102.9 109.8 Imprévus 8.1 Capital Expenditures 107.2 107.2 79.6 76.9 74.1 72.5 76.9 69.1 67.6 71.7 584.0 Total (Curr. Exp. + Imprévus + Cap. Exp.) 595.4 576.0 557.6 538.1 544.9 583.1 168.1 168.4 179.7 2011 Loi de SIGEFI Loi de Execution ratios Finances Reglement Loi de Engageme Ordonnanc Ordonnanc Paiement/L Reglement/ Credit nt Liquidation ement Paiement ement/LFI FI LFI Current Expenditures 537.1 544.6 529.9 451.1 425.3 435.3 551.9 79.2 81.0 102.8 Personnel 272.4 272.2 259.1 181.4 170.7 170.4 272.8 62.7 62.6 100.1 Goods and Services 98.1 103.7 97.9 97.4 93.8 93.6 99.2 95.6 95.4 101.1 Interests 35.6 33.5 32.9 32.3 22.6 33.6 40.2 63.4 94.6 113.0 Transfers 131.0 135.1 140.0 140.0 138.3 137.6 139.7 105.6 105.0 106.6 Imprévus 5.0 Capital Expenditures 127.1 124.6 106.8 106.8 92.2 91.8 106.8 72.6 72.3 84.1 669.2 Total (Curr. Exp. + Imprévus + Cap. Exp.) 669.2 636.8 557.9 517.6 527.1 658.8 151.8 153.3 186.8 Sources: Lois de Finances Initiale 2009-2011, Collectif Budgétaire 2009, Lois de Réglèment 2009-2011, SIGEFI, and authors’ own calculations. 161 Table F.2 Expenditure Chain, 2009 Loi de Execution Rate Finances Credits (Ordonnancement Row Labels Initiale Ouverts Engagement Liquidation Ordonnancement Paiement /LFI) Deuxième Vice-Présidence 1.09 1.17 1.17 1.16 1.12 1.12 102.60 Ministére à la Présidence Chargé de la Bonne Gouvernance et de la Privatisation 2.33 2.33 2.27 2.26 2.20 2.19 94.45 Ministère de la Défénse Nationale et des Anciens Combattants 67.34 69.92 68.12 68.12 66.80 66.84 99.18 Ministère de la Fonction Publique, du Travail et de la Sécurité Sociale 2.68 18.34 24.61 24.54 23.70 22.83 882.69 Ministère de la Jeunesse , des Sport et de la Culture 2.14 2.22 2.20 2.19 2.14 2.13 99.71 Ministère de la Justice et des Gardes Sceaux 15.05 15.59 15.78 14.59 14.38 14.30 95.52 Ministère de la Santé Publique 40.52 43.91 42.41 41.60 41.48 41.35 102.35 Ministère de la Sécurité Publique 47.16 49.04 48.26 48.26 47.50 47.51 100.72 Ministère de la Solidarité Nationale, du Rappatriement, des Réfugiés et de la Réintegration Sociale 6.43 6.37 5.79 5.65 5.50 5.50 85.54 Ministère de l'Agriculture et de l'Elèvage 19.94 16.86 12.75 12.61 11.29 11.29 56.61 Ministère de l'Eau, Environnement , Aménagément du Territoire et de l'Urbanisme 4.32 7.31 5.32 4.84 4.51 4.50 104.10 Ministère de l'Education Nationale et de la Recherche Scientifique 26.85 27.62 24.17 24.03 23.90 23.90 89.00 Ministère de l'Energie et des Mines 14.07 8.36 4.99 4.94 4.56 4.48 32.41 Ministère de l'Information, Communication, Relations avec le Parlement et Porte Parole du Gouvernement 4.43 4.33 4.08 4.07 3.90 3.90 88.06 Ministère de l'Intérieur et du Developpement Communal 2.10 4.17 4.18 4.09 3.89 3.89 184.89 Ministère des Relations Extérieures 15.22 19.44 19.32 19.20 19.05 19.04 125.08 Ministère des Transports, Postes et Télécommunications 3.30 2.99 2.76 2.74 2.70 2.70 81.71 Ministère des Travaux Publics et de l'Equipement 21.16 15.04 13.38 13.35 13.26 12.62 62.65 Ministère du Commerce, de l'Industrie et du Tourisme 1.37 1.32 1.21 1.19 1.14 1.13 83.16 Ministères de l'Economie, des Finances et de la Cooperation au Developpement 112.36 112.24 102.00 88.53 53.61 78.21 47.71 Prémière Vice-Présidence 1.74 1.78 1.79 1.78 1.72 1.72 98.29 Vice-Minisitre Charge du Developpement Comunal 0.54 0.85 0.84 0.83 0.83 0.83 151.83 Vice-Ministere Charge de la Lutte contre le SIDA 2.29 2.29 2.27 2.26 2.26 2.26 98.39 Vice-Ministere Charge de la Planification 2.05 2.07 1.96 1.94 1.92 1.91 93.48 Vice-ministère Charge de l'Enseignement de Base et Sécondaire 107.42 109.53 99.47 96.11 96.09 96.09 89.45 Vice-ministère Charge des Droits de la Personne Humaine et du Genre 0.43 0.42 0.39 0.39 0.39 0.39 89.13 Vice-ministère des Affaires de la Communauté Est Africaine 0.96 0.92 0.81 0.81 0.80 0.80 83.04 Présidence de la République 14.78 18.37 18.40 18.30 17.80 17.80 120.39 Parlement 14.03 13.90 13.78 13.74 13.45 13.46 95.82 Legend 5 Lowest under-executions. Stages of the chain of expenditure exceeding the appropriation authorized in the Loi de Finances 5 Highest over-executions. 162 Table F.3 Expenditure Chain, 2010 Loi de Finances Credits Execution Rate Initiale Ouverts Engagement Liquidation Ordonnancement Paiement (Ordonnancement/LFI) Deuxième Vice-Présidence 1.07 1.82 1.76 1.75 1.29 1.29 120.58 Ministére à la Présidence Chargé de la Bonne Gouvernance et de la Privatisation 2.50 2.52 2.50 2.50 2.49 2.49 99.89 Ministère de la Décentralisation et du Dévéloppement Communal 1.74 1.74 1.72 1.72 1.68 1.68 96.52 Ministère de la Défénse Nationale et des Anciens Combattants 70.38 74.19 71.54 68.10 65.30 65.31 92.79 Ministère de la Fonction Publique, du Travail et de la Sécurité Sociale 4.66 4.67 5.66 5.58 5.55 5.51 118.93 Ministère de la Jeunesse , des Sport et de la Culture 2.07 2.13 2.08 2.06 2.04 2.04 98.64 Ministère de la Justice et des Gardes Sceaux 16.78 17.17 15.45 14.24 13.65 13.56 81.34 Ministère de la Lutte Contre le SIDA 2.57 2.57 2.44 2.34 2.34 2.35 90.87 Ministère de la Planification et de la Reconstruction 3.44 3.44 2.96 2.95 2.94 2.93 85.40 Ministère de la Santé Publique 63.51 63.51 60.94 59.80 59.58 59.51 93.80 Ministère de la Sécurité Publique 44.88 45.94 45.56 42.96 42.33 42.38 94.31 Ministère de la Solidarité, du Rappatriement, des Réfugiés et de la Réintegration Sociale 9.31 9.31 9.11 7.58 7.53 7.53 80.84 Ministère de l'Agriculture et de l'Elèvage 18.04 18.04 14.67 13.85 13.61 13.63 75.45 Ministère de l'Eau, Environnement , Aménagément du Territoire et de l'Urbanisme 12.52 12.52 11.04 10.74 10.26 9.28 81.96 Ministère de l'Energie et des Mines 17.27 17.27 14.61 14.58 13.99 13.87 80.96 Ministère de l'Enseignement des Métiers, De la Formation Professionnelle et de l'Alphébétisation des Adultes 4.21 4.21 3.88 3.70 3.50 3.50 82.99 Ministère de l'Enseignement Primaire et Sécondaire 136.46 136.46 146.21 141.93 140.03 140.03 102.61 Ministère de l'Enseignement Supérieur et de la Recherche Scientifique 30.56 30.56 28.72 28.62 28.62 28.62 93.65 Ministère de l'Information, de la Communication et des Relations avec le Parlement 4.43 4.49 4.48 4.48 4.35 4.36 98.26 Ministère de l'Intérieur 4.06 11.00 10.59 10.56 10.55 10.55 260.03 Ministère des Affaires de la Communauté Est Africaine 1.42 1.54 1.36 1.35 1.33 1.33 93.51 Ministère des Droits de la Personne Humaine et du Genre 0.38 0.70 0.66 0.66 0.65 0.65 172.14 Ministère des Finances 141.61 121.83 115.42 110.01 75.78 88.15 53.52 Ministère des Relations Extérieures et de la Coopération Internationale 20.84 25.80 25.08 25.05 24.92 24.95 119.58 Ministère des Transports, Postes et Télécommunications 1.70 1.70 1.09 1.08 1.07 1.07 63.01 Ministère des Travaux Publics et de l'Equipement 20.93 20.93 14.86 14.82 14.66 14.07 70.06 Ministère du Commerce, de l'Industrie et du Tourisme 2.72 2.72 2.21 2.18 2.03 2.01 74.52 Prémière Vice-Présidence 2.77 3.10 2.66 2.66 2.53 2.53 91.29 Présidence de la République (y compris le Secretariat General du Gouvernement) 18.96 19.90 18.93 18.91 17.61 17.62 92.89 Parlement (Senat + Assemblee Nationale + Cour des Comptes) 13.94 13.94 13.09 13.02 12.97 12.98 93.07 Legend 5 Lowest under-executions. Stages of the chain of expenditure exceeding the appropriation authorized in the Loi de Finances 5 Highest over-executions. 163 Table F.4 Expenditure chain, 2011 Loi de Finances Credits Execution Rate Row Labels Initiale Ouverts Engagement Liquidation Ordonnancement Paiement (Ordonnancement/LFI) Deuxième Vice-Présidence 1.06 1.26 1.22 1.22 1.13 1.13 106.67 Ministére à la Présidence Chargé de la Bonne Gouvernance et de la Privatisation 2.42 2.47 2.45 2.40 2.39 2.39 98.45 Ministère de la Défénse Nationale et des Anciens Combattants 79.67 81.56 81.36 76.01 74.54 74.66 93.57 Ministère de la Fonction Publique, du Travail et de la Sécurité Sociale 24.65 24.65 21.75 21.49 11.72 11.70 47.54 Ministère de la Jeunesse , des Sport et de la Culture 1.98 2.08 2.01 1.90 1.88 1.87 94.60 Ministère de la Justice et des Gardes Sceaux 15.98 15.98 14.43 11.68 11.33 11.25 70.91 Ministère de la Santé Publique et de la Lutte contre le SIDA 72.36 72.36 70.08 58.38 56.06 55.66 77.47 Ministère de la Sécurité Publique 56.25 56.75 54.93 50.61 48.76 48.74 86.69 Ministère de la Solidarité Nationale, des Droits de la Personne Humaine et du Genre 8.18 8.18 7.74 7.64 7.57 7.57 92.46 Ministère de la Telecommunication, de l'Information, de la Communication et des Relations avec le Parlement 3.97 4.00 3.89 3.87 3.84 3.84 96.74 Ministère de l'Agriculture et de l'Elèvage 43.22 43.22 34.36 33.30 31.57 31.70 73.04 Ministère de l'Eau, de l'Environnement , de l'Aménagément du Territoire et de l'Urbanisme 10.50 10.52 9.26 9.04 8.79 8.72 83.71 Ministère de l'Energie et des Mines 16.31 16.31 14.69 14.59 8.21 7.75 50.32 Ministère de l'Enseignement de Base et Secondaire, de l'Enseignement des Metiers, de la Formation Professionnelle et de l'Alphabetisation 148.40 148.43 139.83 88.94 87.24 87.24 58.79 Ministère de l'Enseignement Supérieur et de la Recherche Scientifique 27.45 27.47 26.73 26.59 26.22 26.09 95.51 Ministère de l'Intérieur 3.39 3.44 2.76 2.44 2.42 2.42 71.36 Ministère des Affaires de la Communauté Est Africaine 1.59 2.49 2.45 2.41 2.40 2.40 151.44 Ministère des Finances 137.56 129.02 118.83 117.89 87.98 100.60 63.95 Ministère des Relations Extérieures et de la Coopération Internationale 25.20 25.57 33.35 33.20 32.73 32.78 129.89 Ministère des Travaux Publics et de l'Equipement 15.99 15.99 13.72 13.62 13.27 13.10 83.00 Ministère du Commerce, de l'Industrie des Postes et du Tourisme 2.34 2.34 2.00 1.88 1.71 1.70 73.08 Ministère du Plan et du Developpement Communal 10.38 10.96 9.61 9.51 7.10 7.10 68.44 Prémière Vice-Présidence 2.55 2.71 2.67 2.67 2.65 2.65 103.81 Présidence de la République 15.44 18.21 17.73 17.71 17.44 17.44 112.94 Parlement 12.17 13.04 12.57 12.57 12.29 12.29 101.00 Legend 5 Lowest under-executions. Stages of the chain of expenditure exceeding the appropriation authorized in the Loi de Finances 5 Highest over-executions. 