57880 W O R L D B A N K W O R K I N G P A P E R N O . 2 0 5 A F R I C A H U M A N D E V E L O P M E N T S E R I E S Budgeting for Effectiveness in Rwanda From Reconstruction to Reform THE WORLD BANK W O R L D B A N K W O R K I N G P A P E R N O . 2 0 5 Budgeting for Effectiveness in Rwanda From Reconstruction to Reform Copyright © 2010 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First Printing: October 2010 Printed on recycled paper 1 2 3 4 13 12 11 10 World Bank Working Papers are published to communicate the results of the Bank's work to the development community with the least possible delay. The manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formallyedited texts. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those 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 of the legal status of any territory or the endorsement or acceptance of such boundaries. 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 promptly to reproduce portions of the work. 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, Tel: 9787508400, Fax: 9787504470, 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: 202522 2422, email: pubrights@worldbank.org. ISBN13: 9780821385586 eISBN: 9780821385616 ISSN: 17265878 DOI: 10.1596/9780821385586 Library of Congress CataloginginPublication Data has been requested. Contents Foreword by John Rwangombwa ......................................................................................... ii x Foreword by Yaw Ansu ........................................................................................................ xiii Acknowledgments ................................................................................................................. xiv Acronyms and Abbreviations ............................................................................................... xv Executive Summary ............................................................................................................ xviii . 1. Introduction ........................................................................................................................... 1 . 2. General Budget Support (GBS): Relevance, Rationale, and Remaining Challenges ........................................................................................................................... 6 Historical Background and Trends in Harmonized GBS .............................................. 6 Progress in GBSRelated Processes and Practices ........................................................ 12 . Economic and Structural Reforms 2004­2007 Supported by GBS ............................. 15 . 3. Domestic and External Resources, Recurrent and Capital Expenditures, Overview of Aid Dependency ....................................................................................... 27 Resources ........................................................................................................................... 30 Domestic Resources .......................................................................................................... 35 External Resources ............................................................................................................ 39 Expenditures ..................................................................................................................... 43 . Recurrent Expenditures ................................................................................................... 46 Capital Expenditures ........................................................................................................ 50 Overview of Aid Dependency ........................................................................................ 54 4. Public Expenditure Overview by Ministry Level ......................................................... 64 . Overview of Recurrent and Development Expenditure by Line Ministry ............... 64 Trends in Recurrent and Development Expenditure ................................................... 68 5. Policy Overview, Resource Allocation, and Results ................................................... 103 Priority Sectors ................................................................................................................ 106 Nonpriority Sectors Snapshots ..................................................................................... 224 . 6. Looking Forward: Outstanding Challenges and Concluding Remarks .................. 254 Recent Achievements and Outstanding Challenges .................................................. 254 Concluding Remarks ...................................................................................................... 255 Appendixes ............................................................................................................................. 259 Appendix A. Government Reforms 1995­2003 .......................................................... 261 . Appendix B. Joint Analytical Work Summary 2004­07 ............................................. 262 Appendix C. Case Study of District of Rulindo 2006­08 ........................................... 263 Appendix D. Map of Rwanda ....................................................................................... 269 iii iv Contents Appendix E. Decentralization Structure in Rwanda .................................................. 270 Appendix F. Available Data Sets .................................................................................. 271 References ............................................................................................................................... 273 Tables Table 2.1. Selected Economic and Social Indicators .............................................................. 8 Table 2.2. Development Assistance 1995 Roundtable Conference (US$ millions)............ 9 Table 2.3. Reforms Implemented by the Government of Rwanda (2004­07) .................. 23 Table 3.1. Summary of Net Resources in Nominal Terms 2004­07 .................................. 28 Table 3.2. Summary of Net Resources in Real Terms 2004­07 .......................................... 29 Table 3.3. Line Ministries Expenditures and Resources as a Percentage of GDP ........... 29 Table 3.4. Overview of Resources--Original Projections and Actual 2004­07 ............... 31 Table 3.5. Overall Original Projections for Domestic and External Resources 2004­07 .............................................................................................................................. 33 Table 3.6. Overall Actual Domestic and External Resources 2004­07 .............................. 34 Table 3.7. Overview of Domestic Resources--Original Projections vs. Actual 2004­07 ................................................................................................................. 35 . Table 3.8. Original Projections for Domestic Resources 2004­07 ...................................... 37 Table 3.9. Actual Domestic Resources for 2004­07 ............................................................. 38 Table 3.10. Increase in VAT Tax Projections 2004­07 ......................................................... 39 Table 3.11. Overview of External Resources--Original Projections and Actual 2004­07 .... 39 . Table 3.12. Original Projections for External Resources 2004­07 ...................................... 41 Table 3.13. Actual Funding from External Resources 2004­07 .......................................... 42 Table 3.14. Overview of Expenditures--Original Projections and Actual 2004­07 ....... 43 . Table 3.15. Original Expenditures for 2004­07 .................................................................... 45 Table 3.16. Actual Expenditures 2004­07 ............................................................................. 46 Table 3.17. Overview of Recurrent Expenditures 2004­07 ................................................ 47 . Table 3.18. Original Recurrent Expenditures 2004­07 ........................................................ 49 Table 3.19. Actual Recurrent Expenditures 2004­07 ........................................................... 50 Table 3.20. Overview of Capital Expenditures and Net Lending 2004­07 ...................... 51 Table 3.21. Original Capital Expenditures and Net Lending 2004­07.............................. 52 Table 3.22. Actual Capital Expenditure and Net Lending 2004­07 .................................. 53 Table 4.1. Total Actual Expenditures in Nominal Terms in RwF millions 2004­07 ....... 65 Table 4.2. Total Actual Expenditures in Real Terms in RwF millions 2004­07 ............... 65 Table 4.3. Increase in Actual Expenditure in Real Terms in RwF millions 2004­07 ....... 67 Table 4.4. Allocation to Ministries of Increase in Actual Public Expenditure 2004­07 .............................................................................................................................. 68 Table 4.5. Total Actual Recurrent Expenditure by Ministry in Nominal Terms 2004­07 .............................................................................................................................. 69 Table 4.6. Total Actual Recurrent Expenditure by Ministry in Real Terms 2004­07 ...... 70 Table 4.7. Share of Actual Recurrent Expenditures of the Two Largest Ministries (Central Level Only) 2004­07 ......................................................................................... 70 Table 4.8. Four Largest Ministries by Actual Recurrent Expenditure at Central Level 2004­07 ................................................................................................................... 71 . Contents v Table 4.9. Four Largest Ministries' Share of CentralLevel Recurrent Actual Expenditure ...................................................................................................................... 72 Table 4.10. Total Recurrent Actual CentralLevel Expenditures by Ministries 2004­07 .............................................................................................................................. 72 Table 4.11. Percentage of Total Recurrent Actual Public Expenditure by Economic Classification 2004­07 ..................................................................................................... 73 . Table 4.12. Recurrent Central Only Executed Public Expenditure Trend........................ 74 Table 4.13. Recurrent Central and Decentralized Executed Public Expenditure Trend 2004­07 .................................................................................................................. 75 . Table 4.14. Share of Recurrent Expenditures Transferred to Districts 2004­07 .............. 76 Table 4.15. All Other Ministries' Executed Recurrent Public Expenditure at Central Level 2004­07 ..................................................................................................... 77 . Table 4.16. All Other Ministries' Executed Recurrent Public Expenditure, including District Transfers 2004­07 ............................................................................. 78 Table 4.17. Actual District Transfers 2004­07 ...................................................................... 79 Table 4.18. Execution Rates for Earmarked Transfers to Priority Sectors 2004­07 ......... 80 Table 4.19. Actual District Transfers in Nonpriority Sectors 2004­07 .............................. 80 Table 4.20. Execution Rates for Transfers in Nonpriority Sectors 2004­07 ...................... 81 Table 4.21. Overall Actual Development Expenditure Trends 2004­07 .......................... 82 . Table 4.22. Overall Actual Development Spending by Source of Funding 200407 ....... 83 Table 4.23. Overall Budgeted Development Spending by Source of Funding 2004­ 07 ........................................................................................................................................ 83 Table 4.24. Actual Development Budget for Five Priority Ministries 2004­07 ............... 84 . Table 4.25. Overall Actual Development Expenditures 2004­07 ...................................... 85 Table 4.26. Sources of Original Development Expenditure Allocation 2004­07 ............. 87 Table 4.27. Original Development Expenditure Allocation for Priority Ministries by Funding Sources 2004­07 .......................................................................................... 88 Table 4.28. Comparing Internally Financed Development Original Budget Data 2004­ 07 ........................................................................................................................................ 89 Table 4.29. Comparing Counterpart Development Budgets 2004­07 .............................. 90 Table 4.30. Comparing Donor Development Budgets 2004­07 ......................................... 91 Table 4.31. Comparing Loans Development Budgets 2004­07 ......................................... 92 . Table 4.32. Comparing Sources of Actual Development Budget Funding 2006­07 ....... 93 Table 4.33. Sources of Development Budget Funding for the Five Priority Sectors 2006­07 .............................................................................................................................. 94 Table 4.34. Internal and External Development Funding for MINEDUC 2004­07 ........ 95 Table 4.35. Internal and External Development Funding for MINISANTE 2004­07 ..... 96 Table 4.36. Internal and External Development Funding for MINAGRI 2004­07 .......... 97 Table 4.37. Internal and External Development Funding for MININFRA 2004­07 ....... 98 Table 4.38. Internal and External Development Funding for MINITERRE 2004­07 ...... 99 Table 4.39. Overall Actual Development Expenditure Trends in Nonpriority Ministries 2004­07 ......................................................................................................... 100 . Table 4.40. Overview of Actual Development Budgets of Four Major Nonpriority Ministries 2004­07 ......................................................................................................... 101 . Table 5.1. Poverty Indicators and Geographical Distribution ......................................... 104 Table 5.2. Agriculture Programs and Subprograms ......................................................... 107 vi Contents Table 5.3. Farm Land Distribution by Types of Crop ....................................................... 107 Table 5.4. Proportion of Total Agricultural Output Contributed by Four Major Food Groups 2000­05 .................................................................................................... 108 Table 5.5. Evolution of National Livestock Population 2000­05 ..................................... 108 Table 5.6. Tax Subsidies in the Commercial Agriculture Sector ..................................... 109 Table 5.7. MINAGRI Budget Allocations 2000­07 ............................................................ 112 Table 5.8. MINAGRI and National Budget Allocations 2000­07 .................................... 113 Table 5.9. Budget Allocations to MINAGRI and Other Key Ministries 2007 ................ 114 Table 5.10. Budget Allocations: Agriculture and Social Sectors Compared 2004­07 ... 115 Table 5.11. MINAGRI Budget and Actual Expenditure 2003­05 .................................... 116 Table 5.12. 2006 MINAGRI Execution Rates ...................................................................... 117 Table 5.13. Agriculture Budget as Proportion of Agriculture GDP 2000­06 ................. 117 Table 5.14. Overall Picture of District Allocations for Agriculture in 2007 ................... 118 Table 5.15. Proportion of Agriculture Recurrent Budget and Prevalence of Poverty by Province ..................................................................................................................... 118 Table 5.16. Earmarked Budget Lines in Districts 2007 ..................................................... 119 Table 5.17. Source of Funding of MINAGRI's Development Budget 2000­07 .............. 119 Table 5.18. Main Development Agencies in the Agriculture Sector ............................... 120 Table 5.19. General Budget Support Partners ................................................................... 120 . Table 5.20. Government Funding of MINAGRI's Development Budget 2003­07 ........ 120 Table 5.21. Comparing On and OffBudget Support from Donors in 2006 and 2007 .......................................................................................................................... 121 Table 5.22. Principal AgricultureRelated OffBudget Interventions by Development Partners 2006­07 .................................................................................... 122 Table 5.23. Agriculture Recurrent and Development Budget 2000­07 .......................... 123 Table 5.24. Development Budget Allocations to Recurrent and Capital Cost .............. 124 . Table 5.25. Recurrent Budget Allocations to MINAGRI Departments 2000­05 ........... 124 . Table 5.26. Budget Allocations to MINAGRI by Program 2006 and 2007 ..................... 125 Table 5.27. Agriculture Sector Reform Measures 2004­07 ............................................... 126 Table 5.28. Seed Production vs. Seed Needs 2005­06 ....................................................... 128 Table 5.29. Nutritional Needs Satisfaction between 2001 and 2005 ............................... 128 . Table 5.30. Food Security Profile for Rural Population, 2006 .......................................... 128 Table 5.31. Highest Food Insecurity Zones in Rwanda .................................................... 129 Table 5.32. Development of Total Bank Loans in the Agriculture Sector 2002­05 ....... 129 Table 5.33. Trends in Coffee and Tea Production and in Exports 2000­05 .................... 130 Table 5.34. Literacy Rates in Rwanda ................................................................................. 131 Table 5.35. Education Sector Policies .................................................................................. 132 Table 5.36. Comparative Education Expenditures ............................................................ 134 Table 5.37. Education Sector Recurrent Budget 2001­07 ................................................. 135 . Table 5.38. Education Recurrent Budget by Economic Classification ............................ 136 Table 5.39. Education Recurrent Budget and Budget Execution 2001­06 ..................... 137 . Table 5.40. Teacher Salary Execution by District 2006 ..................................................... 138 . Table 5.41. Teacher Salary Actual District Budget 2006 and Original District Budget 2007 .................................................................................................................... 139 Table 5.42. Average Teacher Salaries 2007 ......................................................................... 140 Table 5.43. Education Budgets for Districts 2007 .............................................................. 141 Contents vii Table 5.44. Education Budget, FTI, and LTSFF Allocations 2006 and 2007 ................... 142 Table 5.45. Higher Education Recurrent Budget 2007 ...................................................... 143 Table 5.46. CNER Income 2001­06 ...................................................................................... 143 Table 5.47. Apparent Unit Government Recurrent Expenditure by Level 2001­06 ..... 144 Table 5.48. Nominal and Real Unit Education Expenditures 2006 ................................. 145 Table 5.49. Education Enrollment and Real Unit Expenditure Change 2001­06 .......... 145 Table 5.50. Sample* Households by School Type and Level** ........................................ 146 Table 5.51. Mean Total Education Expenditure by Household ...................................... 146 . Table 5.52. Mean Total Household Consumption Expenditure by Quintile ................. 147 Table 5.53. Median Education Expenditure per Student in Last 12 Months of Primary Education ......................................................................................................... 147 Table 5.54. Median Education Expenditure per Type of School ..................................... 148 Table 5.55. Median Education Expenditure per Student by Age Category ................... 148 Table 5.56. Reform Measures Undertaken in Education 2004­07 ................................... 149 Table 5.57. Enrollment by Level 2002­06 ........................................................................... 150 Table 5.58. Mean Annual Education Expenditure per Household by Region .............. 151 Table 5.59. Mean Annual Education Expenditure per Funded Individual by Region ............................................................................................................................. 151 . Table 5.60. Mean Number of Persons per Household Funded for Education, by Region ............................................................................................................................. 152 . Table 5.61. Mean Individual Education Expenditure by Age and Location ................. 152 Table 5.62. Mean Individual Education Expenditure by Quintile and Location .......... 153 Table 5.63. Policy Objectives and Expected Impact on Poverty and Growth of the HSSPI (2005­09) ............................................................................................................ 154 Table 5.64. Health Expenditure to National Expenditures (2002­07) ............................ 158 . Table 5.65. Public Health Expenditure by Economic Classification (2002­07) .............. 158 Table 5.66. Sources of Public Finance for Health (2003­07) ............................................. 159 Table 5.67. Rwanda's Child Mortality Rates (in 1000s) .................................................... 163 Table 5.68. Rwanda's Maternal Mortality Ratio per 100,000 Live Births ....................... 164 Table 5.69. Reasons for Not Visiting a Government Service Facility since Last Year .. 168 Table 5.70. Summary of Main Judgments Used in Defining Social Protection ............. 171 Table 5.71. Summary of Main Judgments Not Used in Defining Social Protection ..... 171 Table 5.72. Total Social Protection Spending Overview .................................................. 172 . Table 5.73. Distribution of Recurrent Social Protection Expenditure ............................. 172 Table 5.74. Overview of Allocation of Total Social Protection Funding ........................ 173 Table 5.75. Actual and Budgeted Total Social Protection Expenditures 2004­07 ......... 174 Table 5.76. Recurrent Spending on Social Protection by Ministries and Local Governments .................................................................................................................. 176 Table 5.77. Allocation of Government Social Protection Spending: Budget vs. Actual 2004­07 ............................................................................................................... 177 . Table 5.78. Allocation of OffBudget Donors' Social Protection Expenditure .............. 179 . Table 5.79. Allocation of Total Social Protection Funding 2004­07 ................................ 180 Table 5.80. Water and Sanitation Rehabilitation, Construction, and Management 2003­05 ............................................................................................................................ 183 Table 5.81. Water and Sanitation Financing Mechanisms under MINALOC ............... 183 Table 5.82. Rwanda's Water and Sanitation Sector Policy Overview ............................ 184 . viii Contents Table 5.83. Water and Sanitation CrossSector Involvement .......................................... 184 . Table 5.84. Institutional Roles and Responsibilities in the Water and Sanitation Sector ............................................................................................................................... 185 Table 5.85. Population Trends 1997­2020 .......................................................................... 186 . Table 5.86. Water and Sanitation Resource Flow 2003­05 ............................................... 187 Table 5.87. Water and Sanitation Expenditure by Subprogram 2003­05 ....................... 188 Table 5.88. Multilateral and Bilateral Donors in the Water and Sanitation Sector ....... 189 Table 5.89. Water and Sanitation Project Disbursement and Realization 2003­05 ....... 189 Table 5.90. Water and Sanitation Sector Development Budget 2003­04 ........................ 190 Table 5.91. Breakdown by Program Component Showing New Investment Requirements in Rural Areas to Meet 2020 Objectives ............................................. 191 Table 5.92. Water and Sanitation Investment Requirements by Region ........................ 192 Table 5.93. Water and Sanitation Coverage Targets and Estimated Investment Requirements ................................................................................................................. 192 . Table 5.94. Water and Sanitation Reform Measures 2004­05 .......................................... 193 Table 5.95. Water and Sanitation Performance Indicators 2001­15 ................................ 194 Table 5.96. Urban Sanitation Coverage Rates 2001 ........................................................... 194 Table 5.97. Water and Sanitation Performance Indicators 2001­05 ................................ 195 Table 5.98. Water Infrastructure in Rural and Semiurban Areas 2005 ........................... 197 Table 5.99. Rural Infrastructure Improvements in 2005 ................................................... 197 Table 5.100. Water and Sanitation Sector Actual Expenditure Compared with Outcomes in 2003­05 ..................................................................................................... 198 Table 5.101. Public Budget and the Energy Sector, Rwanda 2001­07 ............................ 202 Table 5.102. Ameliorating Shortages, Expanding Geographic Access ........................... 203 Table 5.103. Initiatives in Nonconventional Energy ......................................................... 203 Table 5.104. Electrogaz Supply/Demand Balance ............................................................. 204 Table 5.105. Implications of Paris Declaration Harmonization and Alignment Agenda for the Energy Sector ...................................................................................... 206 Table 5.106. Energy Sector MediumTerm Expenditure Framework 2008­10 .............. 207 Table 5.107. Key Activities Proposed for 2008­12 ............................................................. 207 Table 5.108. A Typology of Public Finance Instruments for Energy .............................. 216 Table 5.109. Transport Sector Development Expenditures by Program 2004­07 ......... 220 Table 5.110. Transport Sector Recurrent Expenditures by Program 2007 ..................... 221 . Table 5.111. Road Network Progress as of 2007 ................................................................ 221 Table 5.112. Rural Gravel Network Progress as of 2007 (ongoing maintenance contracts) ......................................................................................................................... 222 Table 5.113. 2006 Road Conditions Status Serving as a Baseline for EDPRS 2008­12 .. 224 Table 5.114. Snapshot--PARLEMENT--Office of Parliament (Million RwF) .............. 225 Table 5.115. Snapshot--PRESIREP--Office of the President........................................... 227 Table 5.116. Snapshot--PRIMATURE--Office of Prime Minister .................................. 228 Table 5.117. Snapshot--COURS.SUP--Office of the Supreme Court ............................ 230 Table 5.118. Snapshot--MINADEF--Ministry of Defense .............................................. 232 Table 5.119. Snapshot--MININTER--Ministry of Internal Security .............................. 234 Table 5.120. Snapshot--MINAFFET--Ministry of Foreign Affairs and Cooperation . 236 . Table 5.121. Snapshot--MINICOM--Ministry of Commerce, Industry, Investment Promotion, Tourism and Cooperatives ...................................................................... 238 . Contents ix Table 5.122. Snapshot--MINISTR--Ministry of Science, Technology, and Scientific Research .......................................................................................................................... 240 Table 5.123. Snapshot--MINECOFIN--Ministry of Finance and Economic Planning .......................................................................................................................... 240 Table 5.124. Snapshot--MINIJUST--Ministry of Justice ................................................. 242 Table 5.125. Snapshot--MIJESPOC--Ministry of Youth, Sports, and Culture (prior to March 2008) ................................................................................................................ 244 Table 5.126. Snapshot--PARQUET GENERAL--Office of the Prosecutor General .... 245 Table 5.127. Snapshot--MIGEPROF--Office of Gender and Family Promotion (as of 2006 under Office of Prime Minister)...................................................................... 245 Table 5.128. Snapshot--MIFOTRA--Ministry of Public Service and Labor ................. 247 Table 5.129. Snapshot--MINALOC--Ministry of Local Government, Good Governance, Community Development, and Social Affairs .................................... 249 Table B.1. Joint Analytical Work by Donors and/or Country--Rwanda 2004­07 ........ 262 Table B.2. Joint Technical Assistance Work by Donors and/or Country--Rwanda 2004­07 ............................................................................................................................ 262 Figures Figure 1.1. Overview of the Analytical Structure That Provides the Basis for This Analysis ............................................................................................................................... 5 Figure 2.1. Trends in General Budget Support (in US$ millions) ..................................... 11 Figure 2.2. Share of Current Grants (including GBS) as a Percentage of Total Budget Support ................................................................................................................ 11 Figure 2.3. Share of Total Budget Support as a Percentage of Total Official Development Assistance ................................................................................................. 12 Figure 2.4. Average Budget Aid Arriving on Schedule in Selected Number of Countries ........................................................................................................................... 24 Figure 3.1. Resource Trends Comparing Original Projections vs. Actual 2004­07 ........ 31 . Figure 3.2. Resource Trends by Type of Resource 2004­07 ............................................... 32 Figure 3.3. Overall Original Projections for Domestic and External Resources 2004­07 .............................................................................................................................. 33 Figure 3.4. Overall Actual Resource Trends by Type of Resource 2004­07 .................... 34 . Figure 3.5. Domestic Revenue Trends Comparing Original Projections vs. Actual 2004­07 .............................................................................................................................. 36 Figure 3.6. Domestic Revenue Trends by Type of Revenue Source 2004­07................... 36 Figure 3.7. Domestic Revenue Trends in Original Projections 2004­07 ........................... 37 Figure 3.8. Actual Domestic Revenue Trends 2004­07 ...................................................... 38 . Figure 3.9. External Resource Trends Comparing Original Projections vs. Actual 2004­07 .............................................................................................................................. 40 Figure 3.10. External Resource Trends by Type of Resource 2004­07 .............................. 40 Figure 3.11. Budgeted External Resource Trends 2004­07 ................................................ 41 . Figure 3.12. Actual External Resource Trends 2004­07 ...................................................... 42 Figure 3.13. Expenditure Trends Comparing Original and Actual 2004­07 ................... 