164 Table F.5 Expenditure chain, 2012 (prel.) Loi de Finances Credits Execution Rate Row Labels Initiale Ouverts Engagement Liquidation Ordonnancement Paiement (Ordonnancement/LFI) Deuxième Vice-Présidence 1.06 1.74 1.73 1.72 1.72 1.72 162.91 Ministére à la Présidence Chargé de la Bonne Gouvernance et de la Privatisation 3.15 3.08 3.08 3.05 3.05 3.04 96.62 Ministère de la Défénse Nationale et des Anciens Combattants 89.56 91.04 91.20 87.55 85.69 85.53 95.68 Ministère de la Fonction Publique, du Travail et de la Sécurité Sociale 2.65 6.74 6.37 6.30 6.30 6.30 238.12 Ministère de la Jeunesse , des Sport et de la Culture 2.28 2.25 2.24 2.20 2.20 2.20 96.48 Ministère de la Justice et des Gardes Sceaux 18.08 18.05 18.04 16.66 16.66 16.64 92.15 Ministère de la Santé Publique et de la Lutte contre le SIDA 70.08 68.45 66.07 63.32 63.25 63.25 90.25 Ministère de la Sécurité Publique 58.78 58.94 56.19 51.14 50.90 51.13 86.60 Ministère de la Solidarité Nationale, des Droits de la Personne Humaine et du Genre 9.15 8.54 8.45 8.41 8.38 8.38 91.60 Ministère de l'Agriculture et de l'Elèvage 45.99 40.09 35.50 30.90 30.25 30.33 65.78 Ministère de l'Eau, Environnement , Aménagément du Territoire et de l'Urbanisme 9.57 8.54 6.55 6.53 6.53 6.53 68.22 Ministère de l'Energie et des Mines 25.85 25.15 19.34 19.33 19.29 19.65 74.62 Ministère de l'Enseignement de Base et Secondaire, de l'Enseignement des Metiers, de la Formation Professionnelle et de l'Alphabetisation 162.76 163.74 162.29 147.53 147.52 147.52 90.64 Ministère de l'Enseignement Supérieur et de la Recherche Scientifique 39.58 40.49 40.49 40.45 40.42 40.45 102.11 Ministère de l'Intérieur 3.60 3.60 2.96 2.91 2.91 2.91 80.95 Ministère des Relations Extérieures et de la Coopération Internationale 19.65 22.62 22.49 22.42 22.07 22.09 112.31 Ministère des Transports, des Travaux Publics et de l'Equipement 17.23 15.68 15.58 15.50 15.39 15.38 89.30 Ministère du Commerce, de l'Industrie et du Tourisme 2.58 2.46 2.20 2.13 2.13 2.13 82.54 Prémière Vice-Présidence 1.68 2.16 2.15 2.15 2.14 2.14 127.70 Vice-ministère des Affaires de la Communauté Est Africaine 1.59 1.80 1.78 1.77 1.76 1.76 110.80 Ministère du Developpement Communal 8.17 8.05 8.01 8.01 8.01 8.01 98.03 Ministère des Finances, de la Planification et du Developpement Economique 157.71 142.47 117.79 115.96 85.95 107.92 54.50 Ministère de la Telecommunication, de l'Information, de la Communication et des Relations avec le Parlement 4.56 4.46 4.37 4.37 4.37 4.37 95.73 Présidence de la République (y compris le Secretariat General du Gouvernement et l'Institution Ombudsman) 26.14 26.37 25.47 25.44 25.42 25.44 97.25 Parlement (Senat + Assemblee Nationale + Cour des Comptes) 13.43 13.39 13.26 13.25 13.25 13.25 98.65 Legend 5 Lowest under-executions. Stages of the chain of expenditure exceeding the appropriation authorized in the Loi de Finances 5 Highest over-executions. 165 Table F.6 Items that can be paid through special procedures 2008 2009 2010 2011 2012 Share of Credits Share of Share of Share of Share of opened Credits Credits Credits Credits Billion in Billion registere Billion registere Billion registere registere s of SIGEF s of d in s of d in s of d in Billions d in FBu I FBu SIGEFI FBu SIGEFI FBu SIGEFI of FBu SIGEFI Dette 74.372 16.680 86.540 14.954 87.995 13.022 83.357 11.279 97.277 12.473 Principal 45.186 10.134 53.337 9.217 60.268 8.919 49.844 6.744 59.177 7.588 Intérêts 29.186 6.546 33.204 5.738 27.727 4.103 33.513 4.535 38.100 4.885 Contribution aux Organisations Internationales 0.887 0.199 1.032 0.178 0.180 0.027 0.246 0.033 0.543 0.070 Autres 12.434 2.789 9.579 1.655 8.356 1.237 7.794 1.055 7.909 1.014 ACCUEIL DES DELEGATIONS ETRANGERES 0.040 0.009 0.060 0.010 0.060 0.009 0.053 0.007 0.048 0.006 ACHAT AZOTE LIQUIDE 0.006 0.001 0.003 0.001 0.008 0.001 0.008 0.001 0.007 0.001 ACHAT DES MEDICAMENTS ANTI-TUBERCULEUX 0.052 0.012 0.052 0.009 0.052 0.008 0.000 0.000 0.046 0.006 ACHAT MEDICAMENTS ARV 1.500 0.336 0.500 0.086 1.000 0.148 1.000 0.135 0.649 0.083 ALLOCATION A L'ASSISTANCE HUMANITAIRE D' URGENCE 1.171 0.263 1.121 0.194 1.121 0.166 0.997 0.135 1.094 0.140 ALLOCATION AUX ACTIVITES SPORTIVES ET COMPETITIONS INTERNATIONALES 0.689 0.155 0.862 0.149 0.929 0.138 0.870 0.118 0.985 0.126 CONFERENCES ET SOMMETS 0.120 0.027 0.282 0.049 0.375 0.055 0.100 0.014 0.135 0.017 DEPENSES RELATIVES AUX AFFAIRES CONTENTIEUSES 0.010 0.002 0.010 0.002 0.000 0.000 0.000 0.000 0.000 0.000 ENTRETIEN BATIMENT 0.003 0.001 0.001 0.000 0.003 0.000 0.003 0.000 0.003 0.000 ENTRETIEN DE L ' AVION DU GOUVERNEMENT 0.235 0.053 1.200 0.207 0.287 0.042 0.287 0.039 0.319 0.041 ENTRETIEN DES GROUPES ELECTROGENES DES HOPITAUX 0.014 0.003 0.023 0.004 0.000 0.000 0.000 0.000 0.000 0.000 ENTRETIEN ET REPARATION MACHINES DE BUREAU 0.009 0.002 0.004 0.001 0.009 0.001 0.009 0.001 0.008 0.001 FONCTIONNEMENT DES COMMISSIONS D'ENQUETE 0.159 0.036 0.056 0.010 0.056 0.008 0.050 0.007 0.050 0.006 FRAIS DE LOGEMENT ET RECEPTION POUR HOTES DE MARQUE 0.035 0.008 0.087 0.015 0.128 0.019 0.087 0.012 0.109 0.014 FRAIS DE MISSION A L'INTERIEUR DU PAYS 0.265 0.059 0.402 0.070 0.476 0.070 0.469 0.063 0.443 0.057 FRAIS DES RECEPTIONS OFFICIELLES 0.058 0.013 0.058 0.010 0.060 0.009 0.060 0.008 0.059 0.008 FRAIS FUNERAIRES 0.012 0.003 0.023 0.004 0.022 0.003 0.024 0.003 0.128 0.016 FRAIS INHERANTS A LA DEFENSE DES INTERETS DE L'ETAT 0.040 0.009 0.040 0.007 0.048 0.007 0.048 0.006 0.043 0.006 IMPREVUS 3.862 0.866 0.009 0.002 0.095 0.014 0.003 0.000 0.011 0.001 LUTTE CONTRE LES ENDEMIES 0.064 0.014 0.054 0.009 0.000 0.000 0.000 0.000 0.000 0.000 MAINTENANCE INFORMATIQUE 0.055 0.012 0.070 0.012 0.068 0.010 0.071 0.010 0.042 0.005 MISSION A L'INTERIEUR DU PAYS 0.066 0.015 0.090 0.016 0.089 0.013 0.089 0.012 0.140 0.018 MISSIONS OFFICIELLES A L'ETRANGER 2.600 0.583 2.845 0.492 1.485 0.220 2.108 0.285 2.190 0.281 RAPATRIEMENT DES ETUDIANTS 0.343 0.077 0.362 0.063 0.362 0.054 0.000 0.000 0.000 0.000 RENSEIGNEMENTS 0.200 0.045 0.150 0.026 0.200 0.030 0.200 0.027 0.179 0.023 SOINS DE SANTE 0.379 0.085 0.656 0.113 0.579 0.086 0.635 0.086 0.525 0.067 VALISE DIPLOMATIQUE 0.018 0.004 0.018 0.003 0.218 0.032 0.018 0.002 0.018 0.002 VOYAGES DES DIPLOMATES 0.430 0.096 0.541 0.094 0.625 0.093 0.605 0.082 0.680 0.087 105.72 Total 87.693 19.668 97.152 16.788 96.531 14.285 91.397 12.367 9 13.556 166 The SIGEFI: current functioning and pending challenges 1. The need for a formal system to manage public finance information and data emerged after the adoption of the 2004 budget and chart of accounts. With the support of the World Bank, IMF, the EU and other donors, the SIGEFI has been able to develop the following modules covering the expenditure cycle: • Commitment (“engagement”); • Validation (“liquidation”); • Authorization (“ordonnancement”); • Visa (given by the “Ordonnateur Trésorier du Burundi”); • Payment (“paiement”). 2. Up until recently, the connection between SIGEFI and the Central Bank of Burundi (BRB) was not working efficiently. So, the payment module mentioned above was not providing its full functionality. Since mid-2012, this connection has been working, thus following the expenditure cycle with the Ministry of Finance automatically giving the visa for processing transfer orders (“ordres de virement” – OV) and the BRB doing the payment to suppliers. 3. Since Burundi reached the decision point of the Heavily Indebted and Poor Country (HIPC) initiative in 2006, SIGEFI has been endowed with a HIPC expenditure module that works satisfactorily. The process of installing the civil servant payroll module in SIGEFI concretely started in 2009-2010. Even though the payroll of civil servant uses somehow the OPENPRH software since 2010-2011, its interfacing / connection with SIGEFI is not currently operational. Payroll operations are still carried out almost esntirelly manually between the Ministry of Finance and the Ministry of Civil Servants. 4. SIGEFI offers the possibility to get consolidated budgets, with current expenditures now managed simultaneously with capital expenditures. However, the public accounting (“comptabilité publique”) module is not operational as SIGEFI does not provide yet a comprehensive treatment of the public accounting system. Information and Technology Resources 5. The World Bank and IMF provided technical assistance and training to build up the capacity of the Ministry of Finance to operate an Information and Technology Department destined to manage and maintain the SIGEFI system. This assistance has been very instrumental to the implementation of the public finance management reform’s agenda. There are currently 3 IT engineers who ensure the maintenance of the SIGEFI system and undertake data management and processing tasks. Yet, there is still needs for external assistance from the service provider that developed SIGEFI, especially for medium to complex activities regarding the integration of new functionalities and adaptation to new PFM regulations. The securing and stabilization of the SIGEFI system have been enhanced through time since 2009-2010. Protected cables and high capacity servers have been supplied. But full compliance with security rules and back protocols still pose problems. 167 Critical Challenges 6. So far, the SIGEFI system has been able to incorporate the new procedures (e.g., accounting and budget procedures / regulations, new budget classification / nomenclature, etc.) and regulations resulting from the PFM reforms. For example, there is a bridging table (“tableau de passage”) developed in SIGEFI to ensure the consistency between the classification / nomenclature rules prevailing before 2008 and the new nomenclature issued in 2010, which has now been used in the 2013 budget law. 7. The capacity of SIGEFI is challenged inter alia in the following areas: • TOFE: this module has never been operational in a satisfactorily way. • Interfacing the payroll software to improve harmonization between the Ministry of Finance and the Ministry of Civil Servant and better manage recruitments. • Full integration of the legal and regulations framework: In particular, the implementation of the “Contrôleur des Engagements des Dépenses – CED” entails more responsibility devoted to ministries that have now to handle the budget commitment step of the budget cycle. • The issue of manual payments, explained above, is essential and needs to be solved, otherwise SIGEFI (or any other system) will be extremely limited and provide misleading information. 8. In fact, the main problem that SIGEFI could face down the road is its capacity to properly reflect and manage the new legal and regulations framework of the PFM reforms. However, if adequate resources can be mobilized, the extension of the SIGEFI’s functionalities, as well as additional terminals and capacity building to terminate manual payments, could be achieved in a cost-effective way. 9. For example, the implementation of the CED reform (decentralization of the commitment phase to the ministry level) could be efficiently handled with either a proper web-based interconnectivity (at least when piloting the reform) or a powerful broad-band, both technologies of which are available in Burundi. Normally, these approaches should save Government’s resources and time as they are likely to use the existing SIGEFI’s database with less risks, while still allowing for the existing SIGEFI to progressive comply with the actual legal and regulations framework. To illustrate this possibility, experts think that the codification of the new nomenclature / classification in SIGEFI, which is now being used in the 2013 budget law, would be more resource-demanding than just installing the new functionalities and algorithms required by the interfacing / connection of ministries to the Ministry of Finances to implement effectively the CED reform. 10. Another important aspect of the existing SIGEFI is the encryption and security protocols put in place to ensure the integrity of the existing database. This is the result of a protracted process of success and failures that illustrates well the intrinsic characteristic of reforms in Burundi whereby learning-by-doing plays a critical role. The related modules have been tested for years. However, independent technical audits have to be carried out to confirm the robustness of the security system of SIGEFI datasets. 168 Annex G : Aid management – supplementary analysis CNCA in aid coordination 1. Organizational structure and donor relations: The CNCA is designed to facilitate permanent relations with donors. Furthermore, depending on the priority field of involvement for each donor, the latter act as sector group leaders and fully participate in planning activities. The SP/CNCA is in regular contact with donors to collect data on engagements and disbursements. The new aid management platform implies direct data entry from each donor in the database. 2. Sector Group implementation: Drawing inspiration from the second generation Poverty Reduction Strategic Framework focus areas (Cadre Stratégique de Croissance et de Lutte Contre la Pauvreté/CSLP), the new aid management structure includes Sector Groups. Box 13 The aid architecture in Burundi: an overview 1. Intra-governmental Coordination Aware of the significance and the stakes of coordination, in 2006 the Government of Burundi established the Aid Coordination National Committee (CNCA) which gathers the four relevant ministries in aid management listed above. This committee is chaired by the Second Vice-President of the Republic. The REFES Permanent Secretariat (SP-REFES) is involved in the work of the Committee and is responsible for the preparation as well as the monitoring and evaluation of the Growth and Poverty Reduction Strategic Framework (Cadre stratégique de croissance et de lutte contre la pauvreté/CSLPII) and the production of various related analyses. At the CNCA Permanent Secretariat level, a program to enhance aid coordination and coordination mechanism operations has been developed and implemented since 2006. This program benefited from financial support via the trust fund established to this end and is managed by the UNDP. To this day, the fund has already benefited from funding by the UNDP, the DfID, the Netherlands, Switzerland and the European Union. 2. Government/Partners Dialog Framework With the establishment of the CNCA and its Permanent Secretariat (SP/CNCA), a persistent lack of dialog between the Government and its partners - namely technical and financial partners, civil society, and the private sector - was quickly observed. Thus, all actors collectively decided on the establishment of a dialog framework called Partner Coordination Group (Groupe de coordination des partenaires/GCP). The PCG was fitted with a three-tiered architecture, namely (i) the Political Forum acting as the top body meeting to respond to political orientation issues, (ii) the Strategic Forum, an intermediary level body which technically deals with all socio-economic development related issues, and (iii) the Sector (or Theme) Groups placed under the Monitoring and Evaluation Group umbrella overseeing sector group activities and dealing with in-depth meeting agenda preparation for both forums. The sector groups are in line with the CSLPII focus areas and in charge of sector management and coordination. 169 The SP-CNCA acts a secretariat for the overall GCP. The coordination mechanism is assessed each year and amended accordingly. This group architecture is presented below: Political Forum Strategic Forum Monitoring Group & Evaluation CSLP-II CSLP-II Pillar 1 CSLP-II Pillar 2 CSLP-II Pillar 3 CSLP-II Pillar 4 Governance Economy Social Services Sust. Development MG Water and Environment MG Justice and Rule of Law MG Community Recovery MG Regional Integration Partnership framework MG Energy and Mining MG Good Governance Telecommunications MG Health (CPSD) MG Private Sector MG Transport and MG Agriculture MG Education Development MG Security MG Gender MG Land Positioned above all challenges related to the lack of collaboration between institutions, this very innovative framework proved to be operational. It was used as an intermediary for the exchange of information between development partners and Government actors, the responsibility distribution among the PTF in mentoring sector groups, and the piloting of PCG bodies. Planning improved in many sectors thanks to such sector group efforts. In addition, the successful operation of the PCG enhanced synergies and consistency in PTF interventions thanks to their involvement in planning and budgeting actions across sector groups. Such successes contrast with the real difficulties found at the tool and procedure ownership level of State modernization reform. However, improvements have been observed following Government efforts to respond to the challenges of overlapping and skill-related conflicts among aid coordination piloting institutions. The recent fusion of finance and planning related ministries is a testimony to this fact. 3. CNCA/Donor Coordination In acting as PCG Secretariat, the SP/CNCA has been the privileged interface between Government and partners regarding aid efficiency international agenda implementation, namely monitoring the Paris declaration indicators, implementing international engagement principles in fragile countries (Accra Agenda for Action), international partnership for efficient development cooperation, and the New Deal (Busan). To support the partner coordination mechanism, aid-related statistics collection has proved necessary. The SP/CNCA has thus been fitted with an Aid Management platform to collect, enter and process external aid data. A data collection protocol was developed jointly and approved by 170 both partners and Government. Henceforth, the database will be updated quarterly. The SP/CNCA has acquired a new and improved performance tool which will allow intervention mapping and provide easier data user access. With the new software, partners can themselves enter data in the database. In addition, the SP/CNCA has already participated in three OECD surveys in 2006, 2008, and 2011. With aid policy and its action plan adopted in March 2012 by the PCG Political Forum, the SP/CNCA is looking to hold in-depth discussions within the PCG framework on arrangements to promote budget transparency, transaction cost efficiency through parallel unit reduction, program approach generalization, and aid defragmentation. D. CNCA/Ministry of Finance Coordination Acting as Secretariat of the Inter-ministerial Aid Coordination Committee including the Ministry of Finance and Economic Development Planning, the Ministry of Foreign Affairs and International Cooperation, and the Ministry of Interior, the SP/CNCA must interact with all of them. More specifically as regards the Ministry of Finance, the SP/CNCA, through the aid database, provides the Ministry with disbursement forecast data to feed into central MTEF and budget preparation. Furthermore, the provided data represents a valuable source not only for public expenditure review and State Financial Operation Matrix development, but also for joint annual sectorial reviews. The SP/CNCA also works in partnership with the Ministry of Finance and Economic Development Planning to develop strategic documents such as aid policy, Public Finance Management strategy, etc. The MoF chairs PCG Strategic Forum meetings with the SP/CNCA acting as Secretariat. Both institutions worked closely to prepare and organize the Burundi Development Partners Conference held in Geneva on October 29-30, 2012. Despite recorded progress, collaboration between the MoF and the CNCA should be strengthened in terms of budget preparation so as to lessen the gaps observed in data from both institutions. In the short term, CNCA representation in ad hoc commissions established by the MoF for budget preparation would be significant. E. CNCA/Ministries Coordination The SP/CNCA has privileged relations with technical ministries. Indeed, senior staff and specialists from SP/CNCA oversee all work by sector group. This has enabled almost all ministries to develop current MTEF sector policies, strategies and action plans. Joint annual reviews have been organized through sector group work and the SP/CNCA database has become a critical data source. The same specialists and senior staff have been mentoring and coordinating ministries in partner conference preparation at international and sector levels. Analysis of donor disbursements per ministry 4. Table G.1 displays donor disbursements other than budget support per ministry and highlights excessively marked variations between the various ministries. During the period, external funding reflected a state of emergency marked by prevailing humanitarian aid. Funding granted to the Ministry of National Solidarity reached the US$381 million over the period. This funding was essentially intended to benefit refugee reception and socio-economic reinsertion. 