44 Figure 3.14. Overall Expenditure Trends by Type 2004­07 ............................................... 44 Figure 3.15. Original Overall Expenditure Trends 2004­07 .............................................. 45 . Figure 3.16. Actual Overall Expenditure Trends 2004­07 ................................................. 46 . x Contents Figure 3.17. Recurrent Expenditure Trends Comparing Original vs. Actual 2004­07 .............................................................................................................................. 48 Figure 3.18. Recurrent Expenditure Trends by Type 2004­07 ........................................... 48 Figure 3.19 Original Recurrent Expenditure Trends 2004­07 ........................................... 49 Figure 3.20. Actual Recurrent Expenditure Trends 2004­07 ............................................. 50 Figure 3.21. Capital Expenditures and Net Lending Comparing Original vs. Actual 2004­07 ................................................................................................................. 51 . Figure 3.22. Capital Expenditures and Net Lending Trends 2004­07 .............................. 52 Figure 3.23. Original Capital Expenditures and Net Lending Trends 2004­07............... 53 Figure 3.24. Actual Capital Expenditure and Net Lending Trends 2004­07 ................... 54 Figure 3.25. Total Expenditures and Net Lending 2001­06 ............................................... 57 Figure 3.26. Total Aid as a Percentage of Total Expenditure and a Percentage of GDP 2000­06 .................................................................................................................... 58 . Figure 3.27. Domestic Revenue as a Percentage of GDP 2001­06 ..................................... 58 Figure 3.28. Trends in Total Revenue and Grants, and Total Expenditure and Net Lending 2001­06 .............................................................................................................. 59 Figure 3.29. Net Present Value of DebttoExport Ratio 2001­06 ..................................... 61 . Figure 3.30. Total Aid, Broken Down by Loans and Grants 2000­06 ............................... 62 Figure 3.31. Loans and Grants as Share of Total Aid 2000­06 ........................................... 62 Figure 4.1. Total Actual Expenditure Trend in Real Terms 20042007 ............................. 65 Figure 4.2. Comparison of Actual and Revised Budget Expenditures 2004­07 .............. 66 Figure 4.3. Four Largest Ministries' Actual Recurrent Expenditure Trends 2004­07 .... 71 Figure 4.4. Share of Total Recurrent Expenditures Transferred to Districts' Key Programs ........................................................................................................................... 75 Figure 4.5. Actual Development Spending Trends 2004­07 .............................................. 85 Figure 4.6. Original Development Budget versus Actual Development Expenditures for Five Priority Sectors 2004­07 ........................................................... 86 Figure 4.7. Trends in Development Budget Internally Financed Funding 2004­07 ....... 89 Figure 4.8. Trends in Development Budget Counterpart Funding 2004­07 ................... 90 . Figure 4.9. Trends in Development Budget Donor Funding 2004­07 .............................. 91 Figure 4.10. Trends in Development Budget Loan Funding 2004­07 .............................. 92 Figure 4.11. Internal and External Development Funding for MINEDUC 2004­07 ....... 95 Figure 4.12. Internal and External Development Funding for MINISANTE 2004­07 .............................................................................................................................. 96 Figure 4.13. Internal and External Development Funding for MINAGRI 2005­07 ........ 97 Figure 4.14. Internal and External Development Funding for MININFRA 2004­07 ..... 98 . Figure 4.15. Internal and External Development Funding for MINITERRE 2004­07 .... 99 Figure 5.1. Total 2007 Budget Allocations by Ministry .................................................... 115 Figure 5.2. Total 2007 Recurrent Budget Allocations by Ministry .................................. 116 Figure 5.3. Per Capita Health Expenditure (1998­2006) in 2006 constant US$ Prices .. 159 Figure 5.4. Contributions to the Health Sector as a Percentage of THE (2006) ............. 160 Figure 5.5. OutofPocket Spending by Provider Type .................................................... 161 Figure 5.6. Health Expenditure by Activity ....................................................................... 161 Figure 5.7. Contribution to HIV, Malaria, Reproductive Health, and Other Health Activities by Financing Source ..................................................................................... 162 Figure 5.8. Management of Health Sector Resources ....................................................... 163 Contents xi Figure 5.9. Use of Health Services in Case of Illness in the Overall Population ........... 165 Figure 5.10. Direct DiseaseRelated Spending by Income Quintile (2000­05) .............. 165 . Figure 5.11. Distance to Reach WaterAccess Point in ExProvinces ............................. 196 . Figure 5.12. Transport Transitional Structure ................................................................... 218 . Figure 5.13. Key Budget Lines by Total Development Execution 2007 .......................... 219 Figure 5.14. IsakaKigali Railway Project, "Kigali ­ Dar es Salaam Rail Corridor" ..... 223 Boxes Box 5.1. Results of Analyses on the Impact of Mutuelles on Households ...................... 166 Box 5.2. Results of the PBF Impact Evaluation .................................................................. 167 Box 5.3. Main Findings of the Benefit Incidence Analysis in the Health Sector in Rwanda (2005) ................................................................................................................ 169 Foreword by John Rwangombwa A fter the 1994 genocide, Rwanda went through an intensive reconstruction phase, rebuilding its human resource base as well as basic infrastructure, often through grassroots innovative approaches. By the end of 2003, the Government of Rwanda decided to take a new turn, expanding service delivery and infrastructure through the institutionalizing and deepening of public reforms, particularly public sector reform and decentralization. The 2003­07 years proved a turning point in building the foundations of our new state, and accelerating our progress towards our Vision 2020. This was done through the promotion of innovations and unconventional methods. During these years, a new culture of results has been built in the public sector. Fiscal decentralization has been achieved with the health and education as pilot sectors, building on strong sector strategies. The Government of Rwanda's vision of strong, autonomous providers (health, water, education) was developed. Performance contracts were tested and established at every level. The donor community was very supportive of this approach and showed flexibility. During 2003­07, there was a notable scaling up of grants and a shift from standalone projects towards budget support, our preferred aid modality. We worked closely with our development partners to improve aid effectiveness to make the Paris Declaration a reality. John Rwangombwa Ministry of Finance and Economic Planning Government of Rwanda xii Foreword by Yaw Ansu R wanda has made a remarkable transition from reconstruction to development in the past 14 years. The 1994 genocide decimated Rwanda's fragile economic base, destroyed a large share of its human capital, and eroded its ability to attract private investment. Close to 1 million people died and large numbers of people became refugees. Poverty increased dramatically, reaching 78 percent of the population in 1994 During reconstruction, the government focused on rebuilding institutions, which led to significant improvements in economic outcomes and social indicators. By 2000, the proportion of poor people had declined to 60 percent of the population. Child mortality, which had reached more than 300 per 1,000 live births after the genocide, decreased to 103 per 1,000 from 2000 to 2007, an improvement compared with pre genocide levels. Rwanda has made substantial progress on many of the Millennium Development Goal (MDG) targets. Immunization rates, at 96 percent, are among the highest in SubSaharan Africa. Use of insecticidetreated bed nets increased from 4 percent to 40 percent of the population from 2004 to 2006. Unique in Africa, Rwanda scaled up access to health insurance, leading to increased use of health services (for example, assisted deliveries) and has considerably improved access to clean water, which helped to reduce underfive mortality by 25 percent between 2000 and 2005. Early achievements were the basis for increased support from international partners, provided in the form of budget support since 2002. The government much appreciates the increased alignment of donor activities around the government's priorities that this mode of assistance has made possible. Rwanda, as one of the first three budget support harmonization and alignment pilots, now serves as a good practice example for both development partners and recipient governments. This report reviews Rwanda's budget support in the context of its overall public expenditure and resource trends, fiscal expansion, reform measures, and public spending in prioritysectors, and provides a comprehensive overview of Rwanda's progress through 2007. I hope it provides helpful information to both governments and policy makers considering budget support as a mechanism for achieving development results. Yaw Ansu Director Human Development Department Africa Region, World Bank xiii wledgm Acknow ments J authors d Prosper Musament are Kam Sayinzoga Agnes Souccat, Annika oint Kjellgren, and of this docum afiri. mpeta a, Significant contributions on specific c d chapters were also received from (in alphabetical or ingana, Allison Berg, Yisa Cl rder): Elias Bai n Desai, Pablo laver, Nikhil D a A a, Gottret, Negda Jahanshahi, Agnes Kalibata Aloys Kanam nce mugire, Lauren Lannes, Gayle Martin, C Christophe Prev vost, and Claude Sekabaraga. t c or om The report also includes contributions o comments fro (in alphabet tical order): Paulin Basinga, , Malcolm Cosg grove Davies, Fr rancois Diop, Liz Drake, Kene Ezemenari, m, Erik Fernstrom Alex Kamura aramaga, Fidele Karangwa, Jos ase, Camille Ka seph Kizito, Victoria Kwak M ne kwa, Tembo Mubaruki, Sabin Musange, Br runo Mwanafu unzi, Simon n do Ndutiye, Susan Opper, Fred Quarshie, Guid Rurangwa, Ernest Rwamu ucyo, Abebe Shimeles, Chris stian Shingiro, Nepo Rugemin Ruzindaza, and Mohamed ntwaza, Ernest R d Toure. The author rs would also lik ke to thank the supporters that t made this report possible (in alphabetica order): Bill and Melinda Gates Founda al ation, Global AAlliance for nd on, t Vaccination an Immunisatio Government of Belgium (B y Belgian Poverty Reduction Partnership Tr G f y rust Fund), Government of Japan (Policy and Human Resource n Development F Fund), and Gov vernment of N h Norway (Health Results Innov vation Trust Fund). The author are thankful for the suppor and guidance of James Mus rs rt e soni, former Minister of Fiinance and Eco onomic Plannin of Rwanda and Donald Kaberuka, ng a, he President of th African Dev velopment Ban and former Minister of Finance and nk Economic Plann ning of Rwanda a. The author rs are solely resp e content of this report. ponsible for the xiv Acronyms and Abbreviations AfDB African Development Bank BADEA Arab Bank for Economic Development in Africa BIA benefit incidence analysis BRD Rwanda Development Bank CDF Common Development Fund CEPEX Central Public Investment and External Finance Bureau CFAA Country Financial Accountability Assessment CNER National Examination Council COMESA Common Market for Eastern and Southern Africa COURS.SUP Office of the Supreme Court CP counterpart funding CPAR Country Procurement Assessment Report CPI consumer price index CPIA Country Policy and Institutional Assessment DAD Development Assistance Database DFID U.K. Department for International Development DHS Demographic and Health Survey DRC Democratic Republic of Congo EAC East African Community EC European Commission EDPRS Economic Development and Poverty Reduction Strategy EICV household living conditions survey ESSP Education Sector Strategic Plan EU European Union FARAP Financial Accountability Review and Action Plan FARG Victims of Genocide Fund FI internally financed funding FAO Food and Agriculture Organization of the United Nations FTI FastTrack Initiative GBS general budget support GDP gross domestic product GFATM Global Fund for AIDS, Tuberculosis and Malaria HIMO labor intensive public works HIPC Heavily Indebted Poor Countries HSSP Health Sector Strategic Plan ICT information, communications, and technology IDA International Development Association IMF International Monetary Fund IRC Institutional Reform Credit JAF Joint Action Forum xv xvi Acronyms and Abbreviations JBSR Joint Budget Support Review JGBS joint general budget support LABSF Local Authority Block Support Fund LTSFF LongTerm Strategy and Financial Framework MDG Millennium Development Goal MDRI Multilateral Debt Relief Initiative MIFOTRA Ministry of Public Service and Labor MIGEPROF Ministry of Gender and Family Promotion MIJESPOC Ministry of Youth, Sports, and Culture MINADEF Ministry of Defense MINAFFET Ministry of Foreign Affairs and Cooperation MINAGRI Ministry of Agriculture and Animal Resources MINALOC Ministry of Local Government, Good Governance, Community Development, and Social Affairs MINECOFIN Ministry of Finance and Economic Planning MINEDUC Ministry of Education MINICOM Ministry of Commerce, Industry, Investment Promotion, Tourism, and Cooperatives MINIJUST Ministry of Justice MININFRA Ministry of Infrastructure MININTER Ministry of Internal Security MINISANTE Ministry of Health MINISTR Ministry of Science, Technology, and Scientific Research MINITERRE Ministry of Lands, Environment, Forestry, Water, and Natural Resources MoU memorandum of understanding MSCBP Multisector CapacityBuilding Program MTEF mediumterm expenditure framework NAP National Agriculture Policy NGO nongovernmental organization NHA National Health Accounts NISR National Institute of Statistics of Rwanda NPV net present value NUR National University of Rwanda ODA official development assistance OECD Organisation for Economic Cooperation and Development OPEC Organization of the Petroleum Exporting Countries OVC orphans and vulnerable children PARLEMENT Office of Parliament PARQUET Office of the Prosecutor General GENERAL PBF performancebased financing PER public expenditure review PPCB Planification, Politiques et Renforcement des Capacites PRESIREP Office of the President PRGF Poverty Reduction Growth Facility Acronyms and Abbreviations xvii PRIMATURE Office of the Prime Minister PRS poverty reduction strategy PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper PSTA National Agricultural Policy and Sector Strategic Plan RADA Rwanda Agriculture Development Authority RARDA Rwanda Animal Resources Development Authority RFA République fédérale d'Allemagne RDB Rwanda Development Board RIEPA Rwanda Investment and Export Promotion Agency RMF Road Maintenance Fund RRA Rwanda Revenue Authority RSSP Rural Sector Support Project RwF Rwanda franc RURA Rwanda Utilities Regulatory Agency RWSSP Rural Water Supply and Sanitation Project SFAR Student Financing Agency of Rwanda Sida Swedish International Development Cooperation Agency SP social protection SPA Strategic Partnership for Africa SPTA Strategic Plan for Transformation of Agriculture SPPRMD Strategic Planning and Poverty Reduction Monitoring Department STABEX Système de Stabilisation des Recettes dExportation SWAP sectorwide approach THE total health expenditures U.K. United Kingdom UNHCR United Nations High Commissioner for Refugees USAID U.S. Agency for International Development VAT value added tax WFP World Food Programme WSS water and sanitation services Executive Summary R wanda has made a remarkable transition from reconstruction to development in the past 14 years. The 1994 genocide decimated Rwanda's fragile economic base, destroyed a large share of its human capital, and eroded its ability to attract private investment. Close to 1 million people died and large numbers of people became refugees. Poverty increased dramatically, reaching 78 percent of the population in 1994. During reconstruction, the government focused on rebuilding institutions, which led to significant improvements in economic outcomes and social indicators. As a result of extensive economic and governance reform measures between 1995 and 2005, gross domestic product (GDP) growth averaged around 7.4 percent per year in the period. By 2000, the proportion of poor people had declined to 60 percent of the population. Child mortality, which had reached more than 300 per 1,000 live births after the genocide, decreased to 103 per 1,000 from 2000 to 2007, an improvement compared with pregenocide levels. Substantial progress has been made in stabilizing and rehabilitating the economy, and domestically the country is at peace and secure. At the regional level, Rwanda plays a key role in the peace process. Though not without controversy and risk, the government engaged in several initiatives to promote regional peace and stability. It also actively pursued initiatives that support regional integration (for example, through the Common Market for Eastern and Southern Africa and the East African Community) to reduce political tensions in the region. Separately, governance indicators have improved. Rwanda's overall Country Policy and Institutional Assessment (CPIA) rating improved to 3.5 from 3.4 between 2004 and 2005. Of the 45 SubSaharan African countries rated, 62 percent scored below Rwanda and 29 percent scored above. Another ambitious program pursued by the government is decentralization aimed at further empowering local communities and increasing transparency and accountability. Additionally, Rwanda made substantial progress on many of the MDG targets. Impressive results were achieved in the social sectors: net primary school enrollment reached 95.8 percent in 2007, and completion rates increased to 51 percent in 2006 and remained stable in 2007. Immunization rates, at 96 percent, are among the highest in SubSaharan Africa. Use of insecticidetreated bed nets increased from 4 percent to more than 70 percent of the population from 2004 to 2007. HIV prevalence, at 3 percent, has decreased. Unique in Africa, Rwanda scaled up access to health insurance (through local schemes called mutuelles), from 7 percent to 75 percent of the population between 2003 and 2007, leading to increased use of health services (for example, assisted deliveries). Between 2004 and 2007, an additional 1 million people (14 percent of the population) have gained access to clean water which helped to reduce underfive mortality by 25 percent between 2000 and 2005. These achievements were the basis for increased support from international partners. Rwanda does not have a long history of receiving longterm development assistance. In fact, donor response to the 1994 crisis was hardly coordinated and aid xviii Executive Summary xix delivery was delayed. In the aftermath of the genocide, Rwanda's development partners were neither equipped nor coordinated enough to properly respond to the country's postcrisis needs. Development assistance as a share of GDP has traditionally been high. At times over seventy percent of the country's investment expenditures are financed through development grants and credits. Since the late 1990s, when the country began to gain its reputation of firmness against corruption, demonstrating through results its ability to use aid effectively, official development assistance, and the number of donors active in the country increased year after year. With increases in the amount of aid and number of donors came the burden of responding to the multitude of donor requirements. The burden of managing the flow of development assistance exhausted much of the capacity in place, leaving minimal capacity for proper implementation of government programs. Because budget support entails lower aid management transaction costs than other aid modalities, it remains the government's most favored form of aid. In 2000 the government gradually started taking the leadership role in aid coordination. The U.K. Department for International Development (DFID), the Swedish International Development Cooperation Agency (Sida), the European Commission (EC), and the World Bank first provided budget support and were joined by other donors. In more recent years and after having demonstrated allocative and operational efficiency of budget and resources, the Rwandese government encouraged other donors to provide budget support. It was in response to the government's multiple requests that a harmonization process, initially focused on budget support, was initiated in Rwanda in October 2002, with a mission of the Strategic Partnership for Africa (SPA).1 The government used the results from the SPA mission to influence donor performance. While the SPA focus was limited to budget support, it triggered discussions between donors and the government on how other aid modalities might be better aligned with the poverty reduction strategy and delivered in a more harmonized way. Much progress was made, and many harmonized processes--such as joint budget reviews, joint analytic work, joint dialogue, and so forth--became routine while the government demonstrated its commitment to responsibly using budget support resources to achieve results in priority sectors. However, challenges remain. In particular, development assistance needs to be increasingly channeled through the Ministry of Finance and Economic Planning (MINECOFIN) to allow for more effective planning and budgeting. This document in part reviews the allocative and operational efficiency of government resources during the 2004­07 time frame and aims at further strengthening the basis of the development cooperation dialogue. Rwanda experienced a fiscal expansion in the past several years measured in both nominal and real terms. Analytical work extending as far back as 2001 show a more than twofold fiscal expansion between 2001 and 2007. The largest fiscal expansion occurred between 2003 and 2004, when total expenditures were projected to reach 42 percent of GDP in nominal terms. The fiscal expansion fell short of projections, although it was still much higher in the entire period, because of a move to maintain macroeconomic stability. xx Executive Summary A recent study of fiscal policy and aid in Rwanda showed that the scaling up of aid in the country did not lead to imprudent and unsustainable fiscal policy; that the government has maintained a sound macroeconomic framework and has managed the risk associated with fiscal expansion and increased aid flow; and that while aid remains a significant and important source of resources for the country, increased budget support does not necessarily influence budget allocations away from priority sectors. The government places stronger focus on maintaining macroeconomic stability with increasing levels of development assistance. It is important to note, however, that aidfinanced expansion presents additional challenges, caused primarily by the uncertainty of aid inflows. A developing country such as Rwanda often does not have the luxury of turning down aid because the vast mismatch of its growth and development needs and its public finance capacity may be overwhelming. The country's latest strategy, the Economic Development and Poverty Reduction Strategy (EDPRS) aims to reducing aid as a percentage of GDP between 2008 and 2012. It also aims at doubling aid to support its strategic objectives. Consequently, as Rwanda seeks to further increase its external aid flow, it has aggressively tried to increase its tax base and allocate increased resources into productive sectors--with a net effect of lowering overall aid dependency. Sustainability is one of the most important aspects of aid management, and prior to debt relief, Rwanda's debt stock was high and growing. Rwanda has benefited from debt relief through both the Heavily Indebted Poor Countries (HIPC) and the Multilateral Debt Relief Initiative (MDRI) programs. The country has permanently reduced its net present value of debttoexport ratio below 150 percent--down to 56 percent in 2006--showing again responsible aid management and assuring its budget support donors that the government remains committed to its policies and strategies in the short, medium, and long terms. In the period 2004­07, overall resources increased by 16 percent in real terms. Domestic resources increased by 34 percent, driven by a 28 percent increase in tax revenue. Overall, external resources increased by a mere 2 percent: grants increased by 16 percent, but loans decreased. Tax revenues consistently represented around 90 percent of total domestic resources, showing a steady and consistent increase throughout the period. External resources showed a much more fluctuating trend, again highlighting the challenges faced by the government to manage the unpredictability of aid flow. Grants as a percentage of total external resources increased from 71 percent of actual disbursements in 2004 to 81 percent in 2007. The expansionary trend was paralleled by increases in domestic resources, which showed that the government was prudent in its fiscal expansion by ensuring that increased public expenditures were financed by more than increases in external resources alone. Recurrent expenditures represented the largest portion of total expenditures, fluctuating between 60 percent and 63 percent in the period. Overall, total expenditures increased by 24 percent in real terms, but were offset by a sharp decrease in arrears. Recurrent operating expenditures as well as subsidies and transfers increased by 26 percent and 104 percent, respectively, while interest and public debt Executive Summary xxi saw a sharp decline. Although capital expenditures increased by 65 percent in the period, net lending saw a significant decrease. Looking at key ministries, all sectors but education experienced tripledigit growth in real spending between 2004 and 2007. The ministries of education, health, agriculture, infrastructure, and land and resources increased overall (including recurrent and development spending) by 59 percent, 120 percent, 107 percent, 113 percent, and 351 percent, respectively. Education, health, and agriculture saw their recurrent spending alone increase by 39 percent, 49 percent, and 20 percent, respectively, over the period, but these five ministries experienced tripledigit growth in development spending of 240 percent, 248 percent, and 153 percent, respectively. The five ministries were allocated 67 percent of the total increase of expenditures between 2004 and 2007, with infrastructure and education benefiting the most from the public expenditure expansion in terms of additional resources as a portion of the expenditure increase. MINECOFIN and the Ministry of Defense (MINADEF) remained the largest ministries (excluding district transfers) in terms of recurrent expenditure allocations (although a portion of their budgets are benefiting other ministries/sectors), closely followed by the Ministry of Education (MINEDUC) and the Ministry of Local Government, Good Governance, Community Development, and Social Affairs (MINALOC), combined making up 64 percent of total recurrent ministry expenditures in 2007. Including district transfers, MINEDUC was the second largest ministry measured in recurrent expenditures. District transfers increased significantly in most priority areas following decentralization in 2006, and execution rates improved in the period, indicating that districts are able to cope better with increased resources and responsibilities. However, lack of capacity remained a big challenge facing the government on both central and local levels. In terms of development spending, 75 percent of the government's development budget was spent in five of the key sectors (education, health, agriculture, energy, and land and natural resources) in 2007, up from 40 percent in 2004. Donor funding remained the largest funding source for development spending: 48 percent in 2007, down slightly from 2006. Overall, donor funds and loans represented 67 percent of total development spending in 2007, down from 76 percent in 2004. The largest increase in funding sources was seen in internally financed sourcing (increasing 126 percent) , increasing from 20 percent of total funding in 2004 to 29 percent in 2007, showing the government's continued commitment to managing aid dependency. Among the five key ministries, donors' commitment in 2007 represented 90 percent of the health ministry's total development budget, while donors committed only 15 percent of the education ministry's development budget. Education received the largest portion of its funds from internally financed sources, with other sectors having more diversified sourcing. From 2004 to 2007, counterpart funding, donor funding, internally financed funding, and loans increased in real terms by 170 percent, 93 percent, 496 percent, and 75 percent, respectively, within the five line ministries. In terms of results in priority sectors, several key achievements are noteworthy. Education saw significant progress, with net primary school enrollment rates of 95.8 percent and completion rates consistently increasing to 51 percent by 2006. Both primary school and trunc commun2 is now free of charge, and no significant changes xxii Executive Summary exist between female and male students. Capitation grants were successfully introduced, and special support exists for vulnerable students. An education sector strategic plan, sector mediumterm economic framework (MTEF), organic education law, and higher education law were adopted in the last few years. An independent student financing agency was also established, alongside a national council for higher education. Challenges remain in teacher quality, high dropout rates, low completion rates, and high teachertostudent ratios. Many innovative health sectors programs were developed, including health insurance (mutuelle) with 75 percent of citizens covered in 2007; the government aims at reaching universal health coverage. Dramatic improvements were achieved in the diseases of malaria and HIV/AIDS. For malaria, the trend for both inpatient and outpatient cases decreased significantly, mainly after aggressive promotion and distribution of bed nets and inhouse malaria treatment. HIV testing services have improved and HIV prevalence has decreased since 2004. Utilization of health services also played a key role in improved health, and the introduction of performancebased contracting led to improved health service delivery. Tuberculosis morbidity decreased significantly from 2004 to 2007, despite an increase in the number of cases detected from 28,000 in 2005 to 67,000 in 2007. Immunization increased as well and in some areas reached close to 100 percent coverage. Antenatal care utilization rates reached 96 percent in 2007, while professional assistance at delivery reached 52 percent. Family planning also improved, with utilization rates increasing from 11 percent in 2006 to 27 percent in 2007. This is partially because the Ministry of Health coordinated with HIV/AIDS donors to help promote this area aggressively, as family planning and HIV/AIDS issues are highly related. Many challenges remain, however, especially in the areas of child mortality, maternal mortality, and nutrition. Progress in the area of agricultural transformation included development of an operational policy and improvements in mediumterm planning and budgeting (including a firsttime, resultsbased MTEF), a strategic plan, and piloting of fiscal decentralization. These translated into improvements in budget execution rates. However, the economy remains largely agrarian, and rapid population growth, low agriculture productivity, and high pressure on the land remain key challenges to agricultural transformation. Agricultural production is erratic, irrigation needs to be improved, and extension services remain limited. The food shortages in early 2006 highlighted the country's vulnerability to exogenous shocks and the need for productivityenhancing measures. Remaining challenges include continued agricultural practices with minimum value addition that aligns with competitive markets, and random settlement patterns that limit the opportunities for optimal land use, particularly for mechanized agriculture. The social protection sector received increased attention. In addition to advancements of the social protection strategy under the MINALOC, the sector also aggressively pursued its Vision2020 Umurenge, direct social assistance, public works, and financing support to farmers.3 Social protection is one of the largest public sectors in Rwanda, yet the sector is scattered with many uncoordinated innovations, an issue that the government is currently trying to address through its Vision 2020 Umurenge, EDPRS, and improved coordination among stakeholders. Executive Summary xxiii The government recognized that infrastructure is important to all sectors of the economy and society, and that an effective system of infrastructure and services was crucial for agricultural productivity and poverty reduction, a determinant of business investment, instrumental to human development, and the foundation for private sector development. The government recognized the need to plan for, provide, and manage physical infrastructure that is efficient, costeffective, and financially sustainable, and that supports the social and economic development priorities. Much has already been achieved in recent years. Specifically, the government successfully managed the energy crisis and also successfully uses solar, wind, methane gas, and biogas energy as power production alternatives. Access to potable water increased significantly, to 55 percent of the rural population and 69 percent in urban areas in 2005. Access to sanitation remains low, but the government is addressing the situation with an aim of reaching universal sanitation by 2020. New public expenditure reviews in all three infrastructure sectors will shed further light on progress once completed. A main primary challenge is to focus on investments in the energy and transport sectors that would support Rwanda's competitiveness. Finally, the government focused donors' attention on synergy across sectors, especially between infrastructure and other sectors (for example, export promotion policy and access to basic services, such as water, electricity, and transport). For instance, the encouraging progress made in the area of agricultural transformation will combine with an enhanced human and institutional capacitybuilding initiative. Another example is the need to formally link the export promotion strategy with the National Agricultural Policy and Sector Strategic Plan (PSTA) to ensure that it does not come at the expense of food security, but also with access to basic services (water, energy, transport). In general, infrastructure services need to be better linked with other sectors. Looking at the broader picture beyond Rwanda, despite donors' commitments to scale up aid in line with the 2002 Monterrey Consensus and the 2005 Gleneagles Declaration, the outcome has not been consistent. Official development assistance, according to Organisation for Economic Cooperation and Development (OECD) data, declined in real terms by roughly 5 percent in 2006--the first drop since 1997. In 2007, major donors provided US$103.7 billion in development assistance. Compared with 2006, development assistance decreased by 8.4 percent in real terms. Much of the difference is explained by two factors: (a) the end of debt relief for Iraq and Nigeria, which had translated into exceptionally high levels of development assistance in 2005 and 2006 (US$107 billion and US$104 billion, respectively); and (b) small increases in other categories of official development assistance. With regard to resources needed to achieve the MDGs, donors are not on track to meet their commitments to increase levels of development assistance, which according to the 2005 G8 Summit would rise from the 2004 level of US$80 billion to US$130 billion in 2010. This implies that donors would need to more than double their current rate of increase in development assistance, excluding debt relief, over the remaining years to reach the 2010 target. Moreover, official development assistance as a percent of donors' gross national income fell to 0.3 percent in 2006, after slowly edging up to 0.33 percent in 2005--still well short of the United Nations target of 0.7 percent. As of 2007, xxiv Executive Summary the only countries to exceed the United Nations target of 0.7 percent of gross national income are Denmark, Luxembourg, the Netherlands, Norway, and Sweden. Bilateral development assistance channeled to SubSaharan Africa, excluding debt relief, increased by 10 percent in 2007. In Rwanda, development assistance increased as well in 2007, although at a lower rate. However, donors have yet to meet the rates of growth in development assistance that would support the Gleneagles G8 Summit intention to double their assistance to Africa by 2010. Nevertheless, Rwanda remains highly aid dependent and is prone to fiscal risks associated with aid dependency, for example, vulnerability to exogenous shocks and debt sustainability. Consequently, aid predictability and in particular budget support predictability are important conditions for the planning and uninterrupted implementation of development programs. In the case of budget support, the government adjusts to unexpected underdisbursements through higher domestic financing and reductions in domestic investment spending. To make matters worse, these adjustment needs are sometimes accompanied by tax revenue shortfalls and current expenditure overruns. To smoothly absorb unexpected variations in external financing, the government designed and implemented several flexible and innovative programs that are readily adjustable in terms of their scale (for example, capitation grants in the education sector, or the Common Development Fund) to react to funding variations. Such flexible programs need to be developed in more sectors to better mitigate fiscal risks. The government much appreciates and continues to benefit from the increased alignment of donor activities around the government's priorities and its priority sectors. Rwanda, as one of the first three budget support harmonization and alignment pilots, now serves as a goodpractice example for both development partners and recipient governments. In light of the fact that (a) the government of Rwanda demonstrated allocative efficiency of national and donor resources in the form of budget support; (b) substantive progress was made in terms of results, especially in priority sectors; and (c) Rwanda is at a disadvantage in terms of increases in aid compared with the average increase of 10 percent for Africa and compared with average commitments made by major donors at 2005 G8 meetings, the government of Rwanda welcomes enhanced collaboration and additional resources from its development partners in its efforts to address outstanding challenges. Within the existing framework of partnership and mutual accountability, discretionary resources in the form of budget support would better contribute to enhancing aid effectiveness, as this modality not only represents the lowest transaction costs in terms of aid management, but also enhances collaboration and mutual benefit from increased collaboration. Additionally, to allow for better programming and budgeting, and to reduce unnecessary duplications, the government iterates its preference for "on budget" development assistance, as opposed to resources channeled through nongovernmental entities. Rwanda's high dependence on development resources increases its vulnerability to exogenous shock. Low intraannual as well as multiyear predictability of donor resources only exacerbates this vulnerability in addition to hindering the effectiveness of any mediumterm planning and budgeting process. Executive Summary xxv Notes 1 The SPA had identified three initial countries--Ethiopia, Rwanda, and Senegal--as pilots for budget support harmonization. 2 Trunc commun is Rwandas lower general secondary school. It takes three years to complete; these are the last three years of a "nine year basic education," after the first six years of primary. 3 The government currently envisages VUP funding for public works, direct supports, and credit packages in the ratio 50:20:30. Financial services are not yet operational. VUP complements the delivery of basic health and education services, which is now a responsibility of local government. In addition to the original 30 VUP pilot sectors, 30 additional pilots were added in July 2009 as part of scaling up the VUP, leading to pilots that are being implemented in 60 Sectors with plans to scale up further by 2011. Local governments are playing an expanding role in coordinating the delivery of various services. Communitybased participatory processes continue to play important roles in the identification of VUP beneficiaries and in the selection and implementation of local public works pilots that are currently being implemented to varying degrees in all original pilot Sectors. CHAPTER 1 Introduction T he overall objective of this comprehensive report is to consider Rwanda's budget support in the context of its overall public expenditure and resources to (a) provide an overview of Rwanda's experience with budget support, reform measures, and its progress of budget harmonization, (b) provide the first comprehensive assessment of all of Rwanda's overall public expenditures and resources between 2004 and 2007, and (c) provide the first summary of public expenditure reviews and related analytical work undertaken in priority sectors, covering varying periods between 2000 and 2007. Following this introductory chapter, chapter 2 reviews (a) general budget support relevance, rationale, and outstanding challenges in the context of Rwanda by providing a historical background of budget support; (b) Rwanda's progress in budget support­ related processes and practices; (c) economic and structural reforms to date; and (d) budget support predictability trends. Chapter 3 then assesses the net resources available to the government of Rwanda and how these resources were spent. In this chapter, resources are broken down by domestic revenue (tax revenue, nontax revenue, and other sources), external funding (grants and loans), and other financial resources; expenses are broken down by recurrent expenditures (operational expenditures, interest and commission, reimbursement of public debt, and subsidies and recurrent transfers), capital expenditures and net lending, and arrears. Chapter 4 follows with a detailed review of resource allocations and spending among the government's ministries, including its transfers to districts. Public expenditures are broken down according to the structure of the Organic Budget Law, considering recurrent and development spending by ministry and economic classifications. Chapter 5 reviews all sectors--not only ministerial expenditures, but also other sectorrelated spending across ministries and other expenditures that contribute to a sector but are not part of centralgovernment spending. This sector analysis is primarily based on indepth studies (mainly public expenditure reviews [PERs]) undertaken in agriculture, education, health, social protection, and infrastructure (water and sanitation, energy, and transport) sectors. Chapter 5 gives specific emphasis to these socalled priority sectors and is divided into three main areas: policy and sector overview, resource and spending allocation, and results and challenges. Under infrastructure, water and sanitation, energy and transport receive special attention. While no PER has yet been conducted in the energy and transport subsectors, a snapshot of these subsectors is provided in this chapter. And while the social protection sector, with many programs cutting across multiple ministries including 1 2 World Bank Working Paper MINALOC, is not considered in the overall prioritysector analysis (chapters 3 and 4), it is reviewed in detail in chapter 5. The objective of these sectorspecific analytical summaries (primarily PERs) is to ensure that results of the overall indepth analysis that was conducted on behalf of these key ministries are consolidated to (a) provide an enhanced understanding of the country's overall public expenditure, (b) help put each independent sectorspecific analysis into the context of the overall budget allocation considerations, and (c) enhance the overall prioritysector analysis as presented in this report in chapters 3 and 4. Some discrepancies between PER report data, other sectorspecific analytical work, and the overall data in other chapters of this report can often be explained and, to the extent possible, are noted in endnotes or in the overall text. However, at times underlying assumptions in the analytical approaches provide differences in the data. For sectorspecific differences, this report refers both to the extensive appendixes in this report (Excel spreadsheets that provide the basis for this entire report) and the source documentations used (which can be requested from the specific ministries). Furthermore, it is important to note that the classifications here of priority sectors vary from the priority sectors often listed in International Monetary Fund (IMF) and World Bank Poverty Reduction and Economic Management data. Chapter 5 of this report provides a snapshot of all other sectors between 2004 and 2007. These sectors are regulated by a line ministry but are not considered priority sectors and/or have not been covered by a PER or other analytical work. This chapter allows a quick overview of all sectors for 2004­07 recurrent and development spending. Chapter 6 summarizes the report, addresses outstanding challenges, and offers concluding remarks. Regarding the use of data in this report, the following should be noted: In 2004, IMF and budget data were greatly harmonized. In similar analyses done in the past, covering periods prior to the 2004 harmonization effort, two different data sets needed to be considered. However, as these two data sets have been mostly harmonized, the differences are no longer significant (Sayinzoga 2007). Consequently, and consistent with the entire analysis as presented in this report, budget data by the Ministry of Finance and Economic Planning (MINECOFIN) is the main source for analysis unless otherwise indicated. However, much analytical work looks at budgeted data, either original or revised, and expenditures committed in a fiscal year may not have been paid out by the end of the fiscal year (MINECOFIN 2006). Therefore, this report puts a heavy emphasis on actual data as reported by MINECOFIN by the first or second quarter of the following fiscal year. This report relies on both original and revised budget data in addition to actual data to not only look at trends in actual or projected allocations, but to gain better understanding of discrepancies in budget allocations and related execution rates. This approach is used for the overall data sets and across all data in this analysis (to the extent available). Therefore, original, revised, and actual data are reported for ministry recurrent and development budgets by program and project, including detailed recurrent execution rates for all programs listed under centrallevel district and province budgets. In this context, it should also be noted that although reporting between MINECOFIN and IMF data has been harmonized since 2004, challenges in analyzing Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 3 trends still remain. Reporting in various ministries has changed over the period covered in this report. While 2004 and 2005 data were reported in a fairly consistent manner, 2005/2006 data differ significantly primarily because full decentralization went into effect in 2006. These changes are further complicated by ministerialspecific changes, such as the structural changes taking place within the Ministry of Agriculture and Animal Resources (MINAGRI) in 2005/2006. Additional changes occurred between 2006 and 2007 and further changes are expected. It is also significant that the way MINECOFIN reported data in the past has not been conducive to a thorough analysis as most data were presented in hard copies and in PDF files. As of 2006, MINECOFIN began reporting data electronically in Excel, but more than 98 percent of the data in the data sets of this report consolidate data not previously available in the Excel format. Finally, to the extent possible, data presented in this report were converted into real term 2007 prices. The structure of Rwanda's Organic Budget Law provides the basis for the structure in this analysis. The Organic Budget Law provides several summary tables in its earliest sections, and it is the summary tables for resources and public expenditures in the Organic Budget Law that provide the structure of the analysis in chapter 3. Further details about this chapter, data broken down below subsector levels, are provided for all years in Data Sets 19­23, available on request (see appendix F of this report). In chapter 4, centrallevel recurrent expenditures are listed by ministry, broken down by program, subprogram, and detailed expenses. Details about each ministry's programs can be found in Data Sets 2­5. While trends in recurrent transfers to districts are discussed in chapter 4, further details of programs, subprograms, and expenses by district and province can in varying degrees be found in chapter 5. Data Sets 69 provide further details of program expenditures at decentralized levels. For purposes of this report, these district and province programs are allocated back to assumed related ministries to gain a full sector spending overview. Assumptions in this approach may vary from specific assumptions made in PER reports and other sector specific analytical work, but most of these differences are considered minimal. Whenever further explanation of assumptions made in sectorspecific work is needed, this report refers to the original analytical work and the reference list. Differences between data in chapters 4 and 5 can also partially be explained by the fact that a sector analysis considers expenditures beyond just ministry spending. The details of district and provincelevel programs are found in Data Sets 6­9. The last portion of the Organic Budget Law lists Rwanda's development budget, by ministry, from program down to each project expense. No development budget for districts exists under this structure for the period covered, and at times recurrent expenditures can be found under the development budget. A prime example of this is the salaries for higher education, which are listed as development expenditures under higher education programs in the development budget of the Ministry of Education (MINEDUC). Details about each ministry's development budget are found in Data Sets 10­18. In 2006, the Central Public Investment and External Finance Bureau (CEPEX) released its first annual, comprehensive development budget report showing execution rates for each project listed. This new way of reporting provided a significantly improved source of data for development budget execution. However, there are 4 World Bank Working Paper several key challenges with this reporting. First, data collected for the CEPEX report are not consistent with the budget allocations, at times reflecting additional sources not going through the budget (such as direct transfers from overseas to the project), hence, showing significant overspending in the development budget. Second, CEPEX claims to also include offbudget data, yet this report finds that only three small projects in 2006 and none in 2007 were offbudget. Consequently, a task force was established in April 2008 following the Joint Budget Support Review to address this issue. In light of these issues and for the purposes of this report, actual development spending analysis by ministry in chapters 3 and 4 uses CEPEX executed data for 2006 and the revised development budget for 2007 (unless otherwise indicated). Details of development budget allocations and CEPEXreported spending can be found in Data Sets 10­18. The data sets in this report are available in the form of Excel spreadsheets (all in one comprehensive Excel file). Because these Excel data sets are so large--covering the entire country's recurrent, development, and district centrallevel budgets for four years (2004­2007), including original, revised, and executed data--these data sets are not included in this document but can be requested as a separate file. In addition to the 23 data sets referenced in this report, the Excel file also includes an additional 16 data sets covering relevant data considered in this analysis. All tables and graphs illustrated in this report, all of which are linked to the sources in the data sets, are provided in a separate file. The main objective of this approach is to have one consolidated source of the country's expenditures for the years 2004­2007, a source of data than can then be further extended and updated in the years beyond 2007. Figure 1.1 below illustrates the underlying structure of all data in this report, as presented in the data sets. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 5 Figure 1.1. Overview of the Analytical Structure That Provides the Basis for This Analysis Source: Annika Kjellgren. CHAPTER 2 General Budget Support (GBS): Relevance, Rationale, and Remaining Challenges Historical Background and Trends in Harmonized GBS Rwanda has made a remarkable transition from reconstruction to development in the past 14 years. The 1994 genocide decimated Rwanda's fragile economic base, destroyed a large share of its human capital, and eroded its ability to attract private investment. Close to 1 million people died and large numbers of people became refugees. Poverty increased dramatically, particularly among women, reaching 78 percent of the population in 1994. During reconstruction, the government focused on rebuilding institutions, which led to significant improvements in economic outcomes and social indicators. As a result of extensive economic and governance reform measures taken between 1995 and 2005, gross domestic product (GDP) growth averaged around 7.4 percent a year. By 2000, the proportion of poor people had declined to 60 percent of the population. Child mortality, which had reached more than 300 per 1,000 live births after the genocide, decreased to 103 per 1,000 from 2000 to 2007, an improvement compared with pregenocide levels. Substantial progress was made in stabilizing and rehabilitating the economy. By 1998, GDP had recovered to its pre1994 level, with post conflict reconstruction fueling an initial boom. As the economy moved into the development phase, more recent growth, from 2000 to 2005, averaged 5.3 percent per year. Inflation remains low, at 6­ 6.5 percent per year on average since 1995. Macroeconomic management has been satisfactory, as evidenced by the successful completion of six reviews under the IMF's Poverty Reduction Growth Facility (PRGF) and a threeyear PRGF arrangement approved in 2006. Strong implementation of macroeconomic policies enabled Rwanda to reach the Heavily Indebted Poor Countries (HIPC) Completion Point and qualify for the Multilateral Debt Relief Initiative (MDRI). Domestically, the country is at peace and secure. While the legacy of the genocide persists, the country has made good progress toward resettlement, reconciliation, demobilization, and reintegration of excombatants. About 15,000 members of the former government army were integrated into the Rwandese National Army at various command levels, and about 3.5 million Rwandese refugees were repatriated and resettled (an impressive record compared with other countries). Through the Genocide 6 Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 7 Survivors Fund, the government of Rwanda provides support in education, health, shelter, and incomegenerating activities to the most vulnerable of survivors (for example, widows and orphans). A Unity and Reconciliation Commission, established to consolidate the government's policy in redressing the legacy of divisive politics, continues to raise public awareness through Ingando, civic education initiatives, and there has been extensive dialogue on unity, reconciliation, justice, security, and democratization. Rwanda accelerated Gacaca,1 communitybased legal hearings of those accused of genocide, with jurisdictions expanded to cover the whole country in 2005. Rwanda played a key role in the peace process in the Great Lakes region. Although not without controversy and risks, the government engaged in several initiatives to promote regional peace and stability. The Tripartite Plus One Commission, composed of Rwanda, the Democratic Republic of Congo (DRC), and Uganda, with Burundi as an observer, established an agreement on the prosecution of criminals and "negative forces" in the region. Rwanda and DRC, in cooperation with the UN Mission in DRC, increased pressure on the Forces Démocratiques de Liberation du Rwanda, the Rwandese rebel group in eastern Congo, to honor its commitment to end military activities and return to Rwanda. The government also actively pursued initiatives that support regional integration (for example, through the Common Market for Eastern and Southern Africa [COMESA] and the East African Community) to reduce political tensions in the region. Governance indicators have improved. Rwanda's overall Country Policy and Institutional Assessment (CPIA) rating improved to 3.5 from 3.4 between 2004 and 2005. Of the 45 SubSaharan African countries rated (including Rwanda), 62 percent scored below Rwanda, and 29 percent scored above. According to the Governance Research Indicator Country Snapshot indexes of corruption compiled by the World Bank Institute, between 1998 and 2004 Rwanda improved its record on all six indicators, namely, voice and accountability, political stability, government effectiveness, regulatory quality, rule of law, and control of corruption.2 Compared with the average for SubSaharan Africa, however, Rwanda ranked relatively low on voice and accountability, political stability, and the rule of law, due largely to the previous years of civil conflict. The government worked with bilateral donors to create more open political dialogue. The government also pursued an ambitious decentralization program to empower local communities and increase transparency and accountability. Rwanda made substantial progress on many of the Millennium Development Goal (MDG) targets. Impressive results were achieved in the social sectors: net primary school enrollment reached 95.8 percent in 2007, and completion rates increased to 51 percent in 2006 and remained stable in 2007. Immunization rates, at 96 percent, are among the highest in SubSaharan Africa. Use of insecticidetreated bed nets increased from 4 percent to more than 70 percent of the population from 2004 to 2007. HIV prevalence, at 3 percent, decreased. Unique in Africa, Rwanda scaled up access to health insurance (through local schemes called mutuelles) from 7 percent to 75 percent of the population between 2003 and 2007, leading to increased use of health services (for example, assisted deliveries). From 2004 to 2007, an additional 1 million people (14 percent of the population) gained access to clean water, which helped to reduce under five mortality by 25 percent from 2000 to 2005. Selected economic and social indicators are shown in table 2.1. 