171 This ministry is followed by the Ministry of Public Works with a disbursement level of close to US$214 million. These disbursements were mainly used to finance major road and paving construction work executed over recent years. Free health care and performance-based funding policy implementation in the health sector has allowed it to mobilize significant external funding - close to US$199 million - in favor of the Ministry of Health and 87 Million for AIDS control. 5. Disbursements observed in the Ministry of Finance require careful interpretation given the cross-cutting nature of its projects with other ministries. Key projects remain public finance management strategy funding and European Development Fund-financed projects. The Ministries of Energy and Mines and of Primary Education register the same external funding volumes with close to US$100 million disbursement over the period. In the energy sector, disbursements made are relative to Government support to respond to energy deficit while in education sector, the greatest part comes from the Education Trust Fund. The Ministry of Agriculture and Stock Breeding was granted a significant amount with close to US$143 million essentially to face food shortages with the implementation of several projects. The Ministry of Interior also benefited from a substantial disbursement volume over the period mainly for General Population and Housing Census operations started in 2008. Table G.1 External disbursement evolution per ministry (2007-2011) Year 2007 2008 2009 2010 2011 Total Presidency of the Republic 0.5 0.0 0.0 0.0 0.0 0.5 1st Vice-Presidency 0.3 0.3 0.0 0.0 0.0 0.7 d 2 Vice-Presidency 0.2 0.7 0.3 1.1 2.6 4.9 National Assembly 0.0 0.0 0.0 0.1 0.0 0.2 Ministry of the Presidency in charge of Good Governance, and Privatization 6.9 5.9 8.3 10.5 10.2 41.8 Ministry of Decentralization, and Community Development 6.2 3.6 10.5 20.3 14.3 54.9 Ministry of Defense and Ex-Combatants 18.0 12.5 9.9 6.2 10.4 57.0 Ministry of Public Service, Labor, and Social Security 0.3 0.3 0.1 0.1 0.9 1.8 Ministry of Justice 3.2 4.9 2.1 3.5 4.3 18.0 Ministry of AIDS Control 13.0 18.5 22.2 15.2 18.9 87.7 Ministry of Public Health 7.6 20.6 50.4 69.8 50.9 199.3 Ministry of Public Security 8.3 5.6 5.0 4.7 5.5 29.1 Ministry of National Solidarity, Refugee Repatriation, and Social Reintegration 94.7 93.5 88.0 56.5 48.4 381.2 Ministry of Agriculture and Stock Breeding 19.0 21.6 38.1 26.2 38.7 143.7 Ministry of Energy and Mines 8.0 9.8 17.7 33.9 36.8 106.3 Ministry of Handicrafts, Vocational Training, and Adult Literacy 0.0 0.0 0.0 0.6 1.1 1.7 Ministry of Primary and Secondary Education 12.0 23.5 17.4 17.3 26.9 97.0 Ministry of Higher Education and Scientific Research 3.0 2.4 1.5 2.3 1.6 10.8 Ministry of Water, Environment, Land Management, and Urban Development 10.4 8.9 17.3 24.4 25.4 86.3 Ministry of Information, Communication, and Parliament Relations 0.4 0.2 2.3 2.2 0.9 5.9 Ministry of Interior 10.0 7.8 6.6 35.8 12.8 73.1 Ministry of East-African Community Affairs 0.0 0.4 0.0 5.2 10.1 15.7 Ministry of Human Rights and Gender 4.9 2.4 5.0 2.1 4.0 18.3 Ministry of Foreign Affairs and International Cooperation 0.0 0.4 0.9 1.7 1.6 4.6 Ministry of Transportation, Post, and Telecommunications 3.5 0.4 3.1 2.0 6.3 15.3 Ministry of Public Works and Equipment 25.0 47.3 38.4 75.9 28.1 214.6 Ministry of Trade, Industry, and Tourism 0.6 0.9 0.6 1.9 0.2 4.2 172 Ministry of Planning and Reconstruction 3.9 15.0 11.8 10.4 6.4 47.5 Ministry of Finance 19.5 14.9 17.8 31.9 33.6 117.7 Ministry of Youth, Sports, and Culture 4.6 1.0 1.1 0.7 0.7 8.1 Overall Total 283.9 323.5 376.5 462.5 401.4 1,847.8 Source: CNCA Data 6. To ensure transition from humanitarian aid to development aid, the coming years should be marked by a transition of aid donors from humanitarian to growth potential sectors. External funding is presented in the graph below, per priority sector: Figure G.1: External funding evolution per priority sector Source: CNCA Data 7. PRSP focus areas funding has followed a fluctuating pattern over the PRSP-I implementation period. An excessive volatility level is observed in the public works and equipment sector with an original funding level of US$25 million in 2007 reaching 76 Million in 2010. In 2011, this level almost dropped back to its original 2007 level with close to US$28 million. Over the period, the health sector benefited from the greatest part of the funds granted to the country. Other sectors have registered an increasing trend over the period, except for the defense and security sector which experienced a decreasing trend, going from US$26 million in 2007 to close to 16 Million in 2011. 173 Annex H : Pro-poor spending – supplementary analysis 1. This annex analyses the implementation of the pro-poor budget by evaluating the relevance, continuity, allocation, consistency and execution of Burundi’s pro-poor budget. For this analysis we use the budget and pro-poor budget information extracted from SIGEFI, cleaned according to the BOOST methodology.109 RELEVANCE WITHING THE BUDGET 2. ‘Relevance’ measures the importance of pro-poor spending compared to the entire budget using three measures. This is done with (i) a simple count of the number lines of classified as pro-poor and (ii) a calculation of the pro-poor budget as a share of the total credits. In addition to this, and given that the lines classified as pro-poor change over time, (iii) the percentage share of the pro-poor budget lines is estimated as a percentage of total budget lines. 3. During the period of 2008-2012, the pro-poor budget maintained its importance in budgetary and relative terms. Throughout the sample period, the number of pro-poor budget lines varied in each year; however, in terms of total budget lines, pro-poor budget lines represented more than 45% of the total. More importantly, pro-poor spending passed from a little more than a third of credits in 2008 to 45% in 2012 (Chart 1). Figure H.1: How relevant is the Pro-poor budget? 60% 1,420 50% 1,400 1,380 40% 1,360 30% 1,340 20% 1,320 10% 1,300 0% 1,280 2008 2009 2010 2011 2012 Number of pro-poor lines (RHS) As share of total number of lines As share of approved budget Source: Bank staff calculations based of SIGEFI data 109 The data retrieved from SIGEFI does not include the original budget estimates but the credits available after budget modification and reallocations by the Ministry of Finance. Moreover, as mentioned in the disclaimer at the beginning of the budget analysis, the data extracted from SIGEFI is not entirely reliable. The analysis presented in this section should, therefore, be taken with reserves and care. 174 Continuity 4. While the proportion of the budget assigned to the pro-poor programs is important, another aspect in assessing the relevance and efficacy of pro-poor budget is by looking at its continuity over time. Though the current method to classify pro-poor spending is not really prone to impact evaluation, it is necessary to look, at the very least, at the stability of these expenditures. To evaluate the continuity of the pro-poor budget, we count the number of pro- poor budget lines reported in 2008 in SIGEFI110 that were available in the following years. The results show little continuity. Only half of the lines classified as pro-poor budget in 2008 were repeated in 2010, and in 2011 this figure fell to one third of the total (Chart 2). Similarly, when we compare the pro-poor lines reported in 2009 with the following years, we find that, 50% of the lines were not included in the following year (2010), and this figure further decreased in the second and third year (2011 and 2012) to less than 40%. This indicates that there is a systematic disruption over time in the pro-poor budget lines. Figure H.2: How continuous is the pro-poor budget? 1600 Total number of pro-poor lines 1400 1200 1000 800 600 400 200 0 2009 2010 2011 2012 Repeated lines Non-repeated lines Source: Bank staff own calculation with data from SIGEFI Allocation (by ministry and by economic classification) 5. In addition to relevance and continuity, we analyze the allocation of the resources at the ministry level (including the allocation to priority sectors) and by economic classification. At the ministry level, the analysis shows that the pro-poor budget is concentrated in the priority sectors (Health, Education, Agriculture, Energy and Mining, Public Works), as well as the social services. This indicates that, at least at the allocation level, the pro-poor budget is consistent with the poverty reduction strategy (PRSP-II). 6. The Ministry of Agriculture and Livestock concentrated most of the pro-poor budget lines in every year, accounting for 20% of the total number of lines classified as pro-poor. Yet, in monetary terms, the Ministry of Agriculture only managed 5.5% on average of the pro-poor 110 Though the pro-poor budget was introduced before 2008, counterparts in the Ministry of Finance were only able to provide data from that year onwards. In any case, the results from our sample period show little continuity in the classification of pro-poor spending. 175 budget. As the section on budget analysis showed, the Ministry of Agriculture is consistently among the units with the lowest execution rates according to SIGEFI. The Ministry of Basic and Secondary Education111 managed the highest amount of the pro-poor funds (average of 46%). This is in line with the authorities’ policy goal of expanding access to basic education. Moreover, the Ministry of Superior Education and Research was the second highest recipient of the pro- poor budget, receiving on average 12.1% of the funds. Adding these two ministries, 60% of the pro-poor budget was allocated in Education. The third largest manager of pro-poor spending is the Ministry of Health and Fight Against AIDS112, which received, on average, 12.5% of the pro- poor budget. During these five years, we observe that more resources are constantly allocated to this ministry as its share increased to 15.9% in 2011 and 13.5% in 2012 from 9.4% in 2008. In total, the education ministries and the Ministry of Health manage 71% of the pro-poor budget. Table H.1 Pro-poor budget distribution by administrative classification of spending 2008 2009 2010 2011 2012 Average General Services 9.2% 7.4% 8.7% 15.8% 10.5% 10.3% Social Services 72.3% 76.2% 79.3% 69.5% 72.1% 73.9% Economic Services 18.5% 16.5% 11.9% 14.6% 17.4% 15.8% Source: Bank staff own calculation with data from SIGEFI Table H.2 Pro-poor budget allocation in priority sectors Priority sectors 2008 2009 2010 2011 2012 Average Health/AIDS 9.4% 10.5% 12.9% 15.9% 13.5% 12.5% Education 59.7% 62.7% 63.2% 51.7% 56.8% 58.8% Agriculture 7.3% 5.3% 3.0% 6.3% 5.9% 5.6% Energy and 1.2% 3.6% 1.5% 2.6% 5.6% 2.9% Mining Public Works 8.2% 5.7% 5.4% 3.9% 4.1% 5.5% Grand Total 85.8% 87.7% 86.0% 80.5% 85.9% 85.2% Source: Bank staff own calculation with data from SIGEFI Economic classification expenditure 7. While the pro-poor budget allocation at the ministry level is compliant with the pro-poor budget guidelines and is concentrated in the priority sectors, a breakdown by economic classification suggests that, on average, more than 80% of the pro-poor budget was spent on current expenses (Figure H.3). Within the current expenditure classification we note that Personal expenses took, on average, 50% of the pro-poor budget (Figure H.4).This is driven by the fact that the education-related ministries, which devote most of their resources to wage bill, are the largest managers of pro-poor spending. We also note that this expenditure posted an upward trend, increasing to 52.1% in 2012 from 44.6% in 2008. Moreover, on average, 12% of the budget was spent on Employer social security benefits, social assistance and the other expenses account. The increasingly large share of the wage bill as a proportion of the pro-poor 111 The name of the Ministry has changed over time 112 Before 2011 this ministry was divided in two ministries. We added the two ministries to make this comparison. 176 budget is in similar to allocations within the rest of the budget. As other sections of the PER have highlighted, the share of wage bill in the budget has increased considerably in the last years. Interestingly, tax exemptions (Allocations - contributions et exonerations), represent, on average, 14% of the pro-poor spending. As mentioned before, capital expenditure has 20% of the pro-poor expenditure. Within this classification, the largest expenditure is on fixed assets with an average of 12% of total pro-poor budget. Figure H.3: Where does the Pro-poor budget go? 400,000,000,000 350,000,000,000 19.6% 300,000,000,000 20.3% 250,000,000,000 10.0% 200,000,000,000 19.9% 150,000,000,000 80.4% 14.2% 79.7% 100,000,000,000 90.0% 80.1% 50,000,000,000 85.7% (FBu) 0 2008 2009 2010 2011 2012 Current expenditure Capital Arreas or unpaid (World Bank) Source: Bank staff own calculation with data from SIGEFI Figure H.4: Current expenditure breakdown of Pro-poor budget Securities and 400,000,000,000 Investments 350,000,000,000 Stocks 300,000,000,000 Resources transferred to proyects 250,000,000,000 Intangible assets (FBu) 200,000,000,000 Assets in progress 150,000,000,000 Fixed assets 100,000,000,000 Grants 50,000,000,000 0 External services 2008 2009 2010 2011 2012 Source: Bank staff own calculation with data from SIGEFI 177 Execution 8. Finally we examine the execution rate of the pro-poor budget at the aggregate level but also by economic classification and priority sectors. The execution rates are estimated by dividing commitments (ordonnoncement) by credit. This ratio is consistent with indicator PI-2 of the PEFA and is appropriate considering that treasury management is still weak in Burundi. 9. The charts analysis illustrates that the pro-poor budget has a high execution rate. However, this must be qualified: the execution rate is not calculated with respect to the original budget, but to credits after modification (in the years when there are revisions to the budget) and reallocations by the Ministry of Finance. The reason for this is that public officials from the Ministry of Finance were not able to extract original budget data at the line level from SIGEFI. In any case, on average, during 2008-2012 the execution rate of the pro-poor budget113 was 89%. However, it is worth noticing that in 2011 the execution rate lowered notably to 64% from 100% in 2010 and then rise up again to 92% in 2012. It is unclear whether this volatility is derived from misrecordings in SIGEFI, bad cash management and planning, or both. 10. In line with the rest of the budget, capital expenditures present an execution rate of 69% for the period between 2008 and 2012 (Table 3). During this period, the execution rates of the pro-poor capital expenditures declined to 59% in 2010 from 89% in 2008, and then recovered in the last two years, reaching 80% in 2012. Yet, in comparison to the other capital expenditure execution rates (Ordinary and HIPC), the pro-poor budget execution rate (paid budget/initial credit) is the lowest one114 Figure H.5: Execution rate of the Pro-poor budget 120% (Committed / Initial credit) 100% 80% 60% 40% 20% 0% 2008 2009 2010 2011 2012 Current expenditure Capital expenditure Total expenditure Source: Bank staff own calculation with data from SIGEFI 113 Calculated as (ordonnancement/credit) 114 credit) is the lowest one. Among the priority sectors, we observe that Health/AIDS, Primary Education and Higher Education have the highest executions rates, with an average of 92%, 93% and 90%, respectively. The execution rate for the Ministry of Agriculture was 74% in average, while the Ministry of Energy and Mining registered the lowest execution rate with an average of 56%. 178 Figure H.6: Execution rate of the Pro-poor budget in priority sectors 120% (Committed/ Initial credit) 100% 80% 60% 40% 20% 0% 2008 2009 2010 2011 2012 Health/AIDS Education Agriculture Energy and Mining Public Works Source: Bank staff own calculation with data from SIGEFI 179 Annex I : Burundi Financial and Economic Statistics Table I.1.a Nominal GDP in billion FBu, 2005-2012 PIB Nominal aux prix courants 2005 2006 2007 2008 2009 2010 2011 2012 Primary sector 493.6 532.3 504.0 711.6 791.8 922.2 1,096.2 1,321.4 Agriculture (subsistence farming) 429.9 427.5 426.1 590.1 669.9 777.7 919.7 1,098.2 Agriculture (export) 17.0 51.7 23.8 59.7 21.8 28.8 40.3 61.1 Forestry 9.3 10.6 11.3 10.5 11.8 13.9 16.7 20.3 Livestock 34.1 38.1 40.0 41.1 78.5 90.7 106.8 127.1 Fishing 3.3 4.4 2.8 10.2 9.8 11.1 12.7 14.8 Secondary sector 204.7 200.6 243.5 280.2 324.9 380.9 459.1 550.1 Extraction 10.4 10.4 10.1 11.9 12.3 12.3 12.8 13.4 Industries 143.6 135.8 163.0 183.9 208.0 231.2 262.8 297.2 Electricity, gas and water 10.5 8.5 12.4 16.6 15.9 20.7 27.4 36.6 Construction 40.2 45.9 58.0 67.8 88.7 116.7 156.3 202.8 Tertiary sector 411.0 467.5 602.4 761.3 836.8 976.9 1,161.6 1,385.6 Trade 83.4 94.6 100.9 153.5 88.7 103.6 123.1 148.3 Transports and communications 35.5 40.8 70.4 69.5 98.7 115.3 137.1 165.1 Banks and Insurances 22.3 26.5 49.5 58.9 62.6 73.1 87.0 104.7 Accommodation, restaurants, and other commercial services 190.6 203.6 206.0 279.5 340.8 397.8 473.0 565.8 Public Administration 59.8 63.0 103.1 127.7 149.9 175.0 208.0 241.1 Education 32.4 44.7 55.0 78.3 85.1 99.3 118.1 142.2 Health 8.1 10.3 17.7 17.0 21.0 24.6 29.2 35.2 Personal or collective activities 7.5 10.0 36.6 28.0 44.4 51.8 61.6 74.2 Home services 9.3 9.5 9.7 9.7 10.1 11.8 14.0 16.9 SIFIM -37.9 -35.5 -46.5 -60.8 -64.5 -75.3 -89.6 -107.9 GDP at factor cost 1,109.3 1,200.4 1,349.9 1,753.1 1,953.5 2,280.0 2,717.0 3,257.1 Taxes 99.1 109.3 117.2 157.8 186.7 214.6 253.7 308.7 Nominal GDP 1,208.4 1,309.7 1,467.1 1,910.9 2,140.2 2,494.6 2,970.7 3,565.8 Nominal GDP growth 2005 2006 2007 2008 2009 2010 2011 2012 Primary sector 16.9% 7.8% -5.3% 41.2% 11.3% 16.5% 18.9% 20.5% Agriculture (subsistence farming) 15.5% -0.6% -0.3% 38.5% 13.5% 16.1% 18.3% 19.4% Agriculture (export) 57.4% 204.1% -54.0% 150.8% -63.5% 31.8% 40.0% 51.8% Forestry 20.6% 14.0% 6.6% -7.1% 12.2% 18.0% 20.2% 21.3% Livestock 18.9% 11.7% 5.0% 2.8% 90.9% 15.6% 17.8% 18.9% Fishing 19.5% 33.3% -36.4% 264.3% -3.4% 12.7% 14.8% 15.9% Secondary sector 24.5% -2.0% 21.4% 15.1% 16.0% 17.2% 20.6% 19.8% Extraction 24.5% 0.0% -2.9% 17.8% 3.6% -0.3% 3.8% 4.8% Industries 24.5% -5.4% 20.0% 12.8% 13.1% 11.2% 13.6% 13.1% Electricity, gas and water 24.5% -19.0% 45.9% 33.9% -4.4% 30.1% 32.6% 33.9% Construction 24.5% 14.2% 26.4% 16.9% 30.8% 31.5% 33.9% 29.8% Tertiary sector 21.6% 13.7% 28.9% 26.4% 9.9% 16.7% 18.9% 19.3% Trade 21.6% 13.4% 6.7% 52.1% -42.2% 16.7% 18.9% 20.4% Transports and communications 21.6% 14.9% 72.5% -1.3% 42.1% 16.7% 18.9% 20.4% Banks and Insurances 21.6% 18.8% 86.8% 19.0% 6.4% 16.7% 18.9% 20.4% 180 Accommodation, restaurants, and other commercial services 21.6% 6.8% 1.2% 35.7% 21.9% 16.7% 18.9% 19.6% Public Administration 21.6% 5.4% 63.7% 23.9% 17.4% 16.7% 18.9% 15.9% Education 21.6% 38.0% 23.0% 42.4% 8.7% 16.7% 18.9% 20.4% Health 21.6% 27.2% 71.8% -4.0% 23.7% 16.7% 18.9% 20.4% Personal or collective activities 21.6% 33.3% 266.0% -23.5% 58.6% 16.7% 18.9% 20.4% Home services 21.6% 2.2% 2.1% 0.0% 4.2% 16.7% 18.9% 20.4% SIFIM 21.6% -6.3% 31.0% 30.8% 6.2% 16.7% 18.9% 20.4% GDP at factor cost 20.0% 8.2% 12.5% 29.9% 11.4% 16.7% 19.2% 19.9% Taxes 19.5% 10.3% 7.2% 34.6% 18.3% 14.9% 18.2% 21.7% Nominal GDP growth 19.9% 8.4% 12.0% 30.3% 12.0% 16.6% 19.1% 20.0% Weighting sector GDP Primary sector 41% 41% 34% 37% 37.0% 37.0% 36.9% 37.1% Agriculture (subsistence farming) 36% 33% 29% 31% 31.3% 31.2% 31.0% 30.8% Agriculture (export) 1% 4% 2% 3% 