8 World Bank Working Paper Table 2.1. Selected Economic and Social Indicators Economic Indicators 1990 1994 2005 GDP per capita (US$) 251 152 258 GDP growth (%) ­2 ­50 5.0 Inflation (CPI, %) 4 1.6 (1996­1998) 9.2 Total external debt (% of GDP) 28 127 73 Fiscal deficit (% of GDP, including grants) ­5 ­11 ­1.6 FDI net inflows (US$ million) 8.0 0 3.6 (2004) Population (millions) 7.1 5.5 9.0 Poverty and Social Indicators 1990 1994 Most recent year Poverty incidence (% under $1 per day) 40 78 56.5 Gross primary school enrollment (%) 67 n.a 92 Under-five child mortality (per 1,000 live births) 150 209 (1995) 103 Immunization rate (measles, %) 83 25 96 HIV/AIDS prevalence (%) -- 13 (1997) 3 Source: MINECOFIN macroeconomic tables (2007) and MINECOFIN (200607). n.a. = Not applicable. The country does not have a long history of receiving longterm development assistance. Most assistance until 1997 was humanitarian aid. The country's major donors during the 1990s were Germany, the United Kingdom, the United States, the Netherlands, Belgium, the European Union (EU), the World Bank, the IMF, and the African Development Bank (AfDB). Several years before the 1994 genocide, the international community provided substantial amounts of assistance to Rwanda, contributing to efforts to find a peaceful solution to the many years of conflict in the country. But the conflict eventually culminated in the genocide of 1994, in which during a threemonth period, close to 1 million people were killed. The Rwanda crisis was characterized by rapid and unpredictable changes, including massive population displacements, a breakdown of legal and regulatory institutions and governance, and the collapse of public sector service delivery systems. For example, more than 80 percent of health sector personnel had either fled or been killed. The brutal nature and the extent of the massacre swiftly and profoundly changed the country's socioeconomic foundation. The donor response to the crisis was hardly coordinated, and aid delivery was delayed. Despite the lack of coordinated political, military, or humanitarian strategy or framework, the donor community allocated substantial resources for rehabilitation. Financial support by the international community in response to the genocide was approximately US$1.4 billion in 1994, 85 percent of which came from official donors. Overall, half of donor resources were channeled through the United Nations High Commissioner for Refugees (UNHCR), the World Food Programme, and a few other UN agencies, which in turn disbursed the majority of the funds through implementing nongovernmental organizations (NGOs). EU and U.S. contributions alone represented more than 50 percent of total resources. Nevertheless, even during the period immediately following the 1994 crisis and despite the inkind nature3 of most of the Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 9 emergency and humanitarian relief assistance, there were considerable delays in the delivery of donor assistance. In early 1995 donors made significant pledges at a roundtable of the Rwanda government and donors. The amount of assistance requested by the government, the amounts pledged by donors at the roundtable, the amounts committed, and the amounts disbursed in 1995 are presented in table 2.2. In the aftermath of the genocide, Rwanda's development partners were neither equipped nor coordinated enough to properly respond to the country's postcrisis needs. The commitmentdisbursement gap persisted and is reflected in the postcrisis aid data. The country needed fastdisbursing postcrisis mechanisms circumventing process barriers and avoiding the burden of multiple donor procedures that slow down disbursements. Table 2.2. Development Assistance 1995 Roundtable Conference (US$ millions) Requested Pledged Committed Disbursed Financial support 189.6 186.2 111.2 50.1 Repatriation and reintegration 273.7 65.6 42.7 25.5 Rehabilitation and reconstruction 300.9 314.2 284.5 94.1 Outside the roundtable process and 0.0 141.3 84.6 75.3 unallocated Total 764.2 707.3 523.1 245.1 Source: JEEAR 1996. Development assistance as a share of GDP has traditionally been high in Rwanda. As a result of conflicts and the 1994 crisis, however, the aid resources became for the most part high shares of emergency and humanitarian assistance from an increasing number of donors as well as from international NGOs and faithbased organizations. Since 1998, the country and its development partners have transitioned from short term emergency and humanitarian relief aid to longerterm rehabilitation and actual development assistance. Rwanda remains highly dependent on development assistance. At times over seventy percent of the country's investment expenditures are being financed through development grants and credits. In the late 1990s, Rwanda began to gain a reputation for its firmness against corruption, demonstrating through results its ability to use aid effectively. Official development assistance (ODA) commitments increased year after year, as did the number of donors active in the country. With increases in the amount of aid and number of donors came the burden of responding to the multitude of donor requirements. The burden of managing the flow of development assistance exhausted much of the capacity in place, leaving little capacity for proper implementation of government programs. Budget support, among other desirable features, entails lower aid management transaction costs than other aid modalities, and it has been favored by the government. In 2000 the government gradually started taking the leadership role in aid coordination. Since then, first the U.K. Department for International Development (DFID) in 2000, and then the Swedish International Development Cooperation Agency (Sida) in 2001, and later in 2003 the European Commission (EC), which was providing 10 World Bank Working Paper targeted budget support until then, introduced the first nontargeted budget support intervention. In more recent years, after having demonstrated allocative and operational efficiency of government budget and donor budget support resources, the government was able to encourage other donors to provide assistance in the form of sectoral or general budget support (GBS). It was in response to the government's multiple requests that a harmonization process, initially focused on GBS, was initiated in Rwanda. In October 2002, one month ahead of the governmentdonor roundtable meeting, a first mission of the Strategic Partnership for Africa (SPA)4 took place to consider whether, and how, donor assistance might be harmonized to better support the implementation of Rwanda's poverty reduction strategy.5 The SPA mission findings are well summarized by its recommendations: Budget support donors and development institutions should agree to streamline conditionality within a maximum framework drawn from the poverty reduction strategy matrix of policy and performance indicators. Any requirements external to the framework should be limited to issues of political governance and fiduciary risk, which should be drawn, where possible, from government policy statements. The number of such requirements should be kept to a minimum and be agreed upon by other donors. All donors and development institutions should agree to align their review cycles with the poverty reduction strategy implementation progress reviews. All donors should provide relevant and timely information on commitments and disbursements to increase predictability and transparency of development assistance. The government used the results from this first SPA mission to influence donor performance. While the SPA focus was limited to budget support, it triggered discussions between donors and government on how other aid modalities might be better aligned with the poverty reduction strategy and delivered in a more harmonized way. The government was already expressing its strong preference for development assistance in the form of GBS and argued that despite the costly set of parallel procedures pertaining to budget support (at that time),6 the burden of their management was significantly less than the costs associated with other aid modalities. Recognizing at the time that some donors were reluctant or unable to provide GBS, the government indicated its preference for joint donor arrangements as the next best alternative for reducing the burden of transactions costs. Since 2002, much progress has been made, and many harmonized processes such as joint budget reviews, joint analytic work, joint dialogue, and so forth have become routine. The government has demonstrated its commitment to responsibly using GBS resources toward achievement of results in priority sectors, and the amounts and share of GBS in total development assistance have increased--as have the number of GBS donors. However, the available data highlight a remaining challenge, namely, that all official assistance be channeled thought MINECOFIN. The discrepancy between aid amounts managed by MINECOFIN and those reported by donors as reflected in the Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 11 ODA data of the Organisation for Economic Cooperation and Development (OECD), shown in figure 2.1, provide some indication of the large amounts of offbudget ODA in Rwanda. Rwanda remains heavily aid dependent (see figures 2.2 and 2.3). This is further developed in chapter 3. Figure 2.1. Trends in General Budget Support (in US$ millions) 450 400 350 In million US$ 300 250 200 150 100 50 0 2000 2001 2002 2003 2004 2005 2006 2007 Budget support Source: 2000­03 approximate values in million US$, based on "Budget Support Trend" presentation by Secretary General at donor support retreat (Rwangombwa 2005); MINECOFIN (200407). Figure 2.2. Share of Current Grants (including GBS) as a Percentage of Total Budget Support 100 90 80 70 60 Percent 50 40 30 20 10 0 2004 2005 2006 2007 Total Grant (incl. GBS) Total BS (Current, Capital, Loans) Source: MINECOFIN (2004­07). 12 World Bank Working Paper Figure 2.3. Share of Total Budget Support as a Percentage of Total Official Development Assistance 100 90 80 70 60 Percent 50 40 30 20 10 0 2000 2001 2002 2003 2004 2005 2006 BS Non-BS Sources: MINECOFIN (2004­07); OECD 20042006. Progress in GBS-Related Processes and Practices Since Vision 2020 and the adoption of the Poverty Reduction Strategy Paper (PRSP) as the central expression of Rwanda's own development strategies, the government budget has received increasing attention as the main vehicle for the implementation of the poverty reduction strategy (PRS). Several multilateral and bilateral donors saw an opportunity to support PRS implementation through systematic and more predictable budget financing. Their objective was to strengthen country ownership and sustainability of the development process by reinforcing the country's own policy design, budget, and planning processes rather than setting up parallel systems through donorfinanced projects. As a result of the increasing share of development financing in the form of GBS, (a) the government was able to mobilize substantial technical assistance for making improvements of its budgetary and planning systems, as well as for capacity building; (b) aid in the form of GBS has brought improvements in donor assistance strategy alignment with government priorities, as well as alignment of donor practices among themselves or, whenever possible, with those of the government; and (c) the process of increased share of aid in the form of harmonized GBS has reduced the burden of managing development assistance through better harmonization and alignment across all aid modalities. Development assistance in the form of GBS has come with explicit requirements: description of participatory processes and analytical work, poverty and social impacts analysis, evaluation of fiduciary systems, and assessment of evaluation and monitoring systems. Aid in the form of GBS goes beyond addressing the government's financial needs, as it comes with muchappreciated technical assistance and capacitybuilding activities necessary to (a) better and more systematically identify gaps in country processes and systems; and (b) support their improvement in a mediumterm setting, beyond the fiscal year concerned by firm annual commitments. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 13 The different conditions attached to GBS range from the completion of key policy actions (multilateral banks) to outcome indicators (EC variable tranches) and completion of IMF reviews (bilateral donors and EC fixed tranches). GBS donors have mostly focused on mechanisms that aim at ensuring the coherence of disbursement conditionality, the timing of disbursements, and evaluation of performance in terms of poverty reduction strategy implementation. By comparison, while much progress has also been achieved in the area of technical assistance, its coordination and harmonization have not received the same level of attention. It is a given that a sufficiently transparent budget execution framework is a requisite for the national budget to be an effective instrument for implementing development policies, and that the key elements of such a framework are a transparent budget preparation process, sound accounting for budgetary spending, relatively transparent procurement practices, and operational internal control mechanisms. Additionally, the government needs to prepare a satisfactory macroeconomic framework and revenue forecasts; link budget proposals with mediumterm objectives, especially in priority sectors; and bring together spending and financing in the medium term. The preparation of a mediumterm expenditure framework creates challenges in terms of coordination between the finance ministry and planning departments in line ministries. Nevertheless, the effectiveness of development assistance in the form of GBS depends on the links between budget allocations, goals, and associated financing needs--an operating mediumterm economic framework (MTEF) process. As a result, GBS has attracted a variety of technical assistance aimed at improving technical systems, institutions, and budget planning. In most cases, financing of technical assistance and capacity building activities responded to needs expressed by the government (for example, hiring consultants to assist with the reform of existing systems); in other cases, technical support was embedded in parallel donor projects and programs. The preparation of budget support operations also gives rise to technical support for specific actions that are included in donor conditionality or disbursement triggers and draw directly on expertise of certain donors (for example, preparation of resultsbased MTEFs in the health, education, water, energy, and agriculture sectors). For instance, donors are increasingly aware of the benefits of joint analytic work. The Financial Accountability Review and Action Plan (FARAP) carried out jointly by the government, DFID, and EU was initiated as part of the public expenditure management review process supported by DFID, EU, World Bank, and other donors. While a full Country Financial Accountability Assessment (CFAA) is one of the required underpinnings of World Bank Poverty Reduction Support Credits (PRSCs), the World Bank decided to rely on the FARAP supplemented by other available diagnostic work. Such practices alleviate some of the costs associated with managing development assistance. While recognizing the benefits of joint analytic and diagnostic work, donors also recognize that relying on one donor is sometimes more efficient for all interested parties, provided there is sufficient donor consultation throughout the process. This is the case for the periodic Country Procurement Assessment Reports (CPARs), where donors, instead of mobilizing resources to carry out a joint CPAR, preferred to rely on the World Bank's comparative advantage in this 14 World Bank Working Paper area, on donor consultations during the assessment process, and on sharing final results. Separately, much progress was made in harmonizing capacitybuilding initiatives. The government developed the Multisector CapacityBuilding Program (MSCBP), a national program to build capacity in public, private, and civil society sectors. The program outlined a common framework and identifies capacitybuilding needs with the aim of pooling donor resources. To further facilitate harmonization of capacity building assistance, the government established an autonomous body, the Human Resource and Institutional Development Agency, to coordinate, harmonize, monitor, and evaluate all capacitybuilding efforts in the country under the MSCBP. Separately, DFID carried out a review of donorsupported capacitybuilding activities with MINECOFIN to propose a harmonized approach at the ministry level. Despite these advancements in technical cooperation and capacity building, there appears to be room for further improvement in joint planning of assistance, which could be less fragmented; lowering transaction costs; and more continuous implementation of reforms when only parts of the reform program are funded. Even more effective strengthening of technical and capacitybuilding needs remains a challenge for reaping the potential benefits of joint budget financing. Technical expertise in preparation of the resultsbased MTEF and associated sectoral MTEFs, budget planning (given the resource predictability concerns inherent to the country's current status of being a highly aid dependent), accounting, and auditing will need to be provided in a mediumterm setting to improve country systems and institutions. Only then will the government fully reap the results of lower transaction costs and more timely disbursements. Steady and coordinated technical support as Rwanda's institutions are being strengthened is key to obtaining results under joint budget financing frameworks. Most important, the sequential features of effective budgeting and the alignment of the budget with mediumterm objectives remain a constant struggle in a context of high aid dependency and lax predictability. Alignment of technical support, in terms of both substance and timing, with the actual needs of the government, and the planning of technical assistance and capacitybuilding, preferably during annual PRS reviews, remain crucial for results. As mentioned above, the increased share of aid in the form of harmonized GBS reduced the burden of managing development assistance through better harmonization and alignment across all aid modalities. For example, sectorwide approaches (SWAPs) with harmonized features emerged in the education, health, water, and agriculture sectors. Rwanda's education sector SWAP is the most advanced, and its donor group is led by DFID. The government prefers common aiddelivery approaches at the sector level, relying to the extent possible on national processes and systems--that is, gradually shifting away from a fragmented project approach through the use of common systems for negotiating, planning, implementing, and monitoring-- as this transition reduces aid transaction costs and facilitates the transition toward GBS. SWAPs remain most advanced in the education and health sectors. The Education Sector Donors Group has a memorandum of understanding (MoU) from April 2006 that included pooled funding for the government's education sector capacitybuilding plan. The MoU set out the agreement between development partners and the Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 15 government to support the education sector regardless of the aid modality. The MoU committed to aligned and harmonized planning and policy dialogue, consultation and information sharing, performance monitoring, and pooled support for capacity building. Other examples of progress include the following: A governmentdonor non­budget support harmonization body, the Harmonization and Alignment in Rwanda for Projects and Programs, was created, and the United Nations Development Programme hosts its secretariat. DFID and the World Bank initiated a dialogue on working toward a joint/coordinated country assistance strategy. There remains a proliferation of donors' assistance strategies, each of which requires design, negotiation, implementation, and evaluation missions, which taxes the administrative capacity of the government. Donors are increasingly aware of the benefits of joint or shared analytic work. The FARAP mentioned above is one example of joint analytic work, and appendix B outlines other recent examples. A good example of shared analytic work is the 2004 CPAR, where instead of mobilizing resources to carry out a joint CPAR, donors preferred to rely on the World Bank's comparative advantage in this area, on donor consultations during the assessment process, and on sharing of final results. Delegated cooperation between DFID and Sida in the education sector, with Sida as a silent partner, was a good practice example, and new delegated cooperation initiatives were considered among bilateral donors, for example, the Netherlands and DFID, with the Netherlands as a silent partner. Joint reviews are now common practice in the education and health sectors, where donor coordination is strong. Joint reviews were facilitated by the Harmonization and Alignment Calendar finalized and adhered to during PRSC1 preparation. Donors looked into ways of reducing transaction costs associated with their procurement practices. For example, the World Bank­financed Multisector HIV/AIDS project and the Global AIDS Fund used the same project implementation unit. The World Bank project harmonized its antiretroviral procurement procedures with those of the National Tender Board and the Global AIDS Fund unit. Economic and Structural Reforms 2004­2007 Supported by GBS Budgeting and Planning (including MTEF) The main objectives of this chapter are to broadly outline the budget process and some of the budgeting challenges; to elaborate on how a mediumterm expenditure framework (MTEF) addresses several budgetary challenges; and to describe the links between the PRS, SWAPs, and the MTEF. The longterm goals of Rwanda are elaborated in Vision 2020 and in turn translated into a periodic PRS that outlines mediumterm priorities and related actions. The PRSP guides sector strategies aimed at achieving sectoral outcomes. The implementation of PRSP priorities are elaborated in resultsoriented sector strategies, 16 World Bank Working Paper costed in sectoral MTEFs, and carried over to the annual budget at the sector and local levels. The budget is the government's most important economic and financial planning tool. It covers government revenue and expenditure for the fiscal year, and it embodies the government's development strategy. Budgets have several objectives, including ensuring the balance between aggregate revenue and expenditure, allocating resources in line with priorities, and promoting efficiency of expenditure through a resultsbased approach. The failure to link policy, plans, and budgets is the single most important cause of poor budgeting. A budget elaboration exercise based largely on developing a wish list of needs often disregards policy and planning priorities, undermining the government's objective to deliver quality services when and where they are most needed. To establish these policybudget links, budgets need to take into account macroeconomic realities, projections of revenue (internal and external), public spending (recurrent and development), policies, and longerterm national priorities. Policy planning also needs to recognize constraints and implementation capacity. Broadly speaking the budget cycle has four components: (a) formulation by the executive; (b) enactment by the legislature, whereby the budget plan is discussed, adjusted, and approved; (c) implementation; and (d) expost monitoring of actual expenditure. However, reviews of the PRSP experience across countries have demonstrated that public expenditure management is a key constraint to the implementation of the PRS. Good public expenditure management provides a resource framework to guide formulation and implementation of PRSPs. The MTEF attempts to ensure that expenditures are driven by policy priorities and conditioned by budget realities. The MTEF sets budgets for a threeyear rolling period, which is revised annually, and consists of reconciling bottomup estimates of the implementation cost of government resultsoriented policies with topdown estimates of the resource envelope. The overall MTEF is, to the extent possible, complemented by resultsbased sector MTEFs (that is, education, health, water, agriculture, and energy). The MTEF provides a mediumterm (threeyear) perspective of government and donor resources, thus reducing duplication of donor spending and increasing efficient use of available resources. It sets budget ceilings for sectors and helps shift spending toward policy priorities by planning expenditures over the threeyear period. It promotes a resultsbased approach because resources are allocated to "outputs" rather than to budget line items. The MTEF could improve predictability of resource flows as expenditures are planned over a longer period. The MTEF bridges the PRSP and the budget largely by strategically allocating resources for PRSP implementation. The MTEF also provides the budgetary link between a PRS and SWAPs. Sector MTEFs also serve as frameworks for financing SWAPs (currently in education, health, water, and agriculture). For instance, the health sector program is integrated within a mediumterm budget framework that matches expected government and donor resources to expenditure plans, with a threeyear planning horizon that is rolled forward annually as part of the budget cycle. Resource allocation to the sector is based on national priorities. The budget constraint within which the sector has to manage helps ensure that (a) enthusiasm for sector programs Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 17 does not distort the overall pattern of government spending, and (b) the government is able to meet its own commitments to the sector. Innovative Programs The government implemented a contractual approach to service delivery--in education through the transfer of capitation grants to schools, and in health through performancebased contracting. In particular, capitation grants were successfully transferred from the central government via districts to primary schools to cover the loss in fee income from the 2004 introduction of the government's policy of feefree primary education. The size of the grant increased from RwF 300 in 2004 to RwF 1,000 in 2005 and RwF 2,500 in 2006; it reached RwF 5,300 in 2007.7 District Education Funds to support particularly needy students were set up. In addition to their obvious benefits, these programs demonstrated the flexibility needed to rapidly respond to unexpected fluctuations in donor resources. Mutuelles were piloted successfully over the last 10 years. Mutuelles pool funds from community members to cover a package of basic health services provided at the health center level and the transfer of patients, when needed, to referral hospitals. The mutuelle's objective is to smooth the cost of health services for members, eliminating the hardship of making payment for health services out of pocket. Mutuelles play an important role in intermediating between health centers, district hospitals, and the population. Evaluations show that mutuelles are more effective when they have strong community participation in their governance structures and make payments to the health centers on a capitated basis, essentially transferring all risk to the health centers. In 2005, the focus was on building administrative and management support and technical capacity, including training and development of appropriate tools. In 2006, the government transferred funds to cover premiums for the poorest people in the community (about US$0.15 per capita). This amount increased in 2007, mainly due to a health system grant from the Global Fund to fight AIDS, Tuberculosis and Malaria (GFATM). The government is pursuing district pooled funds as well as a national fund for reinsurance, financed by contributions from formal workers. Budget allocations to support membership by the indigent will need to keep pace. Decentralization The government designed a decentralization policy in 2000. The overall objectives of the decentralization policy were to ensure political, economic, social, managerial/administrative, and technical empowerment of the local population to fight poverty by participating in planning and management of their development process. The decentralization program implementation comprised three phases. The first phase occurred between 2001 and 2005. The second phase began in 2006 and is expected to be completed by 2011. The central government retains the main responsibilities for policy and regulatory frameworks, capacity building for local governments, and monitoring and evaluation. The provinces are part of the central government and assist by providing supervisory functions of the local governments. Decentralization implementation is guided by the Rwanda Decentralization Strategic Framework, which was approved by the cabinet in 2007, and by the fiveyear Decentralization Implementation Program from 2008 to 2012. Decentralized systems 18 World Bank Working Paper are designed to deliver on democratic governance, sustainable local development, and efficient and effective service delivery. Achievements were made in all three areas. These achievements, as well as some remaining challenges and planned next steps, are outlined below. Achievements in democratic governance included (a) participatory processes at the umudugudu level are firmly established and their popularity is increasing; (b) the representative election process at sector and district levels are now better understood; (c) imihigo has now been institutionalized at village, cell, sector, and district levels as a transparent performance management process for local government authorities, and the rate of implementation has increased as local officials are rapidly acquiring capacity; (d) the imihigo process has improved collaboration and cooperation between different state institutions and even between state and nonstate development actors at the local level; (e) partnership building in the Joint Action Development Forum and twinning is taking root at all districts; and (f) every quarter local governments hold a Public Accountability Day, which gives an account of what was achieved, indicates the roles of the population in achieving what is planned, and describes future plans. Achievements in sustainable local development include the following: Participatory development planning was institutionalized. All districts now have a fiveyear District Development Plan, an MTEF that reflects EDPRS and local priorities. Vision2020 Umurenge also have fiveyear development plans to be implemented as rolling annual plans. The Ubudehe process has enabled citizens at the umudugudu level to articulate and prioritize their needs and to dialogue with local government authorities in the development planning, implementation, and monitoring and evaluation processes. Areabased programs offer an opportunity for better coordination of government programs. Joint Action Development Forums were institutionalized at district and sector levels. Itorero (Leadership Training Centers) started at the district level in 2007. Around 24,000 graduates from Itorero have the potential to be effective economic development change agents. Economic Development Advisory Councils were set up at every cell in which the Intore are members. Achievements in terms of effective and efficient service delivery include (a) progress in achieving national and international targets and objectives (for example, Vision2020, MDG targets); (b) approval and production of procedure manuals, including human resource manual, administrative manual, public finance manual, as well as various laws and regulations; (c) delegation to the districts of more service delivery responsibilities in health and education, although the districts require more guidance from central levels; and (d) use of ICT programs in local governments for data management. A clear structural framework lays out the roles, responsibilities, and relationships between different layers of public administration. There are legal frameworks and operating manuals to guide and assist staff, and there are short and mediumterm development strategies that provide an overall framework for improving the delivery Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 19 of government services. Nevertheless, despite notable accomplishments outlined above, many challenges remain: There are insufficient resources to support planned activities. Local governments have a narrow tax base. Although central government transfers have increased in recent years, a wide gap remains between demands on local governments and their capacity to finance the needed services. Facilities remain inadequate.8 Human resource capacity is weak, in terms of both numbers and skills, especially at the umudugudu level. There are only four workers available for every nine required to do the job. Sector service provision manuals are lacking in local government. Sectoral policies, laws, and procedure manuals are not disseminated to the grassroots, and enforcement mechanisms are not yet monitored. Rampant and ad hoc centrally organized meetings, seminars, and workshops consume local government's time and capacity. The decentralization program in Rwanda has progressed well. It is now deeper, wider, and therefore more complex to plan, implement, monitor, and evaluate. As part of its ongoing efforts to improve conditions and facilitate the strengthening of district government services, the Ministry of Local Government, Good Governance, Community Development, and Social Affairs (MINALOC), with support from the Ministry of Public Service and Labor (MIFOTRA), collaborate with the districts and development partners to determine the districtlevel capacity needs. Needs assessments carried out in all 30 districts have been refined and reviewed, and capacitybuilding plans are completed. The district capacity needs assessments provided insights into the challenges ahead, as well as some of the potential solutions: Design a more responsive resourcesharing mechanism between central and local governments. Study and analyze both revenue and investment potential. Encourage and facilitate local investment groups; devise investment incentives at the local level. Devise an interim measure on land management, as is being done by the Ministry of Lands, Environment, Forestry, Water, and Natural Resources (MINITERRE), to reduce land conflicts. Carry out stakeholder studies and provide policy orientation on the privatization of local revenue collection. Strengthen, facilitate, and build the capacity of Itorero as a culturechanging center. Develop and institutionalize production input management, postharvest storage, processing, transformation, and marketing systems as a matter of urgency. Increase the effectiveness in the back and forward links in wealth production. 