1.0% 1.2% 1.4% 1.7% Forestry 1% 1% 1% 1% 0.6% 0.6% 0.6% 0.6% Livestock 3% 3% 3% 2% 3.7% 3.6% 3.6% 3.6% Fishing 0% 0% 0% 1% 0.5% 0.4% 0.4% 0.4% Secondary sector 17% 15% 17% 15% 15.2% 15.3% 15.5% 15.4% Extraction 1% 1% 1% 1% 0.6% 0.5% 0.4% 0.4% Industries 12% 10% 11% 10% 9.7% 9.3% 8.8% 8.3% Electricity, gas and water 1% 1% 1% 1% 0.7% 0.8% 0.9% 1.0% Construction 3% 4% 4% 4% 4.1% 4.7% 5.3% 5.7% Tertiary sector 32% 33% 36% 36% 39.1% 39.2% 39.1% 38.9% Trade 7% 7% 7% 8% 4.1% 4.2% 4.1% 4.2% Transports and communications 3% 3% 5% 4% 4.6% 4.6% 4.6% 4.6% Banks and Insurances 2% 2% 3% 3% 2.9% 2.9% 2.9% 2.9% Accommodation, restaurants, and other commercial services 16% 16% 14% 15% 15.9% 15.9% 15.9% 15.9% Public Administration 5% 5% 7% 7% 7.0% 7.0% 7.0% 6.8% Education 3% 3% 4% 4% 4.0% 4.0% 4.0% 4.0% Health 1% 1% 1% 1% 1.0% 1.0% 1.0% 1.0% Personal or collective activities 1% 1% 3% 2% 2.1% 2.1% 2.1% 2.1% Home services 1% 1% 1% 1% 0.5% 0.5% 0.5% 0.5% SIFIM -3% -3% -3% -3% -3.0% -3.0% -3.0% -3.0% GDP at factor cost 91.8% 91.7% 92.0% 91.7% 91.3% 91.4% 91.5% 91.3% Taxes 8% 8% 8% 8% 8.7% 8.6% 8.5% 8.7% GDP at market price 100.0 100.0 100.0 100.0 100% 100% 100% 100% Contribution to growth Primary sector 6.9% 3.2% -1.8% 15.3% 37.2% 6.1% 7.0% 7.7% Agriculture (subsistence farming) 5.5% -0.2% -0.1% 11.9% 30.90% 5.02% 5.65% 5.98% Agriculture (export) 0.8% 8.0% -0.9% 4.7% 3.10% 0.37% 0.54% 0.89% Forestry 0.2% 0.1% 0.1% 0.0% 0.60% 0.10% 0.11% 0.12% Livestock 0.5% 0.3% 0.1% 0.1% 2.10% 0.57% 0.64% 0.67% Fishing 0.1% 0.1% -0.1% 1.3% 0.50% 0.06% 0.06% 0.07% Secondary sector 4.2% -0.3% 3.6% 2.2% 2.5% 2.8% 3.3% 3.2% Extraction 0.2% 0.0% 0.0% 0.1% 0.02% 0.00% 0.02% 0.02% Industries 2.9% -0.6% 2.2% 1.2% 1.27% 1.04% 1.21% 1.09% Electricity, gas and water 0.2% -0.1% 0.4% 0.3% -0.03% 0.25% 0.30% 0.35% Construction 0.8% 0.5% 1.1% 0.6% 1.28% 1.47% 1.79% 1.70% 181 Tertiary sector 7.0% 4.5% 10.4% 9.5% 5.1% 5.8% 6.5% 6.6% Trade 1.5% 1.0% 0.5% 4.2% -1.75% 0.69% 0.78% 0.85% Transports and communications 0.6% 0.5% 3.5% 0.0% 1.94% 0.77% 0.87% 0.95% Banks and Insurances 0.4% 0.4% 3.0% 0.6% 0.19% 0.49% 0.55% 0.60% Accommodation, restaurants, and other commercial services 3.4% 1.1% 0.2% 5.2% 3.49% 2.67% 3.01% 3.11% Public Administration 1.1% 0.3% 4.5% 1.6% 1.22% 1.17% 1.32% 1.08% Education 0.6% 1.3% 0.9% 1.7% 0.34% 0.67% 0.75% 0.81% Health 0.2% 0.2% 0.9% 0.0% 0.23% 0.16% 0.19% 0.20% Personal or collective activities 0.1% 0.3% 6.7% -0.4% 1.21% 0.35% 0.39% 0.43% Home services 0.2% 0.0% 0.0% 0.0% 0.02% 0.08% 0.09% 0.10% SIFIM -0.7% 0.2% -1.0% -1.0% -0.19% -0.51% -0.57% -0.62% GDP at factor cost 18.33% 7.53% 11.46% 27.40% 10.44% 15.27% 17.53% 18.16% Taxes 1.6% 0.9% 0.6% 2.9% 1.60% 1.28% 1.56% 1.88% Source: IMF Tableau I.1.b Real GDP in billion FBu, 2005-2012 Real GDP (2005 prices) 2005 2006 2007 2008 2009 2010 2011 2012 New GDP at 2005 prices 1,208.4 1,273.5 1,334.4 1,401.8 1,450.4 1,505.3 1,568.4 1,631.2 Primary sector 493.6 509.1 464.3 454.5 468.2 486.5 507.9 535.2 Agriculture (subsistence farming) 429.9 418.7 398.6 393.8 405.8 418.1 430.7 443.8 Agriculture (export) 17.0 41.0 19.3 29.2 30.1 35.2 43.0 56.3 Forestry 9.3 8.9 9.2 6.8 7.1 7.4 7.8 8.1 Livestock 34.1 35.4 34.5 21.3 21.9 22.5 23.0 23.6 Fishing 3.3 5.1 2.6 3.4 3.4 3.4 3.4 3.4 Secondary sector 204.7 215.7 233.3 230.0 241.3 253.2 268.1 279.1 Extraction 10.4 10.2 9.6 6.9 6.1 5.4 4.8 4.4 Industries 143.6 152.8 157.4 146.4 147.5 145.6 144.1 140.7 Electricity, gas and water 10.5 8.6 12.4 12.7 13.9 16.0 18.5 21.4 Construction 40.2 44.1 53.9 64.1 73.9 86.2 100.6 112.7 Tertiary sector 411.0 441.6 530.4 601.1 622.1 644.5 667.7 685.9 Trade 83.4 92.7 97.7 110.6 114.4 118.5 122.8 127.6 Transports and communications 35.5 38.8 59.7 44.5 46.0 47.7 49.4 51.3 Banks and Insurances 22.3 24.9 47.5 60.6 62.7 65.0 67.3 70.0 Accommodation, restaurants, and other commercial services 190.6 179.9 151.7 175.9 182.1 188.7 195.5 201.7 Public Administration 59.8 69.8 111.6 149.5 154.8 160.3 166.1 166.1 Education 32.4 44.5 53.5 67.6 70.0 72.5 75.1 78.1 Health 8.1 5.7 4.5 4.3 4.5 4.6 4.8 5.0 Personal or collective activities 7.5 11.6 42.7 41.7 43.2 44.8 46.4 48.2 Home services 9.3 9.3 8.5 7.7 8.0 8.2 8.5 8.9 SIFIM -37.9 -35.4 -47.0 -61.4 -63.6 -65.9 -68.2 -70.9 GDP at factor cost 1,109.3 1,166.4 1,227.9 1,285.6 1,331.7 1,384.2 1,443.7 1,500.2 Taxes 99.1 107.0 106.5 116.2 118.7 121.1 124.7 131.0 Real GDP (2005 prices) 1,208.4 1,273.5 1,334.4 1,401.8 1,450.4 1,505.3 1,568.4 1,631.2 Real GDP (2005 prices) 2005 2006 2007 2008 2009 2010 2011 2012 182 Real GDP sectoral growth rate Primary sector 1.7% 3.1% -8.8% 2.1% 3.0% 3.9% 4.4% 5.4% Secondary sector 8.3% 5.4% 8.2% -1.4% 4.9% 4.9% 5.9% 4.1% Tertiary sector 5.8% 7.4% 20.1% 13.3% 3.5% 3.6% 3.6% 2.7% GDP at factor cost 4.4% 5.2 3.8 4.5 3.6% 3.9% 4.3% 3.9% Taxes 4.0 8.0 -0.5 9.1 2.2% 2.0% 3.0% 5.0% Real GDP growth 4.4% 5.4% 4.8% 5.0% 3.5% 3.8% 4.2% 4.0% Weighting sector GDP Primary sector 40.8 40.6 34.3 37.2 32.3% 32.3% 32.4% 32.8% Agriculture (subsistence farming) 35.6 32.9 30.0 27.4 28.0% 27.8% 27.5% 27.2% Agriculture (export) 1.4 3.2 1.8 2.3 2.1% 2.3% 2.7% 3.5% Forestry 0.8 0.7 0.8 0.5 0.5% 0.5% 0.5% 0.5% Livestock 2.8 2.8 2.7 1.6 1.5% 1.5% 1.5% 1.4% Fishing 0.3 0.4 0.2 0.2 0.2% 0.2% 0.2% 0.2% Secondary sector 16.9 15.3 16.6 14.7 16.6% 16.8% 17.1% 17.1% Extraction 0.9 0.8 0.7 0.5 0.4% 0.4% 0.3% 0.3% Industries 11.9 12.0 10.3 9.8 10.2% 9.7% 9.2% 8.6% Electricity, gas and water 0.9 0.7 0.9 0.8 1.0% 1.1% 1.2% 1.3% Construction 3.3 3.5 4.1 4.5 5.1% 5.7% 6.4% 6.9% Tertiary sector 34.0 35.7 41.1 39.8 42.9% 42.8% 42.6% 42.1% Trade 6.9 7.3 7.4 7.4 7.9% 7.9% 7.8% 7.8% Transports and communications 2.9 3.0 4.6 3.4 3.2% 3.2% 3.1% 3.1% Banks and Insurances 1.8 1.9 3.7 4.1 4.3% 4.3% 4.3% 4.3% Accommodation, restaurants, and other commercial services 15.8 14.1 12.7 15.5 12.6% 12.5% 12.5% 12.4% Public Administration 5.0 5.5 7.4 9.0 10.7% 10.7% 10.6% 10.2% Education 2.7 3.5 4.0 4.5 4.8% 4.8% 4.8% 4.8% Health 0.7 0.4 0.6 1.1 0.3% 0.3% 0.3% 0.3% Personal or collective activities 0.6 0.9 2.7 2.3 3.0% 3.0% 3.0% 3.0% Home services 0.8 0.7 0.6 0.6 0.5% 0.5% 0.5% 0.5% SIFIM -3.1 -2.8 -3.5 -4.0 -4.4% -4.4% -4.3% -4.3% GDP at factor cost 91.8 91.7 92.0 91.7 91.8% 92.0% 92.0% 92.0% Taxes 8.2 8.4 8.0 8.3 8.2% 8.0% 8.0% 8.0% GDP at market price 100.0 100.0 100.0 100.0 100% 100% 100% 100% Contribution to growth Primary sector 0.0% 0.0% 0.0% 1.0% 1.3% 1.5% 2.0% Agriculture (subsistence farming) -0.9% -1.4% -0.3% 0.85% 0.84% 0.83% 0.82% Agriculture (export) 4.5% -1.0% 1.2% 0.06% 0.40% 0.60% 1.07% Forestry 0.0% 0.0% -0.1% 0.02% 0.02% 0.02% 0.02% Livestock 0.1% -0.1% -0.6% 0.04% 0.04% 0.04% 0.04% Fishing 0.2% -0.1% 0.1% 0.00% 0.00% 0.00% 0.00% Secondary sector 0.0% 0.0% 0.0% 0.9% 1.0% 1.1% 0.8% Extraction 0.0% 0.0% -0.1% -0.05% -0.04% -0.03% -0.03% Industries 0.8% 0.3% -0.7% 0.08% -0.13% -0.09% -0.21% Electricity, gas and water -0.1% 0.4% 0.0% 0.09% 0.16% 0.18% 0.20% Construction 0.3% 0.9% 0.9% 0.78% 0.96% 1.07% 0.83% 183 Tertiary sector 0.0% 0.0% 0.0% 1.4% 1.4% 1.4% 1.0% Trade 0.8% 0.4% 1.0% 0.28% 0.28% 0.28% 0.31% Transports and communications 0.3% 2.5% -0.9% 0.11% 0.11% 0.11% 0.12% Banks and Insurances 0.2% 3.4% 1.1% 0.15% 0.16% 0.15% 0.17% Accommodation, restaurants, and other commercial services -0.8% -2.0% 2.5% 0.44% 0.45% 0.45% 0.40% Public Administration 0.9% 4.4% 3.1% 0.37% 0.38% 0.38% 0.00% Education 1.3% 0.8% 1.2% 0.17% 0.17% 0.17% 0.19% Health -0.1% -0.1% 0.0% 0.01% 0.01% 0.01% 0.01% Personal or collective activities 0.5% 7.3% -0.1% 0.10% 0.11% 0.11% 0.12% Home services 0.0% -0.1% -0.1% 0.02% 0.02% 0.02% 0.02% SIFIM 0.2% -1.1% -1.2% -0.15% -0.16% -0.16% -0.17% GDP at factor cost 4.8% 3.5% 4.1% 3.29% 3.63% 3.95% 3.60% Taxes 0.7% 0.0% 0.8% 0.18% 0.16% 0.24% 0.40% Real GDP Growth 0.1% 0.0% 0.1% 3.5% 3.8% 4.2% 4.0% Source: IMF Table I.2 Monetary Survey, 2005-2012 2005 2006 2007 2008 2009 2010 2011 2012 Monetary Situation Net foreign assets 50.7 53.9 84.9 155.0 168.8 139.9 63.3 59.3 Central Bank 29.2 20.8 51.4 113.0 118.7 104.3 26.2 14.1 Commercial Banks 21.5 33.1 33.5 42.0 50.0 35.7 37.1 45.1 Net domestic assets 247.1 300.0 322.9 395.4 500.3 644.9 771.1 971.2 Domestic credit 277.9 355.9 374.1 436.2 559.2 692.4 887.2 1079.7 Net credit to the government 121.1 159.7 157.4 176.5 252.4 291.9 342.4 378.4 Net credit to the government-TOFE 195.0 268.8 303.1 355.9 390.6 Central government (net) 134.4 173.7 168.9 188.2 267.3 302.7 355.1 354.3 Treasury 170.2 216.3 218.1 234.3 323.8 373.4 416.2 415.3 Other central government -35.9 -42.6 -49.2 -46.1 -56.5 -70.8 -61.2 -61.0 Other government agencies (net) -13.2 -13.4 -10.9 -11.5 -14.4 -10.9 -13.0 0.0 Local government (net) -0.1 -0.6 -0.6 -0.2 -0.5 0.1 0.3 0.0 Public enterprises 5.2 9.2 6.9 21.9 8.4 12.1 3.8 3.8 Private sector 151.6 187.0 209.7 237.8 298.4 388.4 541.0 697.5 Other items net (asset: +) 3/ -30.8 -55.9 -51.2 -40.8 -58.9 -47.5 -116.1 -83.2 M3 297.8 353.9 407.8 550.4 669.1 784.8 834.4 1030.5 foreign deposits 40.2 52.3 77.7 107.3 138.3 151.4 162.2 234.5 Broad money 5/ 257.7 301.6 330.1 443.1 530.8 633.5 672.1 796.0 Currency in circulation 67.9 68.4 84.2 112.6 120.9 139.1 153.2 189.2 Deposits 189.8 233.1 245.9 330.5 409.9 494.4 518.9 606.8 demand deposits 115.6 149.1 151.2 223.3 283.5 341.3 331.8 393.4 time deposits 74.2 84.0 94.7 107.2 126.4 153.1 187.1 213.4 Monetary Authorities Net foreign assets 29.2 20.8 51.4 113.0 118.7 104.3 26.2 14.1 in millions of US$ 29.3 20.8 45.9 91.5 96.5 84.6 19.2 10.1 Foreign assets 114.3 132.7 201.1 332.3 398.8 416.1 411.5 414.1 184 in millions of US$ 114.0 133.0 225.1 410.4 490.7 512.8 560.3 581.4 of which official reserves (BIF) 112.4 131.4 198.5 329.1 397.5 409.5 402.4 405.0 Foreign liabilities 85.0 111.9 149.7 219.3 280.0 311.8 385.4 400.0 in millions of US$ 85.2 111.6 133.7 177.6 227.6 253.0 283.1 284.9 of which: MLT liabilities Net domestic assets 71.3 87.3 73.6 43.4 79.3 105.6 185.1 234.8 Domestic credit 105.6 156.4 138.8 114.2 187.5 182.9 299.3 341.5 Government (net) 113.7 162.9 134.7 120.8 191.7 183.9 266.5 265.6 Government (net)-TOFE 134.1 199.0 189.9 273.4 272.4 Cent. Gov. 118.8 166.3 138.9 127.2 197.5 189.4 272.5 271.7 Treasury 153.1 206.5 180.2 168.0 248.0 250.6 320.5 0.0 Claims 154.4 208.5 181.7 170.8 250.3 253.2 322.0 0.0 Deposits 1.3 2.1 1.6 2.8 2.3 2.6 1.5 0.0 Other central government -34.3 -40.2 -41.2 -40.8 -50.5 -61.1 -48.0 0.0 Other government agencies -4.9 -2.8 -3.7 -6.4 -5.8 -5.5 -6.0 0.0 Local Gov. -0.2 -0.6 -0.5 -0.1 0.0 0.0 0.0 0.0 Nongovernment credit -8.2 -6.6 4.1 -6.6 -4.2 -1.0 32.8 75.9 Private sector 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Public enterprises 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Commercial banks -- 0.0 1.0 0.0 0.0 0.0 25.3 25.3 Discount window Money market Non-Bank Fin. Inst. 3.8 3.9 3.1 5.3 5.8 6.0 7.5 7.5 Liquidity Auction -12.0 -10.5 0.0 -12.0 -10.0 -7.0 0.0 43.1 Other items net (asset:+) -34.3 -69.0 -65.2 -70.7 -108.2 -77.4 -114.2 -81.3 Change/Revaluation (+) -7.2 0.4 13.1 15.4 -0.8 0.5 32.6 12.0 Change/Revaluation (-) 9.6 -0.5 -15.6 -26.0 1.8 -1.0 -66.2 -23.8 Revaluation (cumulative 1996=0) 2.5 2.4 -0.1 -10.7 -9.6 -10.1 -43.7 -55.4 Other items -36.8 -71.4 -65.1 -60.1 -98.6 -67.2 -70.6 -25.9 Reserve money 100.5 108.1 125.0 156.4 198.0 209.8 211.2 249.0 Currency in circulation 67.9 68.4 84.2 112.6 120.9 139.1 153.2 189.2 Commercial bank reserves (including cash in vault, from 2004 onwards) 25.6 28.2 24.9 25.0 53.9 45.7 34.9 53.6 Cash in vault 5.3 8.4 8.8 11.6 15.3 16.7 16.9 Other nonbank deposits 1.7 3.1 7.1 7.2 7.9 8.3 6.2 6.2 Commercial banks Net foreign assets 21.5 33.1 33.5 42.0 50.0 35.7 37.1 45.1 Foreign assets 59.0 74.5 98.3 125.8 155.8 162.9 174.9 182.9 Foreign liabilities 37.5 41.4 64.8 83.7 105.7 127.3 137.8 137.8 Reserves 30.9 36.6 33.8 36.6 69.2 62.4 51.8 53.6 BRB deposits 25.6 28.2 24.9 25.0 53.9 45.7 34.9 36.7 Required reserves 9.5 20.0 9.7 13.1 16.4 19.4 20.4 25.2 Of which: foreign exchange 1.9 Excess reserves 16.1 8.2 15.2 11.8 37.4 26.3 14.5 11.4 BRB borrowing from money market Cash in vault 5.3 8.4 8.8 11.6 15.3 16.7 16.9 16.9 Credit from BRB(rediscount; - liability) 0.0 0.0 0.0 0.0 0.0 0.0 -25.3 -25.3 Domestic credit 172.4 199.5 235.3 322.0 371.7 509.5 613.2 763.5 185 Net claims on the Government 7.4 -3.2 22.7 55.7 60.7 108.0 75.9 112.9 Net claims on the Government-TOFE 28.0 60.9 69.8 113.2 82.6 118.1 Central government (net) 15.6 7.4 30.0 60.9 69.8 113.2 82.6 82.6 Treasury 17.1 9.8 38.0 66.3 75.8 122.9 95.7 0.0 Credit 17.2 9.9 38.0 66.4 75.9 123.0 96.2 0.0 Deposits 0.0 0.1 0.1 0.1 0.1 0.1 0.4 0.0 Other central government -1.5 -2.4 -8.0 -5.3 -6.0 -9.6 -13.2 0.0 Credit 0.2 0.7 1.6 1.7 0.8 0.3 0.7 0.0 Deposits 1.7 3.1 9.6 7.0 6.8 9.9 13.9 0.0 Other government agencies -8.4 -10.6 -7.2 -5.1 -8.6 -5.4 -7.0 0.0 Credit 0.1 0.2 0.0 0.0 0.0 0.0 0.0 0.0 Deposits 8.5 10.8 7.2 5.1 8.6 5.4 7.0 0.0 Local government (net) 0.1 0.0 -0.1 -0.2 -0.5 0.2 0.3 0.0 Public enterprises 5.2 9.1 6.9 21.9 8.4 12.1 3.8 3.8 Private sector 159.8 193.6 205.6 244.5 302.6 389.4 533.6 646.9 Other items net (assets:+) 3.5 13.1 14.0 30.0 49.3 29.9 -1.9 -1.9 Revaluation (+) -3.7 0.2 4.8 6.7 -0.3 0.2 13.3 4.3 Revaluation (-) 5.5 -0.3 -8.7 -10.1 0.5 -0.3 -17.1 -5.4 Deposits 188.1 230.0 238.8 323.3 401.9 486.1 512.7 835.0 Demand deposits 113.9 146.1 144.1 216.1 275.5 333.0 325.6 387.2 Time and savings deposits 74.2 84.0 94.7 107.2 126.4 153.1 187.1 213.4 Foreign Currency deposits 40.2 52.3 77.7 107.3 138.3 151.4 162.2 234.5 PROGRAMING nominal GDP 1208.4 1309.7 1467.1 1910.9 2140.2 2494.6 2970.7 3565.8 Nominal GDP growth 19.9 8.4 12.0 30.3 12.0 16.6 19.1 20.0 Memorandum items (for projection validation) Required reserve ratio 5.0 7.0 3.0 3.0 3.0 3.0 3.0 3.0 effective reserve ratio 13.6 12.2 10.4 7.7 13.4 9.4 6.8 6.1 Growth of private sector credit -8.3 23.3 12.1 13.4 25.5 30.2 39.3 28.9 Central bank rediscount as percent of commercial bank credit 0.0 0.0 0.4 0.0 0.0 0.0 4.1 3.3 M3 velocity 4.1 3.7 3.6 3.5 3.2 3.2 3.6 3.5 M3 multiplier 3.0 3.3 3.3 3.5 3.4 3.7 4.0 4.1 M2 velocity 4.7 4.3 4.4 4.3 4.0 3.9 4.4 4.5 M2 multiplier 2.6 2.8 2.6 2.8 2.7 3.0 3.2 3.2 Gross official reserves (in US$ mill.) 112.7 131.0 177.3 266.5 323.0 332.2 295.6 288.5 Imports of Goods and Services -575.63 -444.17 -583.57 -501.28 -837.48 -896.02 -881.45 -910.43 FBu/USD (end of period) 997.78 1002.47 1119.54 1234.98 1230.5 1232.5 1361.5 1403.91 NIR Target in months of imports 3.6 6.4 4.6 4.4 4.0 3.8 Currency in circulation (as a share of deposits) 29.5 24.0 26.0 25.7 22.1 21.5 22.5 22.5 Credit to the private sector as a share of M2 -6.3 13.7 7.5 8.5 13.7 17.0 24.1 23.3 Source: IMF 186 Table I.3 Balance of Payments, 2005-2012 2005 2006 2007 2008 2009 2010 2011 2012 (In millions of U.S. dollars, unless otherwise indicated) % Current account (including official transfers) -54.8 -273.3 -72.7 GDP 30.5 -248.0 -323.0 -386.9 percent GDP -4.5 -20.9 -5.5 -1.0 1.8 -12.3 -14.8 -16.3 Trade balance -168.2 -317.4 -205.1 -259.1 -258.2 -614.1 -647.7 -697.9 Exports, f.o.b. 56.9 58.2 58.8 69.6 68.4 101.2 124.0 129.8 Of which: coffee 43.1 39.2 38.5 40.2 39.6 69.7 75.6 77.7 Imports, f.o.b. -225.1 -375.6 -264.0 -328.8 -326.6 -715.3 -771.7 -827.7 Of which: petroleum products -38.1 -57.6 -58.7 -90.7 -57.4 -98.4 -164.0 -172.2 Of which: imports related to reconstruction effort -59.6 -104.0 -137.8 -156.7 -158.4 -137.4 -201.4 -170.2 Services (net) -96.9 -165.5 -140.6 -171.6 -124.8 -88.0 -99.3 -103.4 Credits 34.8 34.5 39.6 83.2 49.9 78.9 101.4 106.6 Debits -131.7 -200.0 -180.2 -254.8 -174.7 -166.9 -200.6 -210.0 Income (net) -29.2 -19.4 -20.1 -8.4 -18.3 -11.8 -19.1 -16.3 Of which: Interest on public debt (including IMF and moratorium interest) -11.4 -10.6 -14.2 -4.1 -1.4 -1.0 -1.6 -2.9 Current transfers (net) 239.5 229.0 293.2 423.4 431.8 466.0 443.1 430.7 Private (net) 17.8 29.0 81.4 92.3 158.9 114.6 162.5 165.1 Official (net) 221.7 200.0 211.8 331.1 273.0 351.4 280.6 265.6 Program grants 36.9 65.7 80.8 113.7 65.2 101.3 77.7 79.1 Sectoral program grants 0.0 0.0 0.0 0.0 0.0 0.0 13.5 0.0 Humanitarian aid 73.5 120.0 81.2 84.4 102.4 63.2 51.4 38.7 Special program grants (DDR etc) 8.0 22.8 100.1 133.0 105.4 187.8 138.0 147.8 of which: Technical assistance grants 87.9 94.5 103.7 86.9 98.9 57.5 Other 103.3 -8.5 -50.3 0.0 0.0 -0.9 0.0 0.0 Capital and financial account 87.2 119.7 154.8 287.6 372.6 287.3 408.9 386.3 Capital account 26.2 75.8 170.3 196.4 1249.4 148.2 258.4 223.1 Capital transfers, net 26.2 75.8 170.3 196.4 1249.4 148.2 258.4 223.1 HIPC 8.0 35.7 39.4 39.7 1057.1 0.0 0.0 0.0 Debt forgiveness (IDA and AfDB HIPC and MDRI and bilateral and commercial HIPC SOD) MDRI grant (incl. HIPC) 0.0 0.0 0.0 0.0 51.0 0.0 0.0 0.0 Project grants 18.2 40.1 130.9 156.7 141.3 148.2 258.4 223.1 Other Acquisition or disposal of non-produced, non- financial assets 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Financial account 610 43.9 -15.5 91.2 -876.8 139.1 150.4 163.2 Direct investment 0.6 0.0 0.5 3.3 0.3 0.8 3.4 20.6 Portfolio investment 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Other investment 60.4 43.9 -16.0 88.0 -877.2 138.3 147.1 142.6 Assets -7.8 -30.2 -34.2 -30.8 -28.9 -43.6 -46.8 4.5 Trade credit -3.4 -15.2 -10.3 -7.6 -4.0 -37.8 -37.3 -30.9 Commercial banks -4.5 -15.0 -22.0 -23.2 -24.4 -5.8 -9.5 9.3 Other 0.0 0.0 -1.9 0.0 -0.6 0.0 0.0 26.2 Liabilities 68.2 74.1 18.2 118.8 -848.3 181.9 193.9 138.1 Trade credit 10.0 44.5 -4.3 50.5 124.6 100.9 131.2 125.9 187 Short-term 11.4 46.5 -2.2 52.1 126.2 101.0 127.9 122.5 Long-term -1.4 -2.0 -2.1 -1.6 -1.6 -0.1 3.3 3.4 Monetary authorities (Foreign liabilities of the BRB; assets are below the line) 14.4 25.4 21.1 65.5 50.0 25.4 30.1 20.5 o/w IMF 17.2 21.0 10.9 20.9 -28.3 20.3 18.2 7.7 Public sector 43.2 0.9 -22.2 -16.1 -1043.7 33.7 22.3 11.6 Short-term 6.4 1.4 3.1 2.4 0.0 0.0 0.0 0.0 o/w arrears 6.4 1.4 3.1 2.4 0.0 0.0 0.0 0.0 Long-term 36.8 -0.5 -25.2 -18.5 -1043.7 33.7 22.3 11.6 Disbursements 68.4 39.2 22.9 32.5 13.4 34.9 25.8 15.1 Project loans 41.5 39.2 22.9 32.5 13.4 30.0 25.8 15.1 Special program loans Program loans 26.9 0.0 0.0 0.0 0.0 5.0 0.0 0.0 Repayments -31.6 -39.7 -48.1 -51.0 -1057.1 -1.2 -3.5 -3.5 Other 0.1 -0.4 0.2 -0.1 -1.2 0.4 -0.3 -0.8 Commercial banks 0.6 3.8 23.4 18.9 22.0 21.5 10.5 18.4 Errors and omissions 1/ 3.4 166.8 -52.1 -186.3 -348.1 -25.8 -121.2 0.0 Overall balance 35.8 13.2 30.0 85.5 55.0 13.5 -35.3 -0.6 Financing (- increase in assets) -35.8 -13.2 -30.0 -85.5 -55.0 -13.5 35.3 0.6 Change in foreign assets of the BRB (– increase) -44.6 -17.9 -47.2 -89.5 -55.0 -13.5 35.3 0.6 Gross official reserves (assets only, not on a net basis) -45.5 -18.4 -46.2 -89.2 -56.5 -9.2 36.7 0.0 IMF, net Other, net (= Change in other reserve assets, with new BOP Presentation) 0.8 0.5 -1.0 -0.2 1.5 -4.3 -1.4 0.6 Change in arrears (+ increase) -22.2 0.8 -0.4 -48.0 0.0 0.0 0.0 0.0 of which: principal -13.1 0.8 0.0 -48.0 0.0 0.0 0.0 0.0 interest -9.1 0.0 -0.4 0.0 0.0 0.0 0.0 0.0 Exceptional financing 31.0 3.9 17.7 52.0 0.0 0.0 0.0 0.0 Cancellation 12.2 0.0 13.7 48.0 0.0 0.0 0.0 0.0 Rescheduling 18.9 3.9 4.0 4.0 0.0 0.0 0.0 0.0 Rescheduling of current debt service 5.1 3.9 4.0 4.0 0.0 0.0 0.0 0.0 Rescheduling of arrears 13.