20 World Bank Working Paper Create a district capacitybuilding basket fund that is demand driven and mirrors capacitybuilding plans. MINALOC and development partners are working on this. Create a body that will have the focus and capacity to manage and continuously inform the decentralization programs. Develop and disseminate manuals and guidelines for service provision at the local level to ensure harmonization of procedures, consistency with provisions of the law, and smooth management of local government activities. Reinforce civil society. Develop a local government taxpayer database that shows the fiscal capabilities of all districts and can serve as a tool to reduce tax evasion. Financial Accountability: Public Financial Management and Procurement The government carried out a comprehensive reform program of the national public financial management system, including an expansive decentralization process. New agencies such as the Rwanda Revenue Authority, National Tender Board, and Office of the Auditor General were created by the government with sufficient autonomy to increase transparency and efficiency in their respective domains. The Office of the Auditor General, which reports systematically to Parliament, produces public audit reports, but it does not yet conduct performance audits of government programs that assess progress toward expected results. To coordinate the different efforts in this area, in 2003 the government developed a Financial Accountability Review and Action Plan (FARAP), which provides a comprehensive road map of reforms to guide policy in the coming years. Implementation of the FARAP is coordinated by a Public Financial Management Reform Steering Committee, and the government plans to resort to a Public Expenditure and Financial Accountability Strengthened Performance Measurement Framework as the basis for annual monitoring of public financial management in the future. The Office of the Accountant General, created in 2005, is expected to oversee and coordinate all functions of the Treasury, Public Accounts, and Internal Audit Units. The government has adopted a new Organic Budget Law, which will consolidate all previous laws governing the public financial management system. The 2005 World Bank CPIA performance criterion that assesses the quality of budgetary and financial management places Rwanda at 3.5 on a scale of 1 (very weak) to 6 (very strong). The MINECOFIN created an integrated financial management system known as SIBET, aimed at monitoring transactions in budget preparation and execution. Key departments were connected to the SIBET system through an information system known as Public Books in 2005. For priority ministries (health, education, agriculture, and internal affairs), quarterly reports are produced for the recurrent budget. However, expenditures are tracked on a paymentorder basis--as opposed to regular financial reporting from ministries, agencies, and subnational governments--because of weak capacity in accounting practices and standards. In addition, the government rolled out new budget software known as SmartGov, which provides financial data that feeds a spreadsheet with relevant objectives, indicators, and targets. This software is designed to be linked to the PRS monitoring system. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 21 The government started to address weaknesses in capacity that slowed down reforms in public procurement by deconcentrating public procurement to budget agencies, with oversight by the National Tender Board. A National Procurement Code was adopted in the third quarter of 2004, giving clear guidelines for all government tendering procedures. The Government Procurement Action Plan foresees the development of an improved regulatory framework for procurement, standard bidding procedures and documents, user manuals, and other tools designed to improve institutional capacity. A Public Procurement Law and a law establishing the Public Procurement Regulatory Agency, which will replace the National Tender Board, were submitted to Parliament. An administrative and financial director is responsible for the public tendering process at both the central and provincial levels, and all spending entities are in the process of hiring procurement officers. Rwanda ranked 121st out of 163 on Transparency International's 2006 Corruption Perceptions Index. The country scored 2.5 out of 10, where 0 is most corrupt and 10 is most transparent. Despite the impact of the war and genocide, the government remains organized and has capacity for policy formulation and implementation. The Rwanda Revenue Authority is perceived by many as an efficient and transparent body that has done much to reduce customs and tax corruption nationwide. The 2003 Constitution established an Office of the Ombusdman, which is operational and charged with fighting corruption and injustice at all levels of the administration. Some ministries, departments, and agencies are only partially responsive to the needs of local communities; many have weak capacity for preparing and implementing investment projects, executing budgets, and monitoring and evaluating programs. While weak capacity, especially at umudugudu level, remains a challenge, local government structures have seen improvements in staffing since 2006 contributing to improved service delivery. The central government was reformed to improve efficiency and increase the focus on policy development and monitoring. A case study of the district of Rulindo was carried out in April 2008 to highlight the progress on district levels since 2006 and the progress and challenges to date (see appendix C). Recent Achievements 2004­07 To encourage donors to provide aid in the form of GBS and to align, to the extent possible, their practices around government processes and systems, the government recognized and implemented institutional reforms to further strengthen its systems (monitoring, financial reporting, and accountability). In addition, while the budget continues to rely on development assistance, Rwanda remains active in its efforts to attract international business. These more recent efforts include measures to improve the trade and business environment, such as a regulatory framework for private sector development and land registration. In addition to efforts at the sectoral level, donors supported the government's private sector development efforts with budget support and nonbudget resources, including technical assistance. Results include the formulation and adoption of a private sector strategy, followed by an export promotion strategy and the creation of Rwanda Investment and Export Promotion Agency (RIEPA) in 2004.9 A onestop shop was created in RIEPA to facilitate business development, and the first phase of an Export Processing Zone was initiated. Consequently, exports have increased significantly: receipts from coffee and tea 22 World Bank Working Paper increased in 2005 and exports of hides, skins, cassiterite, and coltan increased. Simultaneously, tourism receipts and numbers of visitors increased. Privatization has begun (for example, state telecom company, major coffee exporter, several tea factories, and commercial banks) and continues. Banking sector regulation and supervision were strengthened, and a regulatory framework for microfinance was created. Implementation of the government's ambitious information communication technology strategy has begun with the laying of fiber optic cable and installation of wireless capability. Rwanda also joined COMESA and the East African Community (EAC). Reforms were implemented in the areas of governance, expenditure management, government revenues, civil service, public enterprise, central bank and monetary policy, finance, social security, exchange rate regime, domestic price policy, and so forth. Rwanda is increasingly considered to be an excellent example of how country owned reform can produce tangible results. Timely donor support in a difficult situation contributed to stabilization and provided the motivation to focus on a medium to longterm reform agenda. GBS donors supported an ambitious and sustainable longterm, governmentdesigned reform agenda, including the creation of core institutions such as the Rwanda Revenue Authority and the National Tender Board. Reforms over the 2004­07 period are summarized in table 2.3, and an overview of completed reforms between 1995 and 2003 is presented in appendix A. Predictability of Budget Support Among the key aid effectiveness issues agreed to by donor and recipient governments in Paris in 2005 was aid predictability. As is widely acknowledged, especially for highly aiddependent countries like Rwanda, aid predictability is a requisite for the government to plan and effectively implement its programs. By contrast, low predictability is costly because it translates into adjustments in government consumption and investment plans, with potentially harmful effects on results and possibly on growth. Donor commitments to address predictability concerns led to some promising results in shortterm predictability of budget support in some good practice countries; nevertheless there remains a high variance in predictability across these countries. A recent IMF study based on a sample of 13 highly aiddependent countries, including Rwanda, highlights that the variance of aidfinanced components in national budgets declined between 1993­99 and 2000­05 (Celasun and Walliser 2007). The problems for the governments stem from both changes in the disbursed amounts (as compared with committed amounts), and changes in the timing or phasing of disbursements. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 23 Table 2.3. Reforms Implemented by the Government of Rwanda (2004­07) Theme/Sector Reform Measures Undertaken Year Private Sector Development and Economic Growth Private Sector Two more tea factories privatized 2007 Development and National microfinance policy adopted 2006 Economic Growth New microfinance law drafted and submitted to Parliament 2006 Rwanda Agriculture Development Authority (RADA) established 2006 Rwanda Animal Resource Development Authority (RARDA) established 2006 Rwandatel, RwandaEx, and two tea factories privatized 2005 Export promotion strategy developed 2005 Land law published in the Official Gazette 2005 Agriculture sector MTEF aligned with Sector Strategic Plan 2005 Regulatory framework for microfinance adopted 2005 National Agricultural Policy and Sector Strategic Plan (PSTA) developed 2004 Premium pricing of tea leaves for producers introduced 2004 Agriculture Guarantee Fund established 2004 Investment code revised to align with revised tax code 2004 Commercial chambers established 2004 Improved Access and Quality of Human Development and Infrastructure Services Education Independent Student Financing Agency of Rwanda established 2006 National Council of Higher Education established 2006 Long-Term Strategy and Financing Framework for education development 2006 Higher Education Law passed 2005 Organic Education Law passed 2004 Education Sector Strategic Plan (ESSP) and sector MTEF developed 2004 Capitation grants established 2004 District Education Funds adopted to support particularly needy students 2004 Health Performance-based scheme expanded to all 12 regions 2006 Health sector strategy and MTEF developed 2004 Performance-based contracting approach for high-impact health services introduced 2004 Mutuelles (local insurance scheme) scaled up 2004 List of essential drugs prices supplied by drug purchasing agency published 2004 Policy and prices for antiretroviral therapy elaborated 2004 Water Sectoral working group established 2005 Water policy drafted 2005 Water supply activities at district and community levels implemented 2004/05 Energy Passage of electricity and gas legislation 2006 Adoption by Cabinet of revised electricity tariffs 2006 Tariff rate increased from RwF 42 to 81 and further to RwF 112 2004/05 Economic Governance, Transparency, and Accountability Public Organic Budget Law adopted 2005 Financial Ministerial accounts converted to zero balance in move to Single Treasury Account 2005 Management, Accountant General appointed 2004 Transparency, and Procurement code published in the Official Gazette 2007 Accountability Financial policies and procedures manuals finalized, adopted, and published 2006 Public Procurement Law submitted to Parliament 2005 Use of citizen report cards piloted 2005 Cacaca process full under way 2004/05 National Decentralization Steering Committee and National Decentralization Implementation 2004 Secretariat operational Strategic plan to combat corruption elaborated 2004 Media further liberalized 2004 Monitoring Monitoring indicators under development at the decentralized level 2005 MINECOFIN reorganized to better handle monitoring and evaluation 2005 PRSP Annual Progress Reports produced 2004/05 Sector Issue Papers produced 2004/05 MTEFs for key sectors produced 2004/05 Source: Data compiled by authors from multiple sources. 24 World Bank Working Paper Trends in predictability, as measured by mean absolute deviation of budget support commitments over several years, show an improvement (see figure 2.4). In 1990­99, Rwanda was one of two countries where budget support was least predictable. In Rwanda, more than twothirds of committed budget support was not disbursed over this period. Fortunately, the situation has since significantly improved. Between 2000 and 2005, the proportion of undisbursed support, although still too high to manage, reduced to less than onesixth of the total committed by budget support donors. However, since then budget support disbursements predictability has improved. With the lead of the World Bank, which committed to early disbursements under the PRSC/G series, the other budget support donors decided to follow suit, and by 2007 most funding was received within the first two quarters of the fiscal year. Although this is considered a major improvement in terms of predictability, the fact that the disbursements are committed only one year at a time still poses challenges to the government in its planning processes. Figure 2.4. Average Budget Aid Arriving on Schedule in Selected Number of Countries 0.9 0.8 Only 65% of aid arrived on schedule 0.7 0.6 0.5 % of GDP 0.4 0.3 0.2 0.1 0 9 5 so so a a nia nia da da da da ar ar ­9 ­0 an an sc sc Fa Fa an an an an za za 93 00 Gh Gh ga ga Ug Ug Rw Rw n n na na 19 20 Ta Ta da da rki rki Ma Ma le le Bu Bu mp mp Sa Sa le le ho ho W W This + mean absolute deviation = average budget aid Mean absolute deviation Source: IMF Finance and Development 2007. In general, the term predictability refers to short, medium, and longterm disbursements as compared with expected commitments, which could also include what is frequently referred to as intraannual disbursements. In this chapter, the focus is on highlighting the importance of budget support predictability in the context of the government's annual budget cycle. The government makes budget allocation decisions within a framework that involves different aid modalities and involves collaboration with a large number of donors. In terms of planning of expenditure, the government, based on planned outcomes, allocates available resources to recurrent and development spending. In Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 25 addition to this constraint of balancing recurrent and development expenditure, there is the constraint of the timing of disbursement of various aid modalities--and the impact on capacity, counterpart funding, and so forth--on top of the unpredictability of aid across all modalities within the budget cycle. Budget support commitments are based on annual reviews, even if some donors active in Rwanda are able to make indicative commitments to facilitate mediumterm planning. GBS is disbursed to the Treasury to finance regular budgetary expenditure, and as such budget support resources are applied to finance both recurrent and development spending. GBS funds can hence be pooled with internal government revenue. In fact, this feature of GBS, which allows it to be seamlessly integrated into national processes, is seen as a major advantage in terms of planning and budgeting. The cost of this advantage is that GBS shortfalls are more difficult to manage than shortfalls in the disbursement of other aid modalities. Low predictability of budget support imposes difficult choices on the government and undermines the effectiveness of budget support in meeting the government's objectives of strengthening internal planning processes. More specifically, when budget support disbursements are unexpectedly cancelled in the middle of the government's budget cycle, the government finds itself obliged to collect more revenue, cut spending, or obtain more domestic financing, unless other financing sources such as debt relief happen to substitute for GBS shortfalls. Similarly, unplanned disbursements could substitute for tax revenue, resulting in unplanned expenditure increases or repayment of debt. Notes 1 Literally, "justice on the grass." 2 Caution is warranted before drawing conclusions as the case of Rwanda's decline from 2004 to 2005, especially on corruption, is very unusual, particularly due to the lack of statistical significance. Because the estimates of governance are based on extremely limited data sources (for corruption, there were four, while typically there are seven for other countries), the margin of error is among the highest of all countries rated. The period 2004­2005 is too short to identify any trends. 3 UN agencies and bilateral donors supplied trained primary health care providers and medical supplies and equipment, and repaired damages to water and sanitation systems. 4 The SPA, an association chaired by the World Bank, coordinates resources and works to improve donor policies and practices for greater aid effectiveness. The partnership currently includes AfDB, Belgium, Canada, Denmark, the Economic Commission for Africa, the EC, Finland, France, Germany, the IMF, Ireland, Italy, Japan, the Netherlands, Norway, the Development Assistance Committee of the Organisation for Economic Cooperation and Development, Portugal, Sweden, Switzerland, the United Kingdom, the United Nations Development Programme, the United States, and the World Bank. 5 The SPA had identified three initial countries, Ethiopia, Rwanda, and Senegal, as pilots for budget support harmonization projects. 6 The number of conditions specified in budget support arrangements is not negligible and represents a heavy burden on the government in terms of processed information. For instance, the IMF's PRGF identifies 13 structural conditionalities (with 6 subconditions) and 49 core actions and reforms to be taken during 2002­2004; the World Bank's Institutional Reform Credit (IRC) outlines 19 conditions, with 48 intermediate benchmarks, to be monitored by more than 50 26 World Bank Working Paper indicators. These operations were developed independently after the finalization of Rwanda's Poverty Reduction Strategic Plan. 7 Before 2008, the amount transferred per student was intended to cover both functional costs and bonuses to teachers. However, some teachers failed to receive their full bonus. Therefore, in 2008, the perstudent capitation amount was split so that RwF 3,800 is being paid for each student plus an additional RwF 12,500 times the number of teachers per school. This was done to avoid problems encountered in previous year. 8 Of the total number of sectors, 66 percent have offices, and only 49 percent of the total number of cells have offices. (The local government structure after decentralization reform now includes five districts [including the City of Kigali], 416 sectors, and 2,150 cells. Cells are the lowest level of local government and consist of a cluster of villages [imidugudu which is the plural form of umudugudu used and explained in other parts of this report]. See appendix E.) The transport policy undertaken by government has greatly improved communication across local governments, but it is putting big strains on district budgets. Enabling software and computers are not fully integrated in district administration, and insufficient numbers of staff have access to the Internet, intranet, and telephone. 9 RIEPA now falls under the recently created Rwanda Development Board (RDB). RDB is modeled after similar economic development agencies created by other emerging nations, particularly some of the "Tiger" countries. RDB was created by merging eight existing government development agencies into one that will report directly to the office of the President, including agencies such as RIEPA, the Centre for Support to Small and Medium Enterprises (CAPMER), Rwanda Commercial Registration of Service Agencies (RCRSA), The Environmental Impact Assessment (EIA) unit and Privatization Secretariat. RDB's main responsibility is to fast track development activities by both the government and the private sector. CHAPTER 3 Domestic and External Resources, Recurrent and Capital Expenditures, Overview of Aid Dependency R wanda has experienced a fiscal expansion in the past several years. Analytical work extending as far back as 2001 shows fiscal expansion that more than doubled between 2001 and 2007.1 The data in this analysis, focusing on 2004­07, confirm this expansionary trend. As the table 3.1 shows, total actual budgetary expenditures in nominal terms--including recurrent expenditure, capital expenditure, net lending, and payments of arrears--increased from RwF 305.9 billion in 2004 to RwF 524.3 billion in 2007, a 71 percent increase in the fouryear period. This fiscal expansion was followed by a consistent increase in domestic revenue. Actual tax revenue increased from RwF 134.6 billion in 2004 to RwF 237.8 billion in 2007 (77 percent increase), while actual nontax revenue increased from RwF 12.4 billion to RwF 15.1 billion in the same period (22 percent increase). The fact that expansionary fiscal policy was paralleled by increases in domestic resources shows that the government was prudent in its fiscal expansion by ensuring that increased public expenditure was financed by more than just increases in external resources. Furthermore, looking at net resources over the same period in real terms, one can also see an expansionary trend, although a less significant one. For example, in real terms, domestic resources increased by 34 percent between 2004 and 2007, while external resources increased by only 2 percent. Tax revenues increased by 28 percent, and there was a 12percent decrease in nontax revenues. Grants also increased by 16 percent in real terms, but were offset by a 33percent decrease in foreign financing and external loans. Table 3.2 further illustrates actual net resources in real terms. As background context for the 2004­07 analytical trend analysis in this report, it helps to briefly review previous analytical work covering 2001­03. As previous analytical work shows, the government underspent by 0.2 percent and 1.2 percent of GDP in 2001 and 2002, respectively, because projected revenue did not fully materialize. Then in 2003, unplanned outlays on presidential and legislative elections and rehabilitation of hotels as part of the tourism strategy implementation led to an overspending of RwF 16 billion (1.8 percent of GDP) (IMF 2005b). 27 28 Table 3.1. Summary of Net Resources in Nominal Terms 2004­07 World Bank Working Paper Source: MINECOFIN. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 29 Table 3.2. Summary of Net Resources in Real Terms 2004­07 REAL Terms (million RwF) 2004 2005 2006 2007 Growth Summary Net Resources Actual Actual Actual Actual 2004-07 I -Domestic Resources 191,805 212,118 228,867 257,700 34% a. Tax Revenues 186,107 206,363 212,479 237,800 28% b. Non-Tax Revenues 17,144 22,514 16,388 15,100 ­12% c. Other Domestic Financing (11,445) (16,758) -- 4,800 ­142% II -External Resources 250,897 272,594 189,357 256,400 2% a. Grants 177,958 208,827 189,357 207,300 16% b. Foreign Financing/External Loans 72,939 63,767 -- 49,100 ­33% Total Budget Resources (I + II) 442,702 484,713 418,223 514,100 16% Source: MINECOFIN (in 2007 prices). The largest fiscal expansion occurred between 2003 and 2004, when the increase in total expenditure as a percentage of GDP, in nominal terms, was projected to reach 42 percent. However, actual total expenditure fell short of projections. This gap between the projected and the actual budget was not so much a result of lack of absorptive capacity, but rather a move to maintain macroeconomic stability, allowing the government to contain domestic demand and inflationary pressures (IMF 2005c). As table 3.3 illustrates, GDP has grown in real terms of about 6 percent in the last four years. Table 3.3. Line Ministries Expenditures and Resources as a Percentage of GDP GDP 2004 2005 2006 2007 Nominal GDP (mill RwF) 1,137,900 1,327,100 1,583,000 1,826,200 Real GDP Growth Rate/year^ 5% 7% 6% 6% OVERALL EXPENDITURES * (Actuals) AS A % OF GDP 2004 2005 2006 2007 Total Expenditures (Recurrent) 208,988 240,270 281,540 356,025 as a % of GDP 18% 18% 18% 19% Total Expenditures (Development) 79,920 120,610 151,691 181,870 as a % of GDP 7% 9% 10% 10% TOTAL Expenditures 288,908 360,880 433,230 537,894 as a % of GDP 25% 27% 27% 29% OVERALL RESOURCES * (Actuals) AS A % OF GDP 2004 2005 2006 2007 Total Domestic Resources 138,716 167,070 203,900 257,700 as a % of GDP 12% 13% 13% 14% Total External Resources 181,451 214,702 168,700 256,400 as a % of GDP 16% 16% 11% 14% TOTAL Domestic & External 320,167 381,771 372,600 514,100 as a % of GDP 28% 29% 24% 28% Source: MINECOFIN. *Shown in nominal terms. ^Based on estimates by MINECOFIN 2007. Furthermore, table 3.3 shows that total recurrent and development public expenditures represented 19 percent and 10 percent, respectively, of GDP in 2007, 30 World Bank Working Paper translating into 29 percent of GDP for total expenditures. Over the period, total recurrent expenditures remained consistent at 18­19 percent of GDP, while development expenditures increased from 7 percent of GDP in 2004 to 10 percent in 2007. As noted in the earlier chapter, Rwanda has experienced a clear fiscal expansion since 2001, with a significant increase starting in 2004, when total actual public expenditure increased in nominal terms by 24 percent from RwF 305.9 billion in 2004 to RwF 379.0 billion in 2006, and RwF 524.3 billion in 2007 (see table 3.1). While a review of actual public resources over the same period yields slightly different numbers, the same trend is confirmed (see table 3.1 and overall resources as a percentage of GDP in table 3.3). In real terms this translates into a 16 percent increase in actual public resources between 2004 and 2007. It is noteworthy that, as can be seen in table 3.1 and table 3.3, there are some discrepancies between actual expenditures and actual resources. While budgeted amounts for total public expenditures and total public resources are equal, actual resources and expenditures show some differences in numbers. This difference can in part be explained by the fact that, as mentioned in earlier chapters, actual development expenditures for the period are taken from different source. For example, in table 3.3, the revised budget development expenditure data is used in 2007 while CEPEX data is used for 2006). Some differences between expenses and resources can also be explained by differences in reporting of numbers for actual data by MINECOFIN. This chapter evaluates the net resources that provide the basis for the fiscal expansion seen in the period, first reviewing the trends in both domestic and external resources; then reviewing the trends in overall expenditures; and finally looking briefly at how the expansion was funded, the level of aid dependency and aid sustainability, and the degree of fiscal space. This chapter illustrates type of public spending by breaking down overall resources by domestic and external resources and overall expenditures by recurrent, capital, net lending, and arrears and shows how this spending was financed. In contrast, chapters 4 and 5 of this report review public spending in terms of central and local government sector allocations, distinguishing between recurrent and development spending. This outline is consistent with the outline of the Organic Budget Law. As is true with this report in general, data in this chapter are primarily based on MINECOFIN data unless otherwise indicated and some differences in underlying assumptions may exist in comparison to other analytical work quoted in this report. Resources This section looks at overall resources (internal and external resources) to illustrate how Rwanda's fiscal spending has being financed. The section below reviews overall resource trends in Rwanda in real terms for 2004­07 based on data in table 3.4. As table 3.4 shows, based on original budget projections, domestic resources were anticipated to increase in the period as a percentage of total resources. In 2004, domestic resources represented 41 percent of projections, increasing to 43 percent and 48 percent, and then down slightly to 46 percent in 2005, 2006, and 2007, respectively. However, as can be noted, actual domestic resources as a percentage of total resources exceeded projections in each of the years listed: 43 percent, 44 percent, 55 percent, and 50 percent for 2004, 2005, 2006, and 2007 respectively. ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 31 nd ­07 Table 3.4. Overview of Resources--Original Projections an Actual 2004­ in million 2004 2004 2005 2005 2006 2006 20007 2007 RwF Orig Actual Orig Acctual Orig Actual Orrig Actual Domestic 91,828 19 191,805 200,576 2,118 212 220,198 228,867 ,700 230, 257,700 External 70,755 27 250,897 267,013 2,594 272 234,099 189,357 ,045 276, 256,400 Total 62,583 46 442,702 467,589 4,713 484 454,297 418,223 ,745 506, 514,100 % of Total 2004 2004 2005 2005 2006 2006 20007 2007 Resources* Orig Actual Orig Acctual Orig Actual Orrig Actual Domestic 41% 4 43% 43% 44% 48% 55% 6% 46 50% External 59% 57% 57% 56% 52% 45% 4% 54 50% Total 100% 1 100% 100% 00% 10 100% 100% 0% 100 100% Source: MINECO OFIN (in 2007 real prices). s should equal tota *Total resources eported data are co al expenditures if re onsistent. Figure 3.1 compares orig ginal resource p projections verssus actual resou urces realized. As the graph shows, Rwand da overestimated in 2004 and 200 d its resources i 06, and it was too conservat a crease in 2007 tive in its projections in 2005. Furthermore, a significant inc showed an an projections s nticipation that resources wo ould increase even further compared wit th previous yea ars. esource Trends Comparing Or Figure 3.1. Re s ons 2004­07 riginal Projectio vs. Actual 2 In million RwF OFIN (in 2007 real prices). Source: MINECO 32 nk World Ban Working Paper Furthermo t y n ore, looking at resources by type--comparing trends in domestic, external, and t --for 2004­07, one can clearly see that whil domestic total resources-- y le resources increeased steadily in the period and also cons or sistently met o exceeded tions, external resources fluct original project tuated more. A figure 3.2 shows, actual As rnal resources did not meet p available exter d 2004 and 2006, and actual projections in 2 external resourrces even decr n nd reased between 2005, 2006, an 2007. In ad ddition, the government's original projec t ctions for 2007 indicated that external reso ources were expected to inccrease significanntly from 2006 to o 2007 to levels in par with 20004 and 2005, while the projeected increase in n internal resouurces was expec e cted to be more moderate. ns r The fluctuation in external resources help explain the flu otal overall uctuations in to resources wher re, in 2004 and 22006, total projections were too optimistic. source Trends by Type of Reso Figure 3.2. Res b ource 2004­07 In million RwF FIN (in 2007 real pr Source: MINECOF rices). The follow wing two graph further snapsho of projected and actual hics provide a f ot ows the projected trends in r resource trends. Table 3.5 sho 2004­07. As real terms for 2 shown in the table, domestic projected rev c venues were pr rojected to incrrease by 20 percent in the p period, while exxternal resourcees were expecte ed to increase by y 2 percent, representing a overall projected increase of 10 percent The trends are further an t. gure 3.3. illustrated in fig ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 33 Domestic and Ex Table 3.5. Overall Original Projections for D ces xternal Resourc 2004­07 2004 2005 2006 2007 Growth in million RwF Orig Orig Orig Orig 2004-07 Domestic 191,828 200,576 220,198 230,700 20% External 270,755 267,013 234,099 276,045 2% Total 462,583 467,589 454,297 506,745 10% Source: MINECO OFIN (in 2007 real prices). verall Original Projections for Domestic and E Figure 3.3. Ov P rces 2004­07 External Resour In million RwF OFIN (in 2007 real prices). Source: MINECO Figure 3. takes a close look at resou .3 er s. urce projections As can be se een, domestic revenues we y ere consistently projected to increase in t the period 200 04­07, while or projections fo external reso a ources reflect a limited expec ased external ctation in increa funding, with h the exception of 2007 when th he projected ex xternal funding jumped from RwF 234.1 bil llion in 2006 to RwF 276.1 billi ion in 2007 (see e table 3.5). When looking at rce actual resour trends, the picture is sligh htly different. Table 3.6 show the actual ws resources ava ese ailable, and the data are cap ptured in figure 3.4 to further illustrate the e ual overall resou trends in actu urces. 34 nk World Ban Working Paper rall rnal Resources 2004­07 Table 3.6. Over Actual Domestic and Exter 2004 2005 2006 2007 Growth in million RwF Actual Actual Actual Actual 2004­07 Domestic 191,805 212,118 228,867 257,700 34% External 250,897 272,594 189,357 256,400 2% Total 442,702 484,713 418,223 514,100 16% FIN (in 2007 real pr Source: MINECOF rices). Table 3.6 sshows that in te erms of actual s spending, the tr rend differs from m projected resources. For 2004­06, actual domestic reven nue in real termms increased by y 34 percent while external resources increeased by 2 perccent, and comb o bined these two subtrends contributed to a 16 percent increase in total actual resources. Figure 3.4 provides a snapshot of thee overall actual r resource trends as discussed ab bove. erall Actual Resource Trends b Type of Reso Figure 3.4. Ove by ource 2004­07 In million RwF FIN (in 2007 real pr Source: MINECOF rices). While dom s ar mestic revenues show a simila picture comp pared with pro ojections--a steady increas r w se--external resources show more fluct tuations comp pared with projections. Exxternal resource projections i 2004­07 were rather flat, but actual in nal available extern resources sh how an increas in 2005, and then a significant drop in se 2006, to be increased again in 22007, showing th npredictability. he pattern of un Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 35 Domestic Resources In this section, domestic revenue will be examined closely, looking at both projected and actual data. Domestic revenue consists of tax revenue, nontax revenue, and other financing. A further, detailed breakdown of each of these domestic revenue categories can be found in Data Sets 20­23, available on request. See appendix F of this report. From a longterm perspective, the level of expenditure is directly related to the capacity of a country to raise domestic revenue effectively (Sayinzoga 2007). Table 3.7 provides an overview of both projected and actual domestic revenues for the 2004­07 period. Table 3.7. Overview of Domestic Resources--Original Projections vs. Actual 2004­07 in million 2004 2004 2005 2005 2006 2006 2007 2007 RwF Orig Actual Orig Actual Orig Actual Orig Actual TOTAL 191,828 191,805 200,576 212,118 220,198 228,867 230,700 257,700 Tax Rev 172,124 186,107 180,902 206,363 187,980 212,479 207,400 237,800 NonTax Rev 12,790 17,144 16,378 22,514 14,372 16,388 9,700 15,100 Other Finan. 6,914 ­11,445 3,295 ­16,758 17,847 0 13,600 4,800 % of Total Domestic 2004 2004 2005 2005 2006 2006 2007 2007 Res. Orig Actual Orig Actual Orig Actual Orig Actual TOTAL 100% 100% 100% 100% 100% 100% 100% 100% Tax Rev 90% 97% 90% 97% 85% 93% 90% 92% NonTax Rev 7% 9% 8% 11% 7% 7% 4% 6% Other Finan. 4% ­6% 2% ­8% 8% 0% 6% 2% Source: MINECOFIN (in 2007 real prices). As is clear from the data above, Rwanda has seen a consistent increase in real terms in its overall domestic revenues, both projected and actual domestic resource generation. Original total projections for 2004 were RwF 191.8 billion, while in 2007 the projections were estimated at RwF 230.7 billion. Of this increase, tax revenue is the main driver of increased domestic revenues, increasing in actual terms from RwF 186.1 billion in 2004 to RwF 237.8 billion in 2007, despite a rather limited tax base in the country. Figures 3.5 and 3.6 further illustrate the trends in domestic revenue generation. Figure 3.5 compares original and actual budget projections for domestic revenues. Figure 3.6 takes the same data but looks at domestic revenue by type of resource. Clearly, overall domestic revenue is closely tracked by overall tax revenues, as nontax revenue and other financing provide only a small portion of overall domestic revenue. Also, actual tax revenue has consistently exceeded projections, resulting in overall domestic revenues also exceeding projections each year except 2004 (as a result of a negative trend in other financing). 36 nk World Ban Working Paper mestic Revenue Trends Compa Figure 3.5. Dom e Projections vs. A aring Original P Actual 2004­07 In million RwF FIN (in 2007 real pr Source: MINECOF rices). mestic Revenue Trends by Typ of Revenue S Figure 3.6. Dom e pe 7 Source 2004­07 In million RwF FIN (in 2007 real pr Source: MINECOF rices). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 37 Clearly, if the government does not manage to raise sufficient revenue, the country will have to rely on external resources to finance its fiscal spending, potentially increasing its aid dependency. An increasing domesticresource base, that is, reducing reliance on external resources, may over time ensure that increased fiscal spending is sustainable. Such increase can be accomplished by increasing the tax base of the country and addressing issues such as tax evasion and the role of the informal sector in the economy. While broadening the tax base and the basis for tax generation, the government must simultaneously balance the desire to increase tax resources with minimizing disincentives for private investments, growth, and further tax evasion (Sayinzoga 2007). Tables 3.8 and 3.9 and figures 3.7 and 3.8 show projected and actual resource growth in the period. Table 3.8. Original Projections for Domestic Resources 2004­07 2004 2005 2006 2007 Growth % in million RwF Orig Orig Orig Orig 2004­07 TOTAL 191,828 200,576 220,198 230,700 20% Tax Rev 172,124 180,902 187,980 207,400 20% NonTax Rev 12,790 16,378 14,372 9,700 -24% Other Finan. 6,914 3,295 17,847 13,600 97% Source: MINECOFIN (in 2007 real prices). Figure 3.7. Domestic Revenue Trends in Original Projections 2004­07 250,000 200,000 in million RwF 150,000 100,000 50,000 0 TOTAL Tax Rev NonTax Rev Other Finan. 2004 2005 2006 2007 Orig Orig Orig Orig Source: MINECOFIN (in 2007 real prices). 38 World Bank Working Paper Table 3.9. Actual Domestic Resources for 2004­07 2004 2005 2006 2007 Growth % in million RwF Actual Actual Actual Actual 2004­07 TOTAL 191,805 212,118 228,867 257,700 34% Tax Rev 186,107 206,363 212,479 237,800 28% NonTax Rev 17,144 22,514 16,388 15,100 ­12% Other ­11,445 ­16,758 0 4,800 ­142% Source: MINECOFIN (in 2007 real prices). Figure 3.8. Actual Domestic Revenue Trends 2004­07 300,000 250,000 200,000 in million RwF 150,000 100,000 50,000 0 -50,000 TOTAL Tax Rev NonTax Rev Other 2004 2005 2006 2007 Actual Actual Actual Actual Source: MINECOFIN (in 2007 real prices). The consistent, increasing trend in domestic revenues is clear in figures 3.7 and 3.8, showing first original projections only and then actual revenue only. The consistent increase in domestic revenue can possibly be attributed to the tax reforms initiated in 1996 to improve the country's tax system, where the aim was to shift the tax burden away from factors of production to a more broadbased consumption tax, using taxes such as the value added tax (VAT) (World Bank 2005). As an example, table 3.10 shows the VAT projections for 2004­07. (Actuals for VAT as a subcategory under Tax Revenue are not available for all years.) Based on the above overall review of domestic resources in 2004­07, the picture becomes clear: domestic revenues, driven by a consistent increase in tax revenue, have increased consistently over the period. However, while actual domestic resources in nominal terms have increased from RwF 138.7 billion in 2004 (in real terms, RwF 191.8 billion) to RwF 257.7 billion in 2007, overall public expenditures increased in nominal terms from RwF 305.9 billion to RwF 524.3 billion in the same period. (Note inconsistencies in data in this chapter compared with chapter 5.) Clearly the increase in domestic revenue was not sufficient to finance the fiscal expansion in the corresponding period, further indicating the significant role of external resources. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 39 Table 3.10. Increase in VAT Tax Projections 2004­07 2004 2005 2006 2007 Growth % Real Terms (in million RwF) Orig Orig Orig Orig 2004­07 Value Added Tax & Sales Tax 59,114 59,997 68,132 73,300 24% % of VAT of Total Budget Resources 13% 12% 16% 14% Source: MINECOFIN (in 2007 real prices). External Resources External resources play a significant role in Rwanda's fiscal spending, providing a significant portion of available financial resources in the country in the form of both grants and loans. Some of this external financing is GBS, while other external spending is more specific and project based. This section will provide an overview of external resources, comparing grants and loans trends in 2004­07, looking at both projected and actual data. Understanding the trends in external resources illuminates the issue of aid dependency and the degree to which aid dependency in this period has increased. Table 3.11. Overview of External Resources--Original Projections and Actual 2004­07 in million 2004 2004 2005 2005 2006 2006 2007 2007 RwF Orig Actual Orig Actual Orig Actual Orig Actual TOTAL 270,755 250,897 267,013 272,594 234,099 189,357 276,045 256,400 Grants 152,915 177,958 197,697 208,827 180,109 189,357 241,515 207,300 Loans 117,840 72,939 69,316 63,767 53,990 0 34,530 49,100 % of Total External 2004 2004 2005 2005 2006 2006 2007 2007 Res. Orig Actual Orig Actual Orig Actual Orig Actual TOTAL 100% 100% 100% 100% 100% 100% 100% 100% Grants 56% 71% 74% 77% 77% 100% 87% 81% Loans 44% 29% 26% 23% 23% 0% 13% 19% Source: MINECOFIN (in 2007 real prices). Table 3.11 shows external resources by projected and actual amounts for 2004­07. As can be seen, total external resources in real terms increased during the period for both projected and actual data although moderately. Total actual grants increased from RwF 177.9 billion in 2004 to RwF 207.3 billion in 2007. Loans show a significant decrease, down from RwF 72.9 billion in 2004 to RwF 49.1 billion in 2007. As a percentage of total external resources, grants increased from a projected 56 percent in 2004 (compared with 71 percent actual) to a projected 87 percent in 2007 (81 percent actual). As these fluctuations in proportions and between projections and actual figures show, external grant predictability is a challenge for the government. Figures 3.9 and 3.10 illustrate the trend seen in table 3.11. 40 nk World Ban Working Paper ernal Resource Trends Compa Figure 3.9. Exte Actual aring Original Projections vs. A 2004­07 In million RwF FIN (in 2007 real pr Source: MINECOF rices). xternal Resource Trends by Type of Resource 2004­07 Figure 3.10. Ex e 300,000 250,000 in million RwF 200,000 150,000 100,000 50,000 0 L TOTAL Grants Loans 2004 2004 2005 2005 2006 2006 2007 07 200 Orig Actual Orig Actual Orig al Actua Orig Actual FIN (in 2007 real pr Source: MINECOF rices). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 41 As can be seen, grants resources consistently exceeded projections in the period until 2007, when actuals went down. Loans were consistent in the period, with only a significant overprojection in 2004. In the following tables and figures, external resources are compared first by budget projections and then by actual contributions for the period. Table 3.12. Original Projections for External Resources 2004­07 2004 2005 2006 2007 Growth % in million RwF Orig Orig Orig Orig 2004­07 TOTAL 270,755 267,013 234,099 276,045 2% Grants 152,915 197,697 180,109 241,515 58% Loans 117,840 69,316 53,990 34,530 -71% Source: MINECOFIN (in 2007 real prices). As table 3.12 shows, grants projections increased from RwF 152.9 billion in 2004 to RwF 241.5 billion in 2007, an overall increase of 58 percent. This is in sharp contrast to the projections of loans in the same period, which declined by 71 percent over the period, from RwF 117.8 billion in 2004 to RwF 34.5 billion in 2007. Overall, this translates to a growth rate in external resources of 2 percent, increasing to RwF 276.1 billion by year 2007. This trend is further illustrated in figure 3.11. Figure 3.11. Budgeted External Resource Trends 2004­07 300,000 250,000 in million RwF 200,000 150,000 100,000 50,000 0 TOTAL Grants Loans 2004 2005 2006 2007 Orig Orig Orig Orig Source: MINECOFIN (in 2007 real prices). 42 World Bank Working Paper Projections for grants between 2005 and 2006 were rather consistent, with a significant increase projected in 2007. Since the majority of external funding comes in the form of grants, projected at 87 percent in 2007, the trend in grant projections in the period is closely reflected in the overall external resource projections, with the exception of 2004, when loans played a greater role in external funding. Since 2004, however, projections for loans measured in real terms consistently decreased. Moreover, the trend in actual external resource data presents a different picture. Data show that total actual external resources in real terms were higher in 2005 compared with both 2004 and 2006 and slightly higher than 2007, showing a fluctuating pattern of reliable funding from donors both in terms of yeartoyear commitments and when compared with promised versus actually disbursed funding. Table 3.13 and figure 3.12 show the actual receipts of grants and loans in the period. Table 3.13. Actual Funding from External Resources 2004­07 2004 2005 2006 2007 Growth % in million RwF Actual Actual Actual Actual 2004­07 TOTAL 250,897 272,594 189,357 256,400 2% Grants 177,958 208,827 189,357 207,300 16% Loans 72,939 63,767 0 49,100 -33% Source: MINECOFIN (in 2007 real prices). Figure 3.12. Actual External Resource Trends 2004­07 300,000 250,000 in million RwF 200,000 150,000 100,000 50,000 0 TOTAL Grants Loans 2004 2005 2006 2007 Actual Actual Actual Actual Source: MINECOFIN (in 2007 real prices). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 43 Expenditures In this section, overall expenditures are analyzed comparing recurrent, capital and net lending, and arrears. This provides a different perspective of expenditure data than the expenditure analysis by recurrent and development spending broken down by sector in later chapters. The expenditure data used to analyze recurrent, capital and net lending, and arrears is presented table 3.14. Table 3.14. Overview of Expenditures--Original Projections and Actual 2004­07 in million 2004 2004 2005 2005 2006 2006 2007 RwF Orig Actual Orig Actual Orig Actual 2007 Orig Actual Recurrent 313,980 254,282 300,277 0 304,448 281,397 0 329,300 Cap&NetLend 125,097 142,445 158,424 0 141,992 144,010 0 186,200 Pmt Arrears 23,506 26,247 8,887 0 7,857 0 0 8,800 TOTAL 462,583 422,974 467,588 0 454,297 425,407 0 524,300 % of Total 2004 2004 2005 2005 2006 2006 2007 Expenditures Orig Actual Orig Actual Orig Actual 2007 Orig Actual Recurrent 68% 60% 64% n.a. 67% 66% n.a. 63% Cap&NetLend 27% 34% 34% n.a. 31% 34% n.a. 36% Pmt Arrears 5% 6% 2% n.a. 2% 0% n.a. 2% TOTAL 100% 100% 100% n.a. 100% 100% n.a. 100% Source: MINECOFIN (in 2007 real prices). Note: 0 indicates that data was not provided in part or in full by MINECOFIN. Figure 3.13 compares original expenditure projections and actual expenditures for each of the three expenditure categories and total expenditures. Although data are missing for actual expenditures in 2005 and for original (projected) expenditures in 2007, overall the graph indicates that at least for 2004 and 2006, total actual expenditure was lower than the projection. In particular, in both 2004 and 2006, actual recurrent expenditures were less than projected, while capital expenditures and net lending were more consistent across the period and with projections for each year. Similarly, looking at the types of expenditure for both budget and actual data, figure 3.14 illustrates the fluctuations between projections and actual spending. Recurrent expenditure represents the largest portion of overall expenditure, and in both 2004 and 2006, actual recurrent expenditure was less than budgeted. 44 World Bank Working Paper Figure 3.13. Expenditure Trends Comparing Original and Actual 2004­07 600,000 500,000 in million RwF 400,000 300,000 200,000 100,000 0 2004 2004 2005 2005 2006 2006 2007 2007 Orig Actual Orig Actual Orig Actual Orig Actual Recurrent Cap&NetLend Pmt Arrears TOTAL Source: MINECOFIN (in 2007 real prices). Figure 3.14. Overall Expenditure Trends by Type 2004­07 600,000 500,000 400,000 in million RwF 300,000 200,000 100,000 0 Recurrent Cap&NetLend Pmt Arrears TOTAL 2004 2004 2005 2005 2006 2006 2007 2007 Orig Actual Orig Actual Orig Actual Orig Actual Source: MINECOFIN (in 2007 real prices). ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 45 ns nda, showing Projection for 2004­07 further support the fiscal expansion in Rwan the intent by the government for each of thee four years to c d significantly consistently and increase overall expenditure. Table 3.15 summmarizes the pr rojected expendditures for the period. ­07 Table 3.15. Original Expenditures for 2004­ 2004 2005 2006 2007 Growth % in million RwF Orig Orig Orig Orig 2004 - 06 Recurrent 313,980 300,277 304,448 0 19% Cap&NetLend 125,097 158,424 141,992 0 40% Pmt Arrears 23,506 8,887 7,857 0 -59% TOTAL 462,583 467,588 454,297 0 21% Source: MINECO OFIN (in 2007 real prices). Note: 0 indicates that data was not pro full N. ovided in part or in f by MINECOFIN Figure 3.1 trates the trends 15 further illust d resources ove s in the projected er the period. Original Overall Expenditure Tr Figure 3.15. O rends 2004­07 In million RwF OFIN (in 2007 real prices). Source: MINECO While pr ws n rojections show an indication of the govern on, nment's intentio table 3.16 shows what was actually spent in the period. The t table shows th hat recurrent expenditures increased 30 ppercent betweenn 2004 and 2007 7, while capital expenditures and net lendiing increased 31 percent. Arrears represent a small portion o of the overall expenditure and declined by 66 percent in the period. Overall, total expenditures b increased by 2 he period. 24 percent in th 46 nk World Ban Working Paper tual Expenditure 2004­07 Table 3.16. Act es 2004 2005 2006 2007 Growth % in million RwF Actual Actual Actual Actual 2004­07 Recurrent 254,282 0 281,397 329,300 30% Cap&NetLend 142,445 0 144,010 186,200 31% Pmt Arrears 26,247 0 0 8,800 -66% TOTAL 422,974 0 425,407 524,300 24% Source: MINECOF rices). FIN (in 2007 real pr at vided in part or in ful by MINECOFIN. Note: 0 indicates tha data was not prov ll ctual Overall Expenditure Trend 2004­07 Figure 3.16. Ac ds In million RwF FIN (in 2007 real pr Source: MINECOF rices). Recurrent Expenditures Recurrent expeenditures const est total overall ex titute the large portion of t xpenditures, accounting for more than 60 percent of total e oth projected an expenditures, bo nd actual, in 2004­07. Recuurrent expendi itures are bro oken down i into recurrent operating interest and com expenditures, i mmissions, reim mbursement of public debt, an subsidies nd transfers. Table 3.17 shows th projected and actual data fo recurrent and recurrent t e he d or expenditures inn the period. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 47 Table 3.17. Overview of Recurrent Expenditures 2004­07 in million 2004 2004 2005 2005 2006 2006 2007 2007 RwF Orig Actual Orig Actual Orig Actual Orig Actual Total 313,980 254,282 300,277 0 304,448 281,397 0 329,300 Recur 176,540 156,454 181,014 0 184,972 187,785 0 197,800 Operat. IR & Comm 18,308 16,762 19,708 0 17,557 12,908 0 10,900 Public Debt 68,065 30,065 36,549 0 34,348 0 0 16,800 Subs&Transf 51,067 51,001 63,006 0 67,571 80,704 0 103,800 % of Total Recurrent 2004 2004 2005 2005 2006 2006 2007 2007 Exp. Orig Actual Orig Actual Orig Actual Orig Actual Total 100% 100% 100% n.a. 100% 100% n.a. 100% Recur 56% 62% 60% n.a. 61% 67% n.a. 60% Operat. IR & Comm 6% 7% 7% n.a. 6% 5% n.a. 3% Public Debt 22% 12% 12% n.a. 11% 0% n.a. 5% Subs&Transf 16% 20% 21% n.a. 22% 29% n.a. 32% Source: MINECOFIN (in 2007 real prices). Note: 0 indicates that data was not provided in part or in full by MINECOFIN. Actual total recurrent expenditures measured in real terms increased consistently in the period. The largest portion of recurrent expenditures is the category of recurrent operations, which increased from RwF 156.5 billion in 2004 to RwF 197.8 billion in 2007. Interest and public debt decreased in the period. Subsidies and recurrent transfers increased from RwF 51.0 billion to RwF 103.8 billion between 2004 and 2007. Figures 3.17 and 3.18 are based on the data in the table above and compare recurrent expenditure trends in both projected and actual data for the period. The emerging pattern in the two graphs shows that projections for recurrent expenditure were too high in the years for which complete data are available (2004 and 2006). This seems to be true for all subcategories except subsidies and recurrent transfers and recurrent operating expenditures in 2006, for which projections were conservative. Figure 3.18 further illustrates that overall projections and actuals for the period fluctuated, with only recurrent operation expenditures and subsidies and recurrent transfers emerging with a consistently increasing trend. Interest and commissions as well as public debt levels remained low for the period. 48 nk World Ban Working Paper ecurrent Expend Figure 3.17. Re Comparing Origi diture Trends C 2004­07 inal vs. Actual 2 In million RwF FIN (in 2007 real pr Source: MINECOF rices). ecurrent Expend Figure 3.18. Re y 7 diture Trends by Type 2004­07 00 350,00 00 300,00 00 250,00 in million RwF 00 200,00 00 150,00 00 100,00 00 50,00 0 Total Recur blic Debt Subs& IR & Comm Pub &Transf Operat. 2004 2004 2005 2005 2006 2006 2007 2007 O Orig Actual Orig Actual Orig Actual Orig al Actua FIN (in 2007 real pr Source: MINECOF rices). ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 49 Total pro nt s ojected recurren expenditures decreased wh in hen measured i real terms 7, and without data for 2007 with the lar rgest increase in subsidies aand recurrent real terms, overall projected recurrent operat transfers. In r res tion expenditur increased, while both interest and publi ic debt decrease ed (see table 3.18 and figure 3.119). However, d when considering projected expenditures, overall recurr res rent expenditur decreased from RwF 313 3.9 billion in 20004 to RwF 304.5 5 billion in 2006. s Table 3.18. Original Recurrent Expenditures 2004­07 2004 2005 2006 2007 Growth % in million RwF Orig Orig Orig Orig 2004­06 Total 313,980 300,277 304,448 0 ­3% Recur Operat. 176,540 181,014 184,972 0 5% IR & Comm 18,308 19,708 17,557 0 ­4% Public Debt 68,065 36,549 34,348 0 ­50% Subs&Transf 51,067 63,006 67,571 0 32% Source: MINECO OFIN (in 2007 real prices). s that data was not Note: 0 indicates or in full by MINE t provided in part o ECOFIN. Original Recurre Expenditure Trends 2004­0 Figure 3.19 O ent 07 In million RwF OFIN (in 2007 real prices). Source: MINECO Actual re ecurrent expen nt nditures presen a slightly di ifferent picture for 2004­07 or where data fo 2007 is avail rent expenditur increased. lable. In this case actual recurr res and interest dec Public debt a creased, while s subsidies and r fers increased recurrent transf by 104 percen nt and recurrent t operating expe enses increased by 26 percent ( (see table 3.19 and figure 3.220). 50 nk World Ban Working Paper tual Recurrent Expenditures 20 Table 3.19. Act E 004­07 2004 2005 2006 2007 Growth % in million RwF Actual Actual Actual Actual 2004 - 07 Total 254,282 0 281,397 329,300 30% Recur Operat. 156,454 0 187,785 197,800 26% IR & Comm 16,762 0 12,908 10,900 ­35% Public Debt 30,065 0 0 16,800 ­44% Subs&Transf 51,001 0 80,704 103,800 104% Source: MINECOF rices). FIN (in 2007 real pr hat data was not provided in part or in full by MINECO Note: 0 indicates th OFIN. ctual Recurrent Expenditure Tr Figure 3.20. Ac rends 2004­07 In million RwF FIN (in 2007 real pr Source: MINECOF rices). The main c conclusion abou penditure is tha ut recurrent exp ions did not at while projecti consistently ma nding, there wa atch actual spen as a slight but co ase in actual onsistent increa ent expenditur driven prim overall recurre re, marily by incr h reases in both operation nd nd expenditure an subsidies an recurrent tra t er ansfers. Interest levels howeve were low and public debt decreased in thhe period. nditures Capital Expen Capital expend diture and net le uted about one ending constitu third of total ex xpenditures g the larger of t (see table 3.1), with capital expenditure being the two elemen nts. As table 3.20 shows, cap ed 6 al pital represente between 85 percent and 96 percent of tota projected capital spendinng between 2004 4 and 2007. ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 51 verview of Capital Expenditure and Net Lend Table 3.20. Ov es ding 2004­07 in million 2004 2 2004 2005 20 005 2006 2006 20007 2007 RwF Orig Actual Orig Acctual Orig Actual Orrig Actual TOTAL 125,097 1 142,445 158,424 0 141,992 144,010 0 186,200 Capital 119,981 1 112,591 134,354 0 125,011 133,234 0 186,200 Net Lend 5,116 29,854 24,070 0 16,980 10,775 0 0 % Tot Capital & 2004 2 2004 2005 20 005 2006 2006 20007 2007 Net Lend. Orig Actual Orig Acctual Orig Actual Orrig Actual TOTAL 100% 1 100% 100% n n.a. 100% 100% a. n.a 100% Capital 96% 79% 85% n n.a. 88% 93% a. n.a 100% Net Lend 4% 21% 15% n n.a. 12% 7% a. n.a 0% Source: MINECO OFIN (in 2007 real prices). s that data was not Note: 0 indicates or in full by MINE t provided in part o ECOFIN. Figures 3 trends seen in table 3.20, firs comparing 3.21 and 3.22 illustrate the t st original projeections and act res ng tual expenditur for capital and net lendin spending. Yearly pattern ns between pro ojected and actu ual total capital spending betw ween 2004 and 2006 fluctuate below RwF 160 billion. Only in 2007 was there a spike in actual total ed n capital spend 21). Similarly, t ding (figure 3.2 total spending by source (cap pital and net lending) fluct tuated between 2004 and 2006 with a spike in actual spen n nding in 2007 (figure 3.22). Capital Expenditures and Net L Figure 3.21. C aring Original vs Actual Lending Compa s. 2004­07 In million RwF OFIN (in 2007 real prices). Source: MINECO 52 nk World Ban Working Paper apital Expenditu Figure 3.22. Ca In million RwF 2004­07 ures and Net Lending Trends 2 FIN (in 2007 real pr Source: MINECOF rices). Table 3.21 and figure 3.2 examine the original (budg 23 xpenditures geted) capital ex and net lendingg for 2004­07. ginal Capital Ex Table 3.21. Orig d 2004­07 xpenditures and Net Lending 2 2004 2005 2006 2007 Growth % in million RwF Orig Orig Orig Orig 2004­06 TOTAL 125,097 158,424 141,992 0 40% Capital 119,981 134,354 125,011 0 28% Net Lend 5,116 24,070 16,980 0 309% Source: MINECOF rices). FIN (in 2007 real pr hat data was not provided in part or in full by MINECO Note: 0 indicates th OFIN. ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 53 Original Capital Expenditures a Figure 3.23. O g ­07 and Net Lending Trends 2004­ In million RwF OFIN (in 2007 real prices). Source: MINECO Similarly d figure 3.24 pre y, table 3.22 and ctual spending in the capital esent data on ac and net lending category. expenditure a ctual Capital Ex Table 3.22. Ac 004­07 xpenditure and Net Lending 20 2004 2005 2006 2007 Growth % in million RwF Actual Actual Actual Actual 2004­07 TOTAL 142,445 0 144,010 186,200 31% Capital 112,591 0 133,234 186,200 65% Net Lend 29,854 0 10,775 0 ­100% Source: MINECO OFIN (in 2007 real prices). s that data was not Note: 0 indicates or in full by MINE t provided in part o ECOFIN. 54 World Bank Working Paper Figure 3.24. Actual Capital Expenditure and Net Lending Trends 2004­07 200,000 180,000 160,000 140,000 Million RwF 120,000 100,000 80,000 60,000 40,000 20,000 0 2004 Actual 2005 Actual 2006 Actual 2007 Actual Year & Source TOTAL Capital Net Lend Source: MINECOFIN (in 2007 real prices). The main conclusion about capital expenditures and net lending is that while moderate fluctuations occurred over the period compared both to original and to actual spending in between yearoveryear, 2007 actual spending saw a significant increase compared to previous years (with no projected data available for 2007). Overview of Aid Dependency External aid plays a major role in promoting poverty reduction and economic growth, and many developing countries like Rwanda rely on external aid to finance their development. Considering that most official development aid is allocated directly to governments, understanding how it influences fiscal behavior in the public sector is important. Furthermore, a sound macroeconomic framework is often considered a prerequisite for development. The World Bank (World Bank 2007a) recommends that international assistance should be targeted at countries with high poverty levels and sound fiscal policies. A study of fiscal policy and aid in Rwanda called "Should Fiscal Policy Make Room For Aid? A Case Study of Rwanda (20002006)" (Sayinzoga 2007) produced several key findings: (a) scalingup of aid has not led to imprudent and unsustainable fiscal policy; (b) governments with a sound initial macroeconomic framework can manage the risks associated with fiscal expansion and increased aid flow; and (c) while aid can be an important component of total revenue, it does not necessarily influence budget allocations to sectors. This study provides the basis for this section of this report, which considers Rwanda in the context of fiscal space and aid sustainability. The definition of "fiscal space" used in this section refers to the availability of budgetary room that allows a government to provide resources for desired purposes without prejudice to the sustainability of a government's financial position (Heller 2005). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 55 Rwanda is highly aid dependent, with a stable macroeconomic environment and an annual per capita income of US$250 (MINECOFIN 2006). GDP growth averaged 7.4 percent per year between 1995 and 2005 as a result of economic and structural reforms. The country has a limited tax base. According to the 2nd Economic Development and Poverty Reduction Strategy (EDPRS), the following aidrelated trends are anticipated between 2008 and 2012 in Rwanda: aid disbursement doubling between 2008 and 2012 to finance the EDPRS aid decreasing as a percentage of GDP from 17 percent in 2008 to 14 percent in 2012 foreign exchange decreasing from five months to four months of imports real exchange appreciating modestly, by less than 1 percent per year. In other words, although external aid is expected to double, aid dependency is expected to fall (MINECOFIN 2007). However, in the medium term, Rwanda will remain highly dependent on aid while increasing its tax base by allocating funds toward productive sectors. Since 1994, Rwanda has seen several aid surges. However, the post2003 aid surge was more steady and prolonged and led to large fiscal expansion (IMF 2005b). A major challenge for Rwanda in 2007 was to manage aidfinanced fiscal expansion (estimated at 3 percent of GDP) while consolidating macroeconomic stability (IMF 2007). In 2007, Rwanda anticipated a substantial surge of grant aid, mainly from these programs and sources: Fast Track Education Initiative (new), USA Millennium Challenge Account (new), U.K. Catalytic Fund (new), AfDB (existing but increasing), and EU (existing but increasing). This anticipated surge of grant aid could potentially have led to inflationary pressure and crowdingout effects. Post2003 aid was more absorbed and better spent than previous aid because of better sustained and aligned government priorities. Nonetheless, the aid also posed some monetary challenges as foreign exchange was not fully absorbed (Foster & Heller 2007), leading to an unwelcome buildup of foreign exchange reserves. The unpredictability of the aid flow led MINECOFIN to increase the international reserves from five months to six months of imports in 2005. However, the biggest risk in a mismanaged expansionary fiscal program is an increase in inflation. In Rwanda, partly as a result of increases in aid, inflation fluctuated between 7.4 percent and 12 percent from 2003 to 2005. To control inflationary pressures, the government used monetary instruments such as Treasury bills to absorb excess liquidity and contain private sector demand (MINECOFIN 2007). Hence, the government did not resort to domestic debt to finance budgetary expenditure, but instead used monetary tools to control inflation. MINECOFIN also had to control the effects of fiscal expansion on the private sector. Any significant increase in aid receipts was expected to further increase the current account deficit (IMF 2002), but trying to limit the widening of the current account deficit could also have led to a squeeze on private investments and consumption. In theory, the widening in the current account deficit would be larger than the actual aid flow due to the multiplier effect. It becomes increasingly difficult to manage widening of a current account deficit while balancing balance of payments gap and fiscal gap if the real exchange rate is not allowed to fluctuate. 