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Debt relief (HIPC bilateral) 0.0 Memorandum items: Nominal GDP (US$) 1211 1306 1310 1547 1739 2024 2182 2372 Trade balance/GDP -13.9 -24.3 -15.7 -16.7 -14.8 -30.3 -29.7 -29.4 Current account/GDP -4.5 -20.9 -5.5 -1.0 1.8 -12.3 -14.8 -16.3 Current account excluding official transfers -276.5 -473.3 -284.5 -346.8 -242.5 -599.4 -603.6 -652.6 % GDP -22.8 -36.2 -21.7 -22.4 -13.9 -29.6 -27.7 -27.5 Overall Balance 3.0 1.0 2.3 5.5 3.2 0.7 -1.6 0.0 Gross official reserves In U.S. dollars 112.7 131.0 177.3 266.5 323.0 332.2 295.6 269.4 In months of imports, c.i.f. 5.1 3.7 6.6 7.9 9.6 4.9 4.0 3.4 In months of imports of goods and services 3.8 2.7 4.8 5.5 7.7 4.5 3.6 3.1 In months of following period's imports of goods and services 2.3 3.5 3.6 6.4 4.4 4.1 3.4 3.3 Net reserves (in months of following period's imports of goods and services) 0.6 0.6 0.9 2.2 1.3 1.0 0.2 0.1 Exchange rate (Burundi francs per U.S. dollar; period average) 1081.6 1028.7 1081.9 1185.7 1230.2 1230.7 1261.1 1440.6 188 Table I.4 Fiscal Operations, 2008-2012 2008 2009 2010 2011 2012 Revenue 735.0 717.3 930.6 1,072.0 1,175.4 Taxes 229.8 278.7 341.7 424.1 477.1 Taxes on income, profits, and capital gains 65.0 84.7 108.7 131.0 173.7 Taxes on goods & services 121.8 147.3 193.7 248.6 263.0 Taxes on international trade & transactions 40.8 44.5 39.3 44.5 40.4 Taxes not elsewhere classified 2.2 2.2 0.0 0.0 0.0 Other revenue [NOTE: includes sale of fixed capital assets] 26.9 25.9 21.6 32.9 50.4 Grants 478.3 412.7 567.3 615.0 647.9 Budget Support 134.8 80.3 124.7 115.0 113.7 Program grants 134.8 72.4 124.7 97.9 113.7 Sectoral Budget Grants (Netherlands) 0.0 0.0 0.0 17.1 0.0 Exceptional COMESA grant 0.0 7.9 0.0 0.0 0.0 Project Support 185.8 173.9 182.4 325.9 321.4 Other Grants and Transfers 157.7 158.6 260.2 174.1 212.8 Amisom Fund 0.0 0.0 0.0 0.0 14.0 DDR support 45.2 0.0 20.0 26.6 0.0 Elections 0.0 2.0 43.5 0.0 0.0 Technical assistance grants 112.5 127.6 126.4 100.7 170.6 Global Fund 0.0 0.0 41.3 46.8 28.3 Special projects grants 0.0 0.0 0.0 0.0 0.0 MDRI grant IMF, WB, ADB, PC 0.0 29.0 29.0 0.0 0.0 Temporary social safety nets 0.0 0.0 0.0 0.0 0.0 Total expenditure 786.7 829.7 1021.5 1,189.5 1,234.7 Expense 515.9 552.0 722.6 726.3 792.3 Compensation of employees 158.4 191.8 214.9 258.2 283.8 Civilian 109.2 135.7 161.8 182.6 202.5 Military 29.5 34.3 31.8 45.6 50.5 National Police 19.8 21.8 21.4 30.0 30.8 Regularization of compensation arrears 0.0 0.0 18.8 22.0 6.0 Purchases/use of goods & services 95.4 104.0 100.3 97.9 99.0 Interest 29.0 26.3 14.9 25.5 33.1 Domestic 16.9 14.8 13.4 22.0 29.4 Foreign 12.1 11.5 1.5 3.5 3.7 Subsidies and Social benefits 71.6 100.3 101.4 140.0 159.3 Other Expense 161.5 129.6 272.3 182.6 211.0 Other Expense, domestically financed 3.8 0.0 41.1 8.6 12.2 Imprévu 3.8 0.0 0.0 0.0 3.8 Special Projects 0.0 0.0 0.0 0.0 0.0 Temporary social safety nets 0.0 0.0 23.4 0.0 0.0 189 Depense de fonds 0.0 0.0 8.1 8.6 8.4 Elections, domestically financed 0.0 0.0 9.7 0.0 0.0 Other Expense, externally financed 157.7 129.6 231.2 174.1 198.8 DDR 45.2 0.0 20.0 26.6 0.0 Elections, externally financed 0.0 2.0 43.5 0.0 0.0 Global Fund 0.0 0.0 41.3 46.8 28.3 Technical assistance 112.5 127.6 126.4 100.7 170.6 Expense not elsewhere classified 233.1 229.9 392.6 344.6 376.3 Net acquisition of nonfinancial assets 270.8 277.7 298.9 463.2 442.5 Domestic resources 46.5 58.4 50.6 104.7 99.3 Project loans 38.5 16.5 36.9 32.6 21.8 Project grants 185.8 173.9 182.4 325.9 321.4 MDRI related expenditure 0.0 29.0 29.0 0.0 0.0 NET LENDING/BORROWING [commitment basis] -51.7 -112.4 -91.0 -117.4 -59.3 Errors and Omissions 17.2 28.0 8.9 17.4 0.0 Net acquisition of financial assets -1.0 14.1 15.1 -11.4 -16.5 Of which, Domestic -1.0 14.1 15.1 -11.4 -16.5 Foreign 0.0 0.0 0.0 0.0 0.0 Net incurrence of liabilities 33.4 98.4 97.2 88.7 42.9 Of which, Domestic 8.0 83.0 55.7 60.3 26.8 Foreign 25.4 15.4 41.5 28.4 16.0 Financing gap 0.0 0.0 0.0 0.0 0.00 Source : IMF Table I.5 Burundi: Inflation Rate Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2010 4.47 5.45 6.14 6.68 8.72 9.70 9.79 8.52 3.50 4.67 5.51 4.08 2011 4.76 4.99 5.70 8.43 7.48 8.60 9.07 11.10 11.71 13.34 16.41 14.86 2012 21.63 21.99 24.51 23.24 22.51 17.28 17.64 15.66 14.25 15.89 11.70 11.84 Source : International Financial Statistics, 2013 Table I.6 Burundi: Exchange rate (BFu/USD) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2010 1230.05 1230.00 1230.00 1230.10 1230.10 1230.32 1230.87 1230.90 1230.91 1231.04 1231.89 1232.80 2011 1233.30 1234.82 1236.37 1237.54 1239.34 1241.35 1246.41 1254.66 1265.58 1280.05 1314.47 1348.99 2012 1375.71 1397.95 1405.40 1394.77 1402.99 1419.16 1456.15 1466.67 1474.39 1486.53 1502.09 1528.26 Source : FX Rates 190 Annex J : List of people consulted during the PER preparation 1. To ensure effective dialogue with government authorities and maximize the joint ownership of the study, the Bank Task team carried out 6 missions to Burundi. Specifically, the consultation process with government and representatives of the donor community (September- November 2012) was followed by two rounds of data collection efforts (December 2012 and February 2013, respectively). Preliminary drafts of the analysis were shared during two subsequent missions (April 2013), to give the authorities the opportunity to react and provide feedback to the team. A final mission to present the full PER report to the authorities is scheduled for June 2013. The list of people met and interviewed during the missions and their respective institutional affiliation is provided below. Table J.1 Government of Burundi Name Title MDA Colette Ndayizeye Directeur du Budget a.i Ministere des Finances Prosper Girukwishaka Directeur du Trésor Ministere des Finances David Kamana Chef des service reddition Ministere des Finances des comptes Nolasque Ndikumana Chef de service prevision et Ministere des Finances preparation budgétaire / Dir budget Thérence Ntabangana Chef du Service Prévision Ministere des Finances Antonine Batungwanayo Directeur de la prevision et Ministere des Finances – Ciza prospective Christian Kwizera Coordonateur adjoint de la Ministere des Finances cellule des réformes Emile Nimbona Expert National en base de C.N.C.A données, suivi et evaluation des programmes Immaculée Bigirimana Secrétaire Permanent Ministère des Finances Rose Marie Kankindi Directrice de la Paie Ministere de la Fonction Publique. Pamphile Muderega Secrétaire Exécutif S.E. / C.N.C.A Pasteur Mpawenimana Directeur de l’informatique Ministere des Finances (SIGEFI) Leonard Kayobera Secrétaire Permanent Ministere de la Fonction Publique Aude Toyi Directrice du Bureau des Ministere des Finances Grands Contribuables / O.B.R Jacques Ngendakumana Coordonnateur de la Cellule Ministere des Finances des Réformes 191 Domitien Ndihokubwayo Commissaire Général adjoint O.B.R / Ministère des Finances Béatrice Samandari Directrice du Budget Ministere des Finances Tesfaye Wondim Conseiller du Commissaire Ministere des Finances des Taxes Interne et des Recettes Non-Fiscales / O.B.R Jacques Ngendakumana Coordonateur de la Cellule Ministère des Finances et de la d’appui au suivi des réformes Planification du et du cadre de partenariat Développement Economique entre le Gouvernement et les Bailleurs de Fonds Désiré Musharitse Coordonateur-Adjoint de la Ministère des Finances et de la cellule d’appui au suivi des Planification du réformes et du cadre de Développement Economique partenariat entre le Gouvernement et les Bailleurs de fonds Nestor Manirakiza Directeur de la comptabilité Ministère des Finances et de la et du Trésor Planification du Développement Economique Kieran Holmes Commissaire Général de Ministère des Finances et de la l’Office Burundais des Planification du Recettes (OBR) Développement Economique Monsieur Nestor Directeur de la comptabilité Ministère des Finances et de la Manirakiza et du Trésor Planification du Développement Économique Monsieur Pasteur Directeur de l’informatique Ministère des Finances et de la Mpawenimana (SIGEFI) Planification du Développement Économique Monsieur Léonce Coordinateur Projet d’Appui à la Gestion Sinzinkayo Économique Monsieur Alexandre Commissaire Général Service Chargé des Entreprises Nakumuryango Publiques (SCEP) Monsieur Nicolas Directeur Général Institut des Statistiques et des Ndayishimiye Études Économiques du Burundi (ISTEEBU) Leonard Kayobera Permanent Secretary Fonction Publique Rose Marie Kankindi Directrice de la Paie Fonction Publique Fonction Publique 192 Tesfaye Wondim Consultant OBR Jean Pierre Amani Directeur Juridique et gestion OBR des contentieux Modeste Kabaka Consultant OBR Felix Niyonzima Responsable adjoint du Central Bank Service études et statistiques Audace Niyonzima Responsable du service Central Bank études et statistiques Antonine Batungwanayo Directeur de la prévision et MTEF team Ciza prospective Nolasque Ndikumana Chef de service prévision et MTEF Team préparation du budget Ntabangana Therence Chef du service prévision MTEF Team Christian Kwizera Coordonnateur adjoint de la MTEF cellule des réformes Ildephonse Rwasamanga Cadre à la direction de la MTEF team programmation Diop Modou Bousso Consultant ARMP Onesphore Baroreraho Directeur Général ONPR Nzigamasabo Philbert Directeur Technique ARMP Njejimana J. Berchmans Attaché de Direction ARFIC Ndayisenga Dieudonné Secrétaire Permanent DNCMP Ndayishimiye Nicolas DG ISTEEBU Tuyaga Anicet DG OTB Ntangibingura Amissi DG UH Birehanisenge Adolphe API Kamikazi Josiane PF Banque Mondiale Ministère des Finances Ruvakubusa Chantal Commissaire OBR Nzeyimana Thacien Directeur OBR Batungwanayo Balthazar Responsible Admnistratif CTI Kamariza Alice F.P. DGI Masenge Evelyne DF INSS 193 Table J.2 Development Partners NAME TITLE Organization Gerard Muringa Conseiller charge de la Ambassade des Pays Bas Coopération et du Secteur privé Theofiel Baert Ministre Conseiller Ambassade de Belgique Yurgen Heimann Chef de Coopération Union Européenne Chargée du Programme Délégation de l’Union Appui budgétaire et Finances Européenne Madame Yata Publiques Charlotte Chargée de Programme Délégation de l’Union Tilli Blomhammar Bonne Gouvernance Européenne Monsieur Eric Charvet Régisseur du Projet Gutwara Délégation de l’Union Neza Européenne Monsieur Marc Gedopt Ambassadeur Ambassade du Royaume de Belgique Monsieur Guy Attaché à la Coopération Ambassade du Royaume Hambrouck de Belgique Monsieur Michel Verge Conseiller Technique chargé Cooperation Technique du dossier de la Bonne Belge (CTB) Gouvernance Monsieur Hervé Corbel Chargé du Programme Coopération Technique Education et Gouvernance Belge (CTB) Claire Galante Représentante AFD Paul Vossen Conseiller Délégation de l’Union Européenne Svenja Ossmann Intérimat à l’appui Délégation de l’Union budgétaire Européenne Augustin Roitman Economiste FMI Koffi Yao Représentant FMI Tokindang Joel Economiste BAD Ulrika Joyce Responsable du Section Ambassade des USA Consulaire Sori-Coulibaly Rosine Représentant Spécial du UN/UNDP Secrétaire Général des Nations Unies / Représentant Résident du PNUD Fritz Meijndert Chef de Mission Adjoint Ambassade des Pays-Bas Droyer Elisabeth Conseiller Chef du Bureau Ambassade de Norvège Bruno Brommer Ambassadeur Ambassade d’Allemagne 194 MAP OF BURUNDI 195