56 World Bank Working Paper Aidfinanced expansion presents additional challenges caused by uncertainty of aid inflows. This in turn affects credibility and realism of expenditure plans (especially in the medium term). Even though governments have direct control over aid disbursements, the complexity of donors' procedures often translate into unanticipated delays (Gupta et al. 2006). Also, few donors are able to provide realistic aid commitments over the medium term, affecting credibility of governments' MTEFs. Aid flows are very vulnerable to political events and conditionalities and can therefore be unexpectedly withdrawn at very short notice. In addition, recipient countries may face a "use it or lose it" constraint as aid is not always intertemporally fungible (MacKinnon et al. 2003). Most donors' assistance is available only for a certain time period, and postponing is often not an option. This does not always allow governments to save resources and spend them on future projects when rates of return may be higher. This constraint can distort the recipient government's fiscal behavior and lead to suboptimal expenditure programs (MacKinnon et al. 2003). It is therefore important to note that most developing countries rarely have the luxury to turn down aid because of the vast mismatch between their growth and development needs and their public financing capacity. Access to grants and borrowing is often the only way to narrow the gap despite associated budgetary uncertainty. The case for increased expenditures in Rwanda is anchored in the need to overcome the socioeconomic legacy of genocide, tackle the high level of poverty, alter the structure of economy, and provide stability through regional integration and improved governance. Concurrently, there is a consensus in the international community that additional assistance should be allocated to highperforming poor countries with sound macroeconomic management (as in Rwanda). Since 1994, the government has been committed to macroeconomic stability and gone to great lengths to design sustainable fiscal and monetary policies. The country has subsequently developed a strong reputation in the region as being high performing and dedicated to economic growth and private sector development (World Bank 2006). Despite several hurdles--high population density, high dependency on lowproductivity agriculture, scarcity of arable land, high poverty rates, low human capital, low infrastructure assets, and regional instability--Rwanda has significantly improved most socioeconomic development indicators. This suggests that the government is resultsoriented and has been and continues to be committed to PRSP/EDPRS implementation. Figure 3.25 illustrates changes in fiscal policy in 2001­06. As can be seen, the largest fiscal expansion occurred in 2004, but it was still short of PRGF projections. Total expenditure was projected to increase from 22 percent of GDP to 28 percent. In nominal terms, this was a projected increase of 42 percent compared with 2003. Despite its planned increase in expenditures, the government did not reach projected total expenditure and undershot the target by 2.2 percent of GDP, spending RwF 275 billion instead of RwF 285 billion. The gap between projections and actuals was to the result of a lack of absorptive capacity, but rather a move to maintain macroeconomic stability, allowing the government to contain domestic demand and inflationary pressures (IMF 2005b). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 57 Figure 3.25. Total Expenditures and Net Lending 2001­06 Expansion trend 450 35 400 30 350 25 300 Billion RwF 250 20 % of GDP 200 15 150 10 100 50 5 0 0 2001 2002 2003 2004 2005 2006 Total expenditures + net lending projections Total expenditures + net lending actuals Total expenditures net lending as % of GDP projected Total expenditures net lending as % of GDP actuals Sources: IMF (2002­07) and MINECOFIN macroeconomics tables (2007). In 2005 and 2006, the government pursued an even more expansionary fiscal stance by systematically outspending projections. Actual total expenditures exceeded projections by 1.5 percent in 2005 and 1.9 percent in 2006, reflecting acceleration of implementation of the PRSP and increased availability of resources. As of 2004, the PRGF introduced a contingent spending approach that was conditional on the availability of grants, ensuring funds were spent on priority areas.2 Total expenditure analysis shows that despite a tight fiscal stance prior to 2003, the government opted for fiscal expansion from 2003 on. As shown in figure 3.25, projected and actual total expenditures in nominal terms or as a percentage of GDP have dramatically increased between 2001 and 2006. Actual total expenditures and net lending as a share of GDP grew from 13 percent to 30 percent between 2001 and 2006, with the sharpest increase after 2003. This growth shows clear signs of an expansionary fiscal policy in Rwanda in the period.3 The changes in aid dependency between 2004 and 2006 were caused by shifts in aid levels (see figure 3.26). The sharp increase in 2004 reflects both delay in disbursements from 2003 and scaling up of budget support by donors. The sharp decrease in 2006 reflects the fact that total aid recorded was much lower than in 2005 (see chapter 2). IMF data indicate that the pre2003 ratio of aid to total expenditure was rather stable. The government's Budget Department data indicate that the pre2003 ratio varied some, with aid dropping to 47 percent of total expenditures in 2001 (compared with 65 percent by IMF data for that year). Analysis of aid as a percentage of GDP confirms that the pre2003 ratio of aid to GDP was highly stable, around 12 percent for IMF data and fluctuating between 10 percent and 12 percent for the Budget Department data. In 2004, aid rose to 70 percent of total expenditures before progressively decreasing to 55 percent in 2006. 58 World Bank Working Paper Figure 3.26. Total Aid as a Percentage of Total Expenditure and a Percentage of GDP 2000­06 Aid dependency trend 80 20 70 18 16 60 % of tot. expenditures 14 50 12 % of GDP 40 10 30 8 6 20 4 10 2 0 0 2000 2001 2002 2003 2004 2005 2006 Aid as % of total expenditures (IMF) Aid as % of total expenditures (MINECOFIN) Total aid as % of GDP (IMF) Total aid as % of GDP (MINECOFIN) Sources: IMF (200207) and MINECOFIN Flash Reports (200206). Figure 3.27 depicts the 2001­06 trend in domestic revenues, with and without grants, as a percentage of GDP. Figure 3.28 shows the 2001­06 trend in total revenues and expenditures. The two graphs demonstrate that the post2003 surge in aid is correlated with Rwanda's fiscal expansion, suggesting that the fiscal expansion was largely financed by international aid. Figure 3.27. Domestic Revenue as a Percentage of GDP 2001­06 250 16 14 200 12 10 Billion RwF 150 % of GDP 8 100 6 4 50 2 0 0 2001 2002 2003 2004 2005 2006 Total domestic revenue projections Total domestic revenue actuals Revenue without grants as %GDP projections Revenue without grants as %GDP actuals Sources: Sources: IMF (200207) and MINECOFIN macroeconomics tables (2007). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 59 Figure 3.28. Trends in Total Revenue and Grants, and Total Expenditure and Net Lending 2001­06 450 400 350 300 250 Billion RwF 200 150 100 50 0 2001 2002 2003 2004 2005 2006 Total revenue and grants projections Total revenue and grants actuals Total expenditures and net lending projections Total expenditures and net lending actuals Sources: IMF (200207) and MINECOFIN macroeconomics tables (2007). Aid became a larger part of GDP from 2004 onward, reaching an unprecedented high level of over 16 in 2005 before declining to below 15 percent by 2006. The "perfect match" between IMF and Budget Department data for aid as percentage of GDP as of 2004 confirms that discrepancies in data in previous years are related to differences in previous data. Unlike many other developing countries, the concerns surrounding aid dependency in Rwanda are not based on a possible large drop in aid receipts but on the cost of aid (MacKinnon et al. 2003), meaning the heavy burden of implementing donor financed projects, the administrative costs of aid management, the constraints of political conditionalities, and the lack of predictability in timing of aid disbursements. The most important aspect of aid management in an aiddependent country is its sustainability. Sustainability is even more critical if aid is delivered through loans. Debt relief initiatives reflect the general concern that developing countries may never pay off their external debt even with high economic growth rates. Whether aid is awarded in the form of grants or loans has major implications for macroeconomic stability. As for the composition of expenditures, government faces a tradeoff between spending in the current period on "donorapproved" projects as opposed to spending "freely" in the future (Heller 2005). Moreover, donors' aid allocations to recipient countries have ceilings on the amount of financing available. Prior to debt relief, Rwanda's debt stocks were increasingly high. In 2003, of the total nominal debt, 88 percent was owed to multilateral creditors, with the International Development Association (IDA), the largest creditor, holding 58 percent of the total outstanding debt. Bilateral debt amounted to US$183.7 million, with US$88.9 million due to Paris Club creditors. With the exception of US$26.2 million, the entire stock of outstanding debt was contracted through ODA (IMF 2005a). 60 World Bank Working Paper Rwanda has since benefited from extensive debt relief. Prior to 2003, Rwanda received marginal debt cancellations form AfDB, the Arab Bank for Economic Development in Africa (BADEA), the Organization of the Petroleum Exporting Countries (OPEC) Fund, the EC, and Paris Club creditors. More important, in April 2005, the government managed to fulfill the stringent requirement needed to reach the HIPC completion point and thus qualified for a total of US$1.4 billion in nominal terms, including US$243.1 million "topping up" (World Bank and IMF 2006). One of the key objectives of the HIPC program was to reduce the net present value (NPV) of debttoexport ratio permanently below 150 percent. Total HIPC debt relief was equivalent to a reduction in NPV terms4 of US$452.4 million. In January 2006, Rwanda was among a select group of 19 countries to qualify for additional debt cancellation through the MDRI (IMF 2007b). Following the HIPC and MDRI, Rwanda's total external debt decreased to an estimated value of US$422 million, equal to an NPV of debttoexport ratio of 56 percent. Analyzing trends in the NPV of debttoexport ratio is central to debt sustainability analysis. A reduction in the ratio reflects resources that can be freed from repayment and interest to finance poverty reduction programs. This can, in turn, determine whether additional borrowing will lead to excessive net payment outflows in future periods. This is one of the key indicators often used by investors to assess macroeconomic stability of an economy. Therefore, a high NPV of debttoexport ratio can act as a deterrent to foreign investors as it is perceived as a lack of commitment by government to a sound macroeconomic program. Nevertheless, the NPV of debtto export ratio also has several shortcomings as an indicator (MacKinnon et al. 2003). Using export as a denominator means that countries with low levels of trade, especially export, will benefit from more debt relief than others. It also implies that they will be more constrained on new lending. Additionally, in countries like Rwanda, which are landlocked and informal trading is common across borders, export figures are likely to be underestimated. On other hand, the debttoexport ratio provides a strong incentive for the government to improve export performance and have access to additional borrowings in the future. In the case of Rwanda, one major weakness of the NPV of debt is that it is influenced by the fluctuations of world interest rates. This means that the NPV can deteriorate even if there has been no significant change in the country's debt stock/flows and export volumes/values. In the past, the fall in world interest rates, which is exogenous, played a substantial role in the deterioration of Rwanda's debt burden indicators (IMF 2005a). Prior to 2004, the NPV of debttoexport ratio was worsening compared with projections (figure 3.29). In 2002, the NPV of debttoexport ratio exceeded projections by 80 percent and rose to 270 percent; the gap widened further in 2003 as the ratio reached 286 percent as opposed to 204 percent originally projected. This was primarily attributable to fundamental changes in the country's economic circumstances resulting from exogenous factors rather than imprudent borrowing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 61 Figure 3.29. Net Present Value of Debt-to-Export Ratio 2001­06 400 350 Threshold 300 250 Percent 200 150 100 50 0 2001 2002 2003 2004 2005 2006 NPV of debt/exports projected NPV of debt/exports actuals Sources: Sources: IMF (200207) and MINECOFIN macroeconomics tables (2007). Indeed, lower export prices, changes in exchange rates, and lowerthanexpected "consessionality" of new borrowing were all unambiguously exogenous and outside the control of government. Lower discount rates were another fundamental factor causing further deterioration of Rwanda's NPV of debttoexport ratio. Also, the unexpected increase of import prices had a lasting negative impact on Rwanda's terms of trade and, consequently, its capacity to repay debt. According to the IMF (2005a), all these factors accounted for more than half of both the unanticipated and the total increase in the NPV of debttoexport ratio. Following the inaccuracy of previous projections and in light of the upcoming HIPC completion point, the IMF and MINECOFIN revised projections downward. According to the Director of Treasury5 (2007) in MINECOFIN, a further deterioration of terms of trade and export performance was anticipated. However, the combination of an aggressive government exportpromotion strategy and debt relief from many creditors (excluding the Bretton Woods Institutions) paid off. Subsequently, in 2004, the NPV of debttoexport ratio improved significantly. Improvements in the NPV of debttoexport ratio in 2005 and 2006 reflect the substantial contribution of the HIPC Initiative and the MDRI as well as favorable export developments. For the first time, the ratio fell to the recommended HIPC threshold (150 percent) in 2005 and to 56 percent in 2006. According to the IMF, the MDRI contributed 76 percentage points to the reduction, while higher exports led to further improvements by 12 percentage points. Export performance was also strong as actual merchandise and service exports in 2005 exceeded projections by 20 percent, largely due to strong export performance in coffee, tea, and minerals (World Bank and IMF 2006). Despite being central to debt sustainability, capping of the NPV of debttoexport ratios could lead to negative net effects. Although a cap limits outflows paid for interest and principal on borrowing, it also constrains the recipient country on future borrowing. Yet these borrowings are used to finance investments necessary to attain poverty reduction and growth targets. In the case of Rwanda, since a large proportion of debt is composed of concessional loans owed to multilateral institutions, this implies that the net effect may actually be negative. 62 World Bank Working Paper The joint IMF and World Bank debt sustainability analysis in 2006 concluded that in the medium term, Rwanda must rely on grant financing. Even though the MDRI substantially improved Rwanda's debt indicators, the country would have to rely mostly on grants to maintain its debt at sustainable levels. With an NPV of debtto export ratio at 55.9 percent in 2006 and barring any exogenous shocks or policy reversals, debtservice payments could remain manageable at below 8 percent of exports over the projection period until 2026. However, the NPV of debttoexport ratio would breach the policy dependent threshold of 150 percent by 2014, indicating that the country would be at a high risk of debt distress beyond the projection period (World Bank and IMF 2006). Because of Rwanda's narrow tax base and constraints on future external borrowing, it is important to examine financing trends (see figures 3.30 and 3.31). Availability of grants is clearly a central feature of fiscal expansion in Rwanda. Figure 3.30. Total Aid, Broken Down by Loans and Grants 2000­06 250 200 Billion RwF 150 100 50 0 2000 2001 2002 2003 2004 2005 2006 Total Aid Grants Loans Source: MINECOFIN Flash Reports (200206). Figure 3.31. Loans and Grants as Share of Total Aid 2000­06 100 90 80 70 60 Percent 50 40 30 20 10 0 2000 2001 2002 2003 2004 2005 2006 Grants as % of total aid Loans as % of total aid Source: MINECOFIN Flash Reports (200206). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 63 Loans as a share of total aid increased steadily from 2000 to 2004, with the exception of 2003. There was a sudden decline in loans as a share of total aid from 28 percent in 2002 to 13 percent in 2003, followed by a sharp increase to 33 percent in 2004. The 2004 increase was prompted by a disbursement of additional budget support from the World Bank and the AfDB (partially loan financed). Implementation of a largescale road construction program financed through loans by BADEA and other Arab donors also contributed to high levels of loans in 2004. From 2004 onward, loans as share of total aid decreased progressively. Grants as a share of total aid declined progressively from a peak of 92 percent in 2000 to 82 percent in 2001 and 72 percent in 2002. Following a oneyear increase in 2003 and a oneyear decrease in 2004, nominal analysis (see figure 3.31) confirms that the scaling up of total aid inflows post2004 was driven mainly by grants. It is clear that grants have always constituted a significant share of total aid, but increasingly so from 2004. The decline of loans and increase of grants post2004 indicate that the government and donors responded to the conclusions of the debt sustainability analysis, which recommended capping Rwanda's nonconcessional borrowing6 to a maximum of US$20 million a year (World Bank and IMF 2006). Following this recommendation, and as the nominal analysis indicates, Rwanda's largest donors-- AfDB, EC, World Bank, and DFID--all started to provide 100 percent grant aid in 2005.7 Where did the money go? Chapters 4 and 5 will take a closer look at where the money was spent since the surge of aid in 2004. Notes 1 Should Fiscal Policy Make Space for Aid? A Case Study of Rwanda (2000­2006) by Kampeta Sayinzoga, University of Nottingham, 2007 is based primarily on IMF data and government of Rwanda data. 2 Contingency expenditures not included in projections can also explain differences between projections and actual, as shown in figure 3.25. 3 Further details of the total expenditures for the period 2004­07 follow in chapters 4 and 5. 4 The NPV of debt is the discounted sum of all future debt service obligations (interest and principal). It is a measure that takes into account the degree of concessionality of a countrys debt stock. Whenever the interest rate on a loan is lower than the prevailing market rate, the resulting NPV of debt is smaller than its face value, with the difference reflecting the grant element. 5 Mr. François Nkulikiyimfura was the Director of Treasury in 2007. 6 Borrowing that does not have a grant element of at least 50 percent. 7 Note that following the recent debt distress reclassification in July 2009, World Bank International Development Association repayment terms are now on a 50 percent grant50 percent loan basis. Source: "World Bank Operational Manual, Operational Policies OP 3.10-- Annex D," July 2009. CHAPTER 4 Public Expenditure Overview by Ministry Level I n this chapter total public expenditures will be considered by looking at both recurrent and capital (development) spending and how this spending is allocated across different line ministries.1 Line ministries are classified based on Rwanda's budget law and may be different from the classification of sectors identified by other sources, such as IMF, and the sectors reviewed in chapter 5 of this report. Furthermore, central budget spending allocated to each province/district was classified based on assumptions of sectorspecific spending and included in the overall ministry analysis when appropriate.2 Specific assumptions of these province/district programspending allocations are illustrated in Data Sets 6­9. Similarly, development budget data are based on ministry allocations contained in the budget law. However, the budget law does not include any district development budget. For easy comparison with chapter 5, which focuses on a more comprehensive analysis of sectorspecific spending in the so called priority sectors (education; health; infrastructure, including water and sanitation, energy and transport; social protection; and agriculture), this chapter will be broken down by (a) ministries directly contributing to these priority sectors being reviewed in chapter 5, and then (b) remaining ministries. Each of the recurrent, district/provinces, and development budget trends will be analyzed in detail in the sections below. Social protection, not a separate line ministry according to the budget law, is considered separately under chapter 5 and is not covered here. (However, a large portion of social protection spending falls under MINALOC, which is covered.) Overview of Recurrent and Development Expenditure by Line Ministry In Rwanda, the level of actual public expenditure considered in nominal terms has increased significantly in years 2004­07 as illustrated in table 4.1. Recurrent expenditures have increased in nominal terms from RwF 208.9 billion in 2004 to RwF 356.0 billion in 2007, while nominal development expenditures have increased from RwF 79.9 billion to RwF 181.9 billion in the same period, all according to government reporting. This represents a 70.4 percent growth rate in recurrent expenditures and a 127.6 percent growth rate in development expenditures. Overall, this reflects a total nominal public expenditure increase of 86.2 percent, from RwF 288.9 billion in 2004 to RwF 537.9 billion in 2007. 64 Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 65 Table 4.1. Total Actual Expenditures in Nominal Terms in RwF millions 2004­07 Growth % 2004 2005 2006 2007 2004­07 Recurrent 208,988 240,270 281,540 356,025 70.4% Capital 79,920 120,610 151,691 181,870 127.6% TOTAL 288,908 360,880 433,230 537,894 86.2% Source: MINECOFIN. Considering the level of actual public expenditure in Rwanda over the same period in real terms (2007 prices), one can see that expenditure has increased overall, but at a less significant rate (see table 4.2 and figure 4.1). Real recurrent expenditures have increased from RwF 288.9 billion in 2004 to RwF 356.0 billion in 2007, while real development expenditures have increased from RwF 110.5 billion to RwF 181.9 billion in the same period, all according to government reporting. This represents a 23.2 percent growth rate in recurrent expenditures measured in 2007 prices and a 64.6 percent growth rate in development expenditures. Overall, this reflects a increase of 34.6 percent (compared with 86 percent when measured in nominal terms), from RwF 399.5 billion total expenditures in 2004 to RwF 537.9 billion in 2007. Table 4.2. Total Actual Expenditures in Real Terms in RwF millions 2004­07 Growth % 2004 2005 2006 2007 2004­07 Recurrent 288,972 305,057 316,013 356,025 23.2% Capital 110,508 153,131 170,265 181,870 64.6% TOTAL 399,480 458,188 486,278 537,894 34.6% Source: MINECOFIN. Figure 4.1. Total Actual Expenditure Trend in Real Terms 2004-2007 600,000 500,000 400,000 Millions RwF 300,000 200,000 100,000 0 2004 2005 2006 2007 Year Recurrent Capital Source: MINECOFIN (in 2007 real prices). 66 World Bank Working Paper The increase in overall public expenditure is due to several factors. First, the increase reflects an increase in actual domestic resources (measured in real terms) from RwF 191.8 billion to RwF 257.7 billion3 in the period. Second, there was a significant reduction in debt services, linked primarily to the debt relief in 2005 and 2006. Third, Rwanda has also experienced an increase in actual aid flow (grants) from RwF 128.7 billion to RwF 207.3 billion4 (in nominal terms) over the same period, in particular as it relates to budget support. Figure 4.2 compares overall actual and budgeted public expenditures in 2004­07. Figure 4.2. Comparison of Actual and Revised Budget Expenditures 2004­07 600,000 500,000 400,000 Millions RwF 300,000 200,000 100,000 0 Recurrent Recurrent Capital Capital Total Actual Total Actual Budget Actual Budget Budget (Revised) (Revised) (Revised) Type 2004 2005 2006 2007 Source: MINECOFIN (in 2007 real prices). As figure 4.2 shows, 2004 and 2005 actual recurrent, capital, and total public expenditure were less than budgeted, with the exception of the 2005 capital budget where actual capital expenditures exceeded the budget. Overall, the government kept its actual overall public expenditures close to its budget allocations. Between 2004 and 2007, as overall recurrent, development, and total public expenditures measured in nominal terms increased by 70 percent, 127 percent, and 86 percent, respectively, overall recurrent expenditure in real terms increased by only 23 percent and development expenditure increased by 64 percent, resulting in an overall increase of 34 percent. As table 4.3 illustrates, key ministries experienced variations in growth as measured in real terms. Overall public expenditures increased in the period in real terms from RwF 399.5 billion to RwF 537.9 billion, an RwF 138.4 billion increase. All ministries listed in the table experienced significant increases in their respective recurrent and development budget spending. Overall, all ministries experienced an overall increase in their budgets, with land and resources showing the highest overall increase in budget allocations. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 67 Table 4.3. Increase in Actual Expenditure in Real Terms in RwF millions 2004­07 Growth in RwF Growth % REAL Terms 2004 2005 2006 2007 2004­07 2004­07 Overall Recurrent 288,972 305,057 316,013 356,025 67,053 23% Development 110,508 153,131 170,265 181,870 71,362 65% Total 399,480 458,188 486,278 537,894 138,415 35% Ministry--Recurrent Education 54,151 57,726 63,708 75,423 21,272 39% Health 16,224 18,179 20,256 24,124 7,899 49% Agriculture 5,308 6,784 3,846 6,385 1,076 20% Infrastructure incl. Energy 7,688 15,738 14,139 26,861 19,173 249% Land & Resources 1,214 3,420 3,682 3,087 1,873 154% Ministry--Development Education 5,854 12,653 11,864 19,862 14,008 239% Health 9,052 19,238 49,694 25,684 16,632 184% Agriculture 9,977 14,468 16,517 13,517 3,540 35% Infrastructure incl. Energy 30,712 47,098 32,050 52,125 21,413 70% Land & Resources (incl. WSS) 5,157 14,767 20,214 25,903 20,746 402% Ministry--Total Education 60,005 70,379 75,572 95,285 35,280 59% Health 25,276 37,417 69,950 49,808 24,531 97% Agriculture 15,285 21,252 20,363 19,902 4,616 30% Infrastructure incl. Energy 38,400 62,836 46,189 78,986 40,586 106% Land & Resources (incl. WSS) 6,371 18,187 23,896 28,990 22,619 355% Source: MINECOFIN (in 2007 real prices). More specifically, as table 4.3 shows, education recurrent expenditures increased 39 percent, development expenditures increased 239 percent, representing an overall increase of 59 percent. Health experienced a similar growth in recurrent expenditures at 49 percent and a higher increase in development expenditures at 184 percent, which translates into an overall growth in health expenditures of 97 percent. Agriculture experienced a 20 percent increase in recurrent expenditures, but its development expenditures increased only 35 percent, for an overall increase in expenditures in the agriculture sector of 30 percent. All ministries experienced increases in their total expenditures, with agriculture showing the least growth in the period, only 30 percent. The most significant increase in both development and recurrent expenditure occurred in the land and natural resource sector, which includes water and sanitation. This ministry grew recurrent and development spending by 355 percent during the period. As table 4.4 shows, of the overall increase in public expenditures measured in nominal terms in 2004­07, the largest share of the increase for recurrent expenditures went to the education ministry, with a 25 percent share, and infrastructure captured the highest share of the increase in development spending, with 14 percent. 68 World Bank Working Paper Table 4.4. Allocation to Ministries of Increase in Actual Public Expenditure 2004­07 % of sectors' share of % of sectors' share of % of sectors' share of total increase in total increase in total increase in total Nominal Terms recurrent 2004-07 developm. 2004-07 2004-07 Education 25% 15% 21% Health 8% 19% 13% Agriculture 2% 6% 4% Infrastructure (incl. Energy) 14% 29% 21% Land & Natural Resources (incl WSS) 2% 22% 10% TOTAL Ministry' Share 51% 91% 67% Source: MINECOFIN. Note: % calculated based on sector increases in public expenditures as % of tot overall increase in public expenditure. Infrastructure and health captured the highest percentage of the overall public expenditure increase in the period, with 21 percent, and agriculture saw the lowest overall allocation with only 4 percent of the total increase. About 67 percent of the government's increase in public expenditure as measured in nominal terms in the period was allocated to the five ministries listed in the table. Trends in Recurrent and Development Expenditure With the addition of the Ministry of Science, Technology and Scientific Research (MINISTR) in 2007 and the Office of the Prosecutor General (PARQUET GENERAL) in 2006, the government of Rwanda reached a total of 20 ministries in 2004­07. However, the Ministry of Gender and Family Promotion (MIGEPROF) was discontinued as an independent ministry after 2005 and became a part of the Office of the Prime Minister (PRIMATURE). This section will review the resource allocations to the various ministries and the spending trends for 2004­07. Trends in Recurrent Expenditure Overall Recurrent Expenditure Trends Measured in nominal terms between 2004 and 2007 and excluding transfers to districts, the Ministry of Infrastructure (MININFRA) saw the largest percentage increase in recurrent spending, at 390 percent, followed by the Ministry of Public Service and Labor (MIFOTRA) with an increase of 217 percent and the Ministry of Lands, Environment, Forestry, Water, and Natural Resources (MINITERRE) with an increase of 214 percent. A full overview of all ministries' actual recurrent expenditures in nominal terms and each ministry's recurrent expenditure increase in the period are listed in table 4.5. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 69 Table 4.5. Total Actual Recurrent Expenditure by Ministry in Nominal Terms 2004­07 NOMINAL TERMS Growth % (in million RwF) 2004 2005 2006 2007 2004-07 01 - PARLEMENT 4,098 3,962 4,706 5,234 28% 02 - PRESIREP 6,132 7,257 7,985 8,649 41% 04 - PRIMATURE 1,805 2,070 2,747 3,307 83% 05 - COURS.SUP 2,476 2,273 3,252 3,709 50% 06 - MINADEF 25,498 35,315 41,990 32,924 29% 07 - MININTER 5,883 6,891 8,394 9,292 58% 08 - MINAFFET 3,625 4,662 5,489 6,906 91% 09 - MINAGRI 3,137 4,732 3,426 4,229 35% 10 - MINICOM 1,954 3,697 4,626 3,739 91% 11- MINISTR -- -- -- 1,210 n.a. 12 - MINECOFIN 74,373 60,310 72,796 98,875 33% 13 - MINIJUST 2,425 6,584 4,039 6,812 181% 14 - MINEDUC 21,186 25,902 34,126 35,626 68% 15 - MIJESPOC 960 1,457 1,527 1,773 85% 16 - MINISANTE 7,747 9,674 11,989 13,957 80% 17 - PARQUET GENRERAL -- -- 1,715 1,832 n.a. 18 - MININFRA 5,477 12,338 12,597 26,830 390% 19 - MIGEPROF 419 494 -- -- n.a. 20 - MIFOTRA 699 2,519 2,232 2,219 217% 22 - MINITERRE 815 2,613 3,280 2,556 214% 23 - MINALOC 10,735 14,694 19,275 17,291 61% TOTAL: 179,444 207,444 246,190 286,972 60% Source: MINECOFIN (in 2007 real prices). Measured in real terms between 2004 and 2007 (see table 4.6), MININFRA saw an increase of 254 percent in recurrent spending, followed by MIFOTRA with an increase of 130 percent and MINITERRE with an increase of 127 percent. The only other ministry with a tripledigit percentage growth was the Ministry of Justice (MINIJUST) with a growth increase in real recurrent expenditures of 103 percent. A slight decrease in real recurrent expenditures occurred in the Office of the President (PRESIREP), the Ministry of Defense (MINADEF), the Ministry of Agriculture (MINAGRI), and the Ministry of Finance and Economic Planning (MINECOFIN), although these were singledigit decreases. The Ministry of Education (MINEDUC) and the Ministry of Health (MINISANTE) saw their centrallevel real recurrent expenditure in the period increase by 22 percent and 30 percent, respectively. Noteworthy is that MINAGRI, identified as a significant contributor to GDP, experienced a decrease in recurrent expenditures in the period. Overall, total actual recurrent expenditure, in real terms, increased in the period by 16 percent. 70 World Bank Working Paper Table 4.6. Total Actual Recurrent Expenditure by Ministry in Real Terms 2004­07 REAL TERMS Growth % (in million RwF) 2004 2005 2006 2007 2004-07 01 - PARLEMENT 5,666 5,030 5,282 5,234 ­8% 02 - PRESIREP 8,478 9,214 8,963 8,649 2% 04 - PRIMATURE 2,496 2,628 3,083 3,307 33% 05 - COURS.SUP 3,424 2,885 3,651 3,709 8% 06 - MINADEF 35,257 44,837 47,131 32,924 ­7% 07 - MININTER 8,134 8,749 9,422 9,292 14% 08 - MINAFFET 5,013 5,919 6,161 6,906 38% 09 - MINAGRI 4,338 6,008 3,846 4,229 ­2% 10 - MINICOM 2,702 4,693 5,192 3,739 38% 11- MINISTR -- -- -- 1,210 n.a. 12 - MINECOFIN 102,837 76,572 81,709 98,875 ­4% 13 - MINIJUST 3,353 8,360 4,533 6,812 103% 14 - MINEDUC 29,295 32,887 38,305 35,626 22% 15 - MIJESPOC 1,327 1,850 1,714 1,773 34% 16 - MINISANTE 10,713 12,282 13,457 13,957 30% 17 - PARQUET GENRERAL -- -- 1,925 1,832 n.a. 18 - MININFRA 7,574 15,664 14,139 26,830 254% 19 - MIGEPROF 579 628 -- -- n.a. 20 - MIFOTRA 966 3,198 2,505 2,219 130% 22 - MINITERRE 1,127 3,318 3,682 2,556 127% 23 - MINALOC 14,844 18,656 21,635 17,291 16% TOTAL: 248,121 263,380 276,335 286,972 16% Source: MINECOFIN (in 2007 real prices). MINECOFIN and MINADEF are the largest ministries in terms of actual recurrent spending, together accounting for 56 percent, 46 percent, 47 percent, and 46 percent of total centrallevel recurrent expenditures in 2004, 2005, 2006, and 2007, respectively. Table 4.7 summarizes the two ministries' actual recurrent expenditures in real terms and their share of total actual centrallevel expenditures. Table 4.7. Share of Actual Recurrent Expenditures of the Two Largest Ministries (Central Level Only) 2004­07 In million RwF MINECOFIN & MINADEF 2004 138,094 Share of 2 lartest ministries of Total RT 56% MINECOFIN & MINADEF 2005 121,410 Share of 2 lartest ministries of Total RT 46% MINECOFIN & MINADEF 2006 128,840 Share of 2 lartest ministries of Total RT 47% MINECOFIN & MINADEF 2007 131,800 Share of 2 lartest ministries of Total RT 46% Source: MINECOFIN (in 2007 real prices). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 71 The four largest ministries based on actual recurrent expenditure are MINECOFIN, MINADEF, MINEDUC, and the Ministry of Local Government, Good Governance, Community Development, and Social Affairs (MINALOC) when considering the entire period between 2004 and 2007. In 2004, only these ministries had actual recurrent expenditures above RwF 11 billion, and in 2007, only these ministries had recurrent spending over RwF 17 billion with the exception of MININFRA that actually ranked fourth, before MINALOC, with RwF 26.8 billion in 2007. Combined, the largest four ministries (excluding MININFRA in 2007) executed 73 percent, 66 percent, 68 percent, and 64 percent of total actual centrallevel recurrent expenditures in 2004, 2005, 2006, and 2007, respectively (see table 4.8). Noteworthy is that only one ministry contributing to a priority sector, MINEDUC, is included in this category throughout the entire period, while MININFRA and MINISANTE had recurrent spending above RwF 13 billion in 2007, ranking fourth and sixth among all ministries that year. Also, in 2007, MINEDUC exceeded MINADEF in terms of recurrent spending with RwF 35.6 billion compared with MINADEF's RwF 32.9 billion. Table 4.8. Four Largest Ministries by Actual Recurrent Expenditure at Central Level 2004­07 REAL Terms (in million RwF) 2004 2005 2006 2007 12 - MINECOFIN 102,837 76,572 81,709 98,875 06 - MINADEF 35,257 44,837 47,131 32,924 14 - MINEDUC 29,295 32,887 38,305 35,626 23 - MINALOC 14,844 18,656 21,635 17,291 TOT 4 Largest Ministries 182,232 172,953 188,780 184,716 TOT ALL MINISTRIES 248,121 263,380 276,335 286,972 % Share of 4 largest of Total 73% 66% 68% 64% Source: MINECOFIN (in 2007 real prices). Figure 4.3. Four Largest Ministries' Actual Recurrent Expenditure Trends 2004­07 120,000 100,000 80,000 Million RwF 60,000 40,000 20,000 0 2004 2005 2006 2007 Year MINECOFIN-12 MINADEF-06 MINEDUC-14 MINALOC-23 Source: MINECOFIN (in 2007 real prices). 72 World Bank Working Paper Table 4.9 further breaks down the four ministries' share of total centrallevel current expenditures. As can be seen, MINECOFIN decreased its share of total actual centrallevel recurrent spending from 41 percent in 2004 to 34 percent in 2007, and MINADEF decreased its recurrent spending share from 14 percent to 11 percent in the period. No change is seen in the portion of total recurrent expenditures going to MINEDUC and MINALOC, both spending a consistent portion of recurrent expenditures in the range of 12­14 percent and 6­8 percent, respectively. Table 4.9. Four Largest Ministries' Share of Central-Level Recurrent Actual Expenditure REAL Terms 2004 2005 2006 2007 12 - MINECOFIN 41% 29% 30% 34% 06 - MINADEF 14% 17% 17% 11% 14 -MINEDUC 12% 12% 14% 12% 23 -MINALOC 6% 7% 8% 6% TOTAL % of 4 Largerst of TOT RT 73% 66% 68% 64% Remaining recurr. % for all other ministries 27% 34% 32% 36% Source: MINECOFIN (in 2007 real prices). Table 4.10. Total Recurrent Actual Central-Level Expenditures by Ministries 2004­07 in million RwF 2004 2005 2006 2007 Growth % 04­07 18 -MININFRA 7,574 15,664 14,139 26,830 254% 16 - MINISANTE 10,713 12,282 13,457 13,957 30% 07 - MININTER 8,134 8,749 9,422 9,292 14% 02 - PRESIREP 8,478 9,214 8,963 8,649 2% 08 - MINAFFET 5,013 5,919 6,161 6,906 38% 01 - PARLEMENT 5,666 5,030 5,282 5,234 ­8% 10 - MINICOM 2,702 4,693 5,192 3,739 38% 13 - MINIJUST 3,353 8,360 4,533 6,812 103% 09 - MINAGRI 4,338 6,008 3,846 4,229 ­2% 22 - MINITERRE 1,127 3,318 3,682 2,556 127% 05 - COURS.SUP 3,424 2,885 3,651 3,709 8% 04 - PRIMATURE 2,496 2,628 3,083 3,307 33% 20 - MIFOTRA 966 3,198 2,505 2,219 130% 15 - MIJESPOC 1,327 1,850 1,714 1,773 34% 17 - PARQUET GENRERAL -- -- 1,925 1,832 n.a. 19 - MIGEPROF 579 628 -- -- n.a. 11- MINISTR -- -- -- 1,210 n.a. TOTAL All Other Ministries 65,888 90,427 87,555 102,255 55% Source: MINECOFIN (in 2007 real prices). Table 4.10 shows actual recurrent expenditures, in real terms, for the ministries excluding the four largest (where MININFRA only in 2007 reached recurrent expenditurelevels above those of MINALOC). For these sectors, MININFRA and MINISANTE were the largest in 2007, followed by MININTER and PRESIREP, the only other ministries with actual recurrent spending in excess of RwF 7.0 billion in the 2004­ Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 73 07 period. Among the ministries that have existed in all four years (that is, excluding new and ceased ministries), the smallest ministries in terms of recurrent spending (less than RwF 3.2 billion throughout the period) were the Ministry of Youth, Sports, and Culture (MIJESPOC) and MIFOTRA, although MIFOTRA saw one of the largest increases in recurrent spending at 130 percent. While MINEDUC, MININFRA, and MINISANTE rank 3rd, 5th (4th in 2007) and 6th (5th in 2004) among all sectors, the two other major ministries contributing significantly to the priority sectors, MINAGRI and MINITERRE, rank only 13th and 14th with less recurrent centrallevel spending than PRESIREP and the Ministry of Foreign Affairs and Cooperation (MINAFFET), for example. While these two latter sectors grew their recurrent expenditures by 2 percent and 38 percent, respectively, in the period, MINAGRI's recurrent expenditures decreased by 2 percent in the period. At the same time, MINITERRE saw an increase in recurrent spending of 127 percent, the highest increase after MININFRA. Yet in absolute terms, MINITERRE's centrallevel spending was still relatively small, less than half of the recurrent centrallevel spending of any of the eight largest ministries, including PRESIREP. While recurrent centrallevel spending does not consider transfers to districts, and not considering the fact that some ministries have large development budgets, it is still noteworthy that two key ministries--MINAGRI and MINITERRE--still lag behind many other sectors in both absolute terms and in terms of growth at the central level. A closer look at transfers and development allocations in later chapters of this report will put these ministries' overall spending in further perspective. Trends by Economic Classification Another interesting recurrent expenditure breakdown is the allocation of spending based on economic classification. This breakdown is done annually by MINECOFIN.5 In this section, only recurrent expenditure is included, as capital expenditures are discussed in a different section of this chapter. Overall, the breakdown of public expenditures by economic classification changed slightly in the period (see table 4.11). Wages and salaries went from 23.2 percent of total recurrent public expenditure in 2004 to 22.9 percent in 2007. Goods and services Table 4.11. Percentage of Total Recurrent Actual Public Expenditure by Economic Classification 2004­07 2004 2005 2006 2007 (1) Wages & Salaries 23.2% 21.3% 22.1% 22.9% (2) Goods & Services 22.2% 26.6% 25.4% 24.2% (3) Transfers 17.6% 22.1% 25.8% 32.4% (4)Exceptional Expenditures 8.7% 14.7% 11.9% 14.7% (5) Debt 19.1% 11.8% 7.6% 0.0% (6) Interest 0.0% 0.0% 4.2% 5.8% (7)Arrears 9.1% 3.5% 3.0% 0.0% (8) Capital n.a. n.a. n.a. n.a. Source: MINECOFIN (in 2007 real prices). Note: Data for debt and arrears in 2007 was not provided by MINECOFIN and, hence, affects the trend in the period. 74 World Bank Working Paper fluctuated between 22.2 percent and 24.2 percent. A less steady trend was seen in transfers for the period, from 17.6 percent in 2004 to 32.4 percent in 2007, a reflection of the ongoing decentralization efforts. Exceptional expenditures fluctuated in the period between 8.7 percent and 14.7 percent. A more significant trend can be observed in debt, which declined significantly and consistently over the period from 19.1 percent of the overall recurrent expenditure to only 7.6 percent in 2006 (2007 data not reported), primarily linked to debt relief in 2005 and 2006. Recurrent Expenditure Trends by Line Ministry Of the five priority line ministries, all but MINAGRI have benefited from the overall increase in resources (see table 4.12). This can be seen in real terms in recurrent public expenditures for education, health, infrastructure, and land and natural resources, which increased (on a central level) by RwF 6.3 billion, RwF 3.2 billion, RwF 19.3 billion, and RwF 1.4 billion, respectively. Table 4.12. Recurrent Central Only Executed Public Expenditure Trend Growth RwF 2004-07 % of Total Recurr* % of Total Recurr* % of Total Recurr* % of Total Recurr* Growth % 2004-07 Execution % 2004 Execution % 2005 Execution % 2006 Execution % 2007 REAL 2004 2005 2006 2007 Terms MINEDUC 29,295 13% 104% 32,887 12% 100% 38,305 13% 98% 35,626 11% 95% 6,331 22% MINISANTE 10,713 5% 94% 12,282 5% 92% 13,457 5% 100% 13,957 4% 100% 3,245 30% MINAGRI 4,338 2% 93% 6,008 2% 101% 3,846 1% 79% 4,229 1% 95% ­108 ­2% MININFRA 7,574 3% 75% 15,664 6% 84% 14,139 5% 100% 26,830 8% 104% 19,256 254% MINITERRE 1,127 0% 65% 3,318 1% 60% 3,682 1% 79% 2,556 1% 86% 1,429 127% Source: MINECOFIN (in 2007 real prices). *Total is based on total recurrent, excluding debt and capital. Noteworthy at the central level, however, is that while education and health consistently executed their recurrent budgets at 92 percent to 104 percent during the period, the other sectors showed more fluctuating recurrent execution rates. Agriculture showed recurrent execution rates in line with education and health for 2004 and 2005, but a sudden drop in execution in 2006, down to 79 percent. This lack of execution within the agriculture ministry was partially attributed to the restructuring during the decentralization process that took place in full force in 2006 and left agriculture with limited resources on the central level and limited capacity on district levels to fully execute in the sector. Also, the infrastructure sector and the land and natural resource sector showed high variations in execution in the period. While infrastructure showed progress from a recurrent execution rate of 79 percent in 2004 to 104 percent in 2007, the land and natural resource sector went from 65 percent to 86 percent recurrent execution. Although this was an improvement in MINITERRE's execution rate, this sector significantly lags behind the other priority sectors in terms of recurrent expenditure execution. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 75 Table 4.13 shows the same sector breakdown in real term, actual recurrent public expenditures, but includes both central and decentralized spending for the five prioritysectors. Education, health, agriculture, infrastructure, and land and natural resources increased by RwF 21.2 billion, RwF 7.8 billion, RwF 1.1 billion, RwF 19.2 billion, and RwF 1.8 billion, respectively. Measured in terms of percentagegrowth over the period, infrastructure saw the largest amount of increase in recurrent expenditures, increasing by 249 percent, followed by a 154 percent increase by land and natural resources. In terms of both recurrent expenditures increase measured in million RwF and recurrent expenditure growth as a percentage over the period, agriculture experienced the lowest level of increase among the five prioritysectors. Table 4.13. Recurrent Central and Decentralized Executed Public Expenditure Trend 2004­07 Growth RwF 2004-07 % of Total Recurr* % of Total Recurr* % of Total Recurr* % of Total Recurr* Growth % 2004-07 Execution % 2004 Execution % 2005 Execution % 2006 Execution % 2007 REAL 2004 2005 2006 2007 Terms MINEDUC 54,151 23% 101% 57,726 21% 99% 63,708 22% 99% 75,423 24% 101% 21,272 39% MINISANTE 16,224 7% 96% 18,179 7% 94% 20,256 7% 94% 24,124 8% 106% 7,899 49% MINAGRI 5,308 2% 93% 6,784 3% 96% 3,846 1% 79% 6,385 2% 97% 1,076 20% MININFRA 7,688 3% 75% 15,738 6% 94% 14,139 5% 100% 26,861 8% 104% 19,173 249% MINITERRE 1,214 1% 65% 3,420 1% 83% 3,682 1% 79% 3,087 1% 89% 1,873 154% Source: MINECOFIN (in 2007 real prices). *Total is based on total recurrent, excluding debt and capital. Figure 4.4. Share of Total Recurrent Expenditures Transferred to Districts' Key Programs 80,000 70,000 Central Districts 60,000 Million RwF 50,000 40,000 30,000 20,000 10,000 0 Sector & Year Source: MINECOFIN (in 2007 real prices). 76 World Bank Working Paper As part of the recent decentralization efforts, these priority sectors have seen changes in the way districts receive their funding. Although the government continued to show education and health transfers to districts in its budget law, no agricultural allocations to districts were shown as of 2006, when decentralization went into full effect. However, this was a change not in allocations, but in how transfers were reported as MINECOFIN started to transfer agriculture funds directly to the districts, whereas in previous years, the funds went through MINAGRI. Table 4.14. Share of Recurrent Expenditures Transferred to Districts 2004­07 Central Districts TOTAL Edu 2004 54% 46% 100% Health 2004 66% 34% 100% Agri 2004 82% 18% 100% Infra 2004 99% 1% 100% Land&Re2004 93% 7% 100% Edu 2005 57% 43% 100% Health 2005 68% 32% 100% Agri 2005 89% 11% 100% Infra 2005 100% 0% 100% Land&Re2005 97% 3% 100% Edu 2006 60% 40% 100% Health 2006 66% 34% 100% Agri 2006 100% 0% 100% Infra 2006 100% 0% 100% Land&Re2006 100% 0% 100% Edu 2007 47% 53% 100% Health 2007 58% 42% 100% Agri 2007 66% 34% 100% Infra 2007 100% 0% 100% Land&Re2007 83% 17% 100% Source: MINECOFIN (in 2007 real prices). As table 4.14 clearly indicates, recurrent funds from centrallevel priority sectors transferred to districts remained low for all sectors other than education and health over the entire period, with an increase shown in transfers to districts' agriculture programs. Education transfers were between 40 percent and 53 percent of total recurrent education spending in the period, showing a declining trend, while health showed a transfer range of 32 percent to 42 percent of total recurrent health spending. Agriculture showed an 18 percent transfer rate in 2004 and no transfers in 2006, again due to structural changes in reporting (further details about agriculture transfers can be found in the section review of PER findings), but in 2007 the transfers were up to 34 percent of total ministry spending. Transfers to infrastructure and land and natural resources remained low over the period. On a central level in real terms, actual recurrent public expenditures for all other ministries increased over the period by RwF 7.8 billion. Average execution rates for Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 77 these nonpriority sectors at the central level were 83 percent, 88 percent, 101 percent, and 97 percent in 2004, 2005, 2006, and 2007, respectively. Table 4.15 shows a summary of all the nonpriority sectors centrallevel actual recurrent spending, nonpriority sectors' recurrent spending as a share of total recurrent expenditure (excluding debt and capital), and execution rates in the period. Table 4.15. All Other Ministries' Executed Recurrent Public Expenditure at Central Level 2004­07 Growth RwF 2004-07 % of Total Recurr* % of Total Recurr* % of Total Recurr* % of Total Recurr* Growth % 2004-07 Execution % 2004 Execution % 2005 Execution % 2006 Execution % 2007 REAL 2004 2005 2006 2007 Terms 01 - PARLEMENT 5,666 2% 95% 5,030 1% 89% 5,282 2% 88% 5,234 2% 98% ­432 ­8% 02 - PRESIREP 8,478 3% 98% 9,214 3% 102% 8,963 3% 102% 8,649 3% 101% 171 2% 04 - PRIMATURE 2,496 1% 82% 2,628 1% 91% 3,083 1% 91% 3,307 1% 109% 812 33% 05 - COURS.SUP 3,424 1% 60% 2,885 1% 92% 3,651 1% 89% 3,709 1% 104% 285 8% 06 - MINADEF 35,257 11% 100% 44,837 13% 109% 47,131 14% 109% 32,924 10% 100% ­2,333 ­7% 07 - MININTER 8,134 3% 87% 8,749 3% 89% 9,422 3% 95% 9,292 3% 97% 1,158 14% 08 - MINAFFET 5,013 2% 98% 5,919 2% 97% 6,161 2% 102% 6,906 2% 100% 1,894 38% 10 - MINICOM 2,702 1% 80% 4,693 1% 85% 5,192 2% 107% 3,739 1% 90% 1,038 38% 11- MINISTR -- -- -- -- -- -- -- -- -- 1,210 n.a. 81% n.a. n.a. 12 - MINECOFIN 102,837 32% 77% 76,572 22% 72% 81,709 25% 101% 98,875 31% 113% ­3,961 ­4% 13 - MINIJUST 3,353 1% 81% 8,360 2% 148% 4,533 1% 96% 6,812 2% 93% 3,459 103% 15 - MIJESPOC 1,327 0% 99% 1,850 1% 94% 1,714 1% 89% 1,773 1% 94% 446 34% 17 - PARQUET GENRERAL -- -- -- -- -- -- 1,925 1% 87% 1,832 1% 115% n.a. n.a. 19 - MIGEPROF 579 0% 59% 628 0% 60% -- -- -- -- -- -- n.a. n.a. 20 - MIFOTRA 966 0% 57% 3,198 1% 93% 2,505 1% 93% 2,219 1% 75% 1,253 130% 23 - MINALOC 14,844 5% 88% 18,656 5% 109% 21,635 7% 97% 17,291 5% 102% 2,447 16% TOTAL RT 195,075 60% 83% 193,220 57% 88% 202,907 62% 101% 203,773 64% 97% 7,832 4% Non-Priority Sectors: Source: MINECOFIN (in 2007 real prices). *Total is based on total recurrent, excluding debt and capital. Overall nonpriority recurrent spending as percentage of total recurrent spending remained consistent at 60 percent in 2004 to 64 percent in 2007. Nonpriority recurrent spending grew by 4 percent in real terms, from RwF 195.1 billion to RwF 203.8 billion. Table 4.16 shows the same sector breakdown in real terms in recurrent public expenditures, including both central and decentralized spending for all nonpriority sectors, which increased in total by RwF 5.9 billion. 78 World Bank Working Paper Table 4.16. All Other Ministries' Executed Recurrent Public Expenditure, including District Transfers 2004­07 Growth RwF 2004-07 % of Total Recurr* % of Total Recurr* % of Total Recurr* % of Total Recurr* Growth % 2004-07 Execution % 2004 Execution % 2005 Execution % 2006 Execution % 2007 REAL 2004 2005 2006 2007 Terms 01 - PARLEMENT 5,666 2% 95% 5,030 1% 89% 5,282 2% 88% 5,234 2% 98% ­432 ­8% 02 - PRESIREP 8,478 3% 98% 9,214 3% 102% 8,963 3% 102% 0 0% 0% ­8,478 ­100% 04 - PRIMATURE 2,496 1% 82% 2,628 1% 91% 3,083 1% 91% 3,307 1% 109% 812 33% 05 - COURS.SUP 3,424 1% 60% 2,885 1% 92% 3,651 1% 89% 3,709 1% 104% 285 8% 06 - MINADEF 35,257 11% 100% 44,837 13% 109% 47,131 14% 109% 32,924 10% 100% ­2,333 ­7% 07 - MININTER 14,979 5% 87% 16,027 5% 95% 14,083 4% 97% 15,277 5% 98% 298 2% 08 - MINAFFET 5,013 2% 98% 5,919 2% 97% 6,161 2% 102% 6,906 2% 100% 1,894 38% 10 - MINICOM 3,237 1% 80% 5,025 1% 81% 5,192 2% 107% 3,739 1% 90% 502 15% 11- MINISTR -- -- -- -- -- -- -- -- -- 1,210 0% 81% n.a. n.a. 12 - MINECOFIN 102,837 32% 77% 76,572 22% 72% 81,709 25% 101% 98,875 31% 113% ­3,961 ­4% 13 - MINIJUST 3,353 1% 81% 8,360 2% 148% 4,533 1% 96% 6,812 2% 93% 3,459 103% 15 - MIJESPOC 1,724 1% 99% 2,175 1% 91% 1,798 1% 90% 1,855 1% 94% 130 8% 17 - PARQUET GENRERAL -- -- -- -- -- -- 1,925 1% 87% 1,832 1% 115% n.a. n.a. 19 - MIGEPROF 579 0% 59% 628 0% 60% -- -- -- -- -- n.a. n.a. n.a. 20 - MIFOTRA 966 0% 57% 3,198 1% 93% 2,505 1% 93% 2,219 1% 75% 1,253 130% 23 - MINALOC 16,377 5% 88% 19,904 6% 104% 24,365 7% 97% 27,596 9% 101% 11,220 69% TOTAL RT 204,386 63% 83% 202,403 59% 88% 210,383 64% 101% 211,496 66% 100% 5,997 3% Non-Priority Sectors: Source: MINECOFIN (in 2007 real prices). *Total is based on total recurrent, excluding debt and capital. The differences between centrallevel spending and combined central and transfer todistrict spending were minimal as many ministries do not have transfers to districts.6 In addition to MINEDUC, MINISANTE, and MINAGRI, MINALOC--with its budget support transfers to the Local Authority Block Support Fund (LABSF) starting in 2006--and the Ministry of Internal Security (MININTER)--with its prisons--had significant transfers. A further evaluation of district transfers follows in the section below. Trends in Recurrent Transfers to Districts Overall Transfer Trends. Transfers to districts and provinces went through a transformation, which could be seen in particular in 2006, when the full decentralization went into effect. In 2004 and 2005, the district portion of the budget law was divided into 12 provinces, and the programs and subprograms for these two years for each of the 12 provinces were rather detailed and consistent. In 2006, however, Rwanda changed from its 12 provinces into 4 provinces and 30 districts, and the budget showed transfers to only three main programs: education, health, and good governance. That did not mean, however, that nothing was transferred to districts in other sectors. For example, the district section of the budget showed no transfers to agriculture; but instead of going through MINAGRI, funding was transferred directly Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 79 from MINECOFIN. In 2007, the number of programs receiving transfers in the budget increased to a total of 17. While major sectors such as education, health, and agriculture showed larger amounts of transfers to districts, smaller amounts were also allocated for sectors such as energy, land and resources, prisons, good governance, and so forth. Attempts to classify these expenses to estimate total recurrent spending of onbudget sector allocation are based on assumptions of district programs and subprograms. Data Sets 6­9 show in detail each province's and district's program expenditures and execution rates between 2004 and 2007 and also indicate assumptions for sector allocations. Trends in Transfers by Line Ministry For health and education, decentralized expenditures increased in the period by RwF 4.7 billion and RwF 14.9 billion, respectively. This represents a 60 percent increase in education transfers and a 84 percent increase in health transfers between 2004 and 2007. As previously stated, transfers to agriculture, infrastructure, and land and natural resources were not recorded in 2006 but were rather small compared with education and health, as shown in table 4.17. Table 4.17. Actual District Transfers 2004­07 Decentralized Total Change in Total Change in (Transfers) Transfers to Districts Transfers to Districts Actuals 2004 2005 2006 2007 (RwF) 2004-07 (%) 2004-07 Education 24,856 24,839 28,514 39,797 14,941 60% Health 5,512 5,897 7,632 10,166 4,654 84% Agriculture 971 776 n.a. 2,155 1,185 122% Infrastructure 114 74 n.a. 31 ­83 ­73% Land & Natural 87 101 n.a. 532 445 511% Resources Source: MINECOFIN (in 2007 real prices). Furthermore, execution rates varied in districts and provinces during the period. Table 4.18 shows 2004­07 execution rates earmarked for priority sectors for all districts and provinces combined. As the table shows, education had the highest execution rate in the entire period, at or around 101 percent on average. Health also showed a high execution rate in 2004 and 2005 at 100 percent; despite a sudden drop in 2006 to only 84 percent, the average execution rate for the period was 100 percent. In contrast, the three remaining earmarked transfers to priority sectors show average execution rates of only 87 percent in agriculture, 78 percent in infrastructure, and 63 percent in land and resources. However, it is noteworthy that both agriculture and infrastructure earmarked transfers were executed at 100 percent in 2007, an improvement from previous years. 80 World Bank Working Paper Table 4.18. Execution Rates for Earmarked Transfers to Priority Sectors 2004­07 Decentralized (Transfers) Average Execution Execution 2004 2005 2006 2007 Rate (%) 2004­07 Education 98% 98% 100% 107% 101% Health 100% 100% 84% 117% 100% Agriculture 90% 72% n.a. 100% 87% Infrastructure 71% 64% n.a. 100% 78% Land & Natural Resources 74% 62% n.a. 53% 63% Source: MINECOFIN (in 2007 real prices). For nonpriority sectors' earmarked transfers, MININTER received the most transfers to districts, primarily because of the prisons spread out in many districts (see table 4.19). Also, MINALOC transferred funds to districts and provinces mostly under programs identified as good governance. In 2007, LABSF (block transfers to support operations at the district level) was reported under program 18 in districts' budgets; assuming that LABSF falls under MINALOC (see Data Sets 6, 7, 8, and 9), MINALOC's transfers increased 572 percent in the period, while MININTER's transfers decreased 13 percent. Also MIJESPOC's allocation to districts decreased from RwF 398 million to only RwF 82 million in real terms, divided over the 30 districts, which represents a 79 percent decrease for an already small sector. Table 4.19. Actual District Transfers in Nonpriority Sectors 2004­07 Total Change in Total Change Transfers to in Transfers Decentralized Districts (RwF) to Districts (Transfers) Actuals 2004 2005 2006 2007 2004-07 (%) 2004-07 07 - MININTER 6,844 7,278 4,661 5,985 -859 -13% 10 - MINICOM 536 332 n.a. n.a. n.a. n.a. 13 - MINIJUST 1 0 n.a. n.a. n.a. n.a. 15 - MIJESPOC 398 325 84 82 -316 -79% 23 - MINALOC 1,533 1,248 2,730 10,305 8,773 572% Source: MINECOFIN (in 2007 real prices). Execution rates in nonpriority sectors in districts have varied as well. While districts have executed MININTERrelated expenses (mostly prisonrelated recurrent expenditures) consistently around 100 percent, the other sectors have struggled to execute consistently, although both MIJESPOC and MINALOC improved their execution rates (see table 4.20). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 81 Table 4.20. Execution Rates for Transfers in Nonpriority Sectors 2004­07 Decentralized (Transfers) Average Execution Execution 2004 2005 2006 2007 Rate (%) 2004­07 07 - MININTER 99% 102% 100% 100% 100% 10 - MINICOM 78% 49% n.a. n.a. n.a. 13 - MINIJUST 75% n.a. n.a. n.a. n.a. 15 - MIJESPOC 87% 78% 95% 100% 90% 23 - MINALOC 75% 60% 98% 101% 83% Source: MINECOFIN (in 2007 real prices). Again, refer to Data Sets 6, 7, 8, and 9 for further details of provinces' and districts' execution rates and assumptions for classification of programs to these ministries. Trends in Development Expenditure by Line Ministry When analyzing development budget expenditure in Rwanda, it is important to keep in mind that executed data are not allinclusive. First, MINECOFIN reported that information on execution of donor funds is not consistently made available to the ministry, and hence, often external funding that is assumed to be 100 percent funded may not always be. In particular, this was the case prior to 2006. In 2006, CEPEX, funded by MINECOFIN, published its first comprehensive development budget report, which was supposed to closely capture all onbudget internally and externally financed projects, as well as many offbudget projects. However, it was difficult to make a precise comparison of the original development budget law, as presented by MINECOFIN and approved by Parliament, and the revised and actual budget, as reported by CEPEX, as the two sources include different data. Specifically, CEPEX captured some spending not reflected in the budget law. Furthermore, in the budget law, no development budget allocation is provided for the provinces and districts as is done for recurrent expenditures. Therefore, for the purpose of this report, actual development spending analysis by ministry in chapters 3 and 4 uses CEPEXexecuted data for 2006 and the revised development budget law for 2007. Details of development budget allocations and CEPEXreported spending can be found in Data Sets 10­18. Overall Development Expenditure Trends Overall, actual development spending in real terms went from RwF 110.5 billion in 2004 to RwF 181.9 billion in 2007, representing an overall increase of 64 percent in the period. Furthermore, in 2005, the largest allocations of development budget spending went to MININFRA, with 31 percent of total development spending, and MINISANTE and MINALOC, with 13 percent of total allocation each, representing a combined total of 57 percent of the total actual development spending that year. These ministries were followed by MINITERRE, MINAGRI, MINEDUC, and MINECOFIN with 10 percent, 9 percent, 8 percent, and 8 percent of total development spending, respectively. Table 4.21 shows the executed development budget (that is, actual development spending) by ministry and the portion of each ministry's executed development budget as a portion of the total development budget. 82 World Bank Working Paper Table 4.21. Overall Actual Development Expenditure Trends 2004­07 REAL TERMS Actual % of Total Actual % of Total Actual % of Total Actual % of Total (in million RwF) 2004 DB 2005 DB 2006 DB 2007* DB 01 - PARLIAMENT data missing n.a. 28.3 0% 0.0 0% 0.0 0% 04 - PRIMATURE 4.3 0% 0.0 0% 1,491.8 1% 2,454.0 1% 05 - COURS.SUP data missing n.a. 2,414.5 2% 794.5 0% 674.0 0% 06 - MINADEF data missing n.a. 803.0 1% 712.9 0% 1,060.0 1% 07 - MININTER 259.1 0% 1,547.9 1% 2,156.9 1% 2,484.0 1% 08 - MINAFFET data missing n.a. 0.0 0% 0.0 0% 1,115.0 1% 09 - MINAGRI 9,977.0 9% 14,468.1 9% 16,517.3 10% 13,517.0 7% 10 - MINICOM 532.0 0% 713.5 0% 3,774.1 2% 5,357.0 3% 12 - MINECOFIN data missing n.a. 12,554.7 8% 7,131.2 4% 5,799.0 3% 13 - MINIJUST data missing n.a. 2,402.7 2% 4,411.3 3% 3,555.0 2% 14 - MINEDUC 5,853.9 5% 12,653.3 8% 11,864.3 7% 19,862.4 11% 15 - MIJESPOC data missing n.a. 107.5 0% 998.9 1% 1,313.0 1% 16 - MINISANTE 9,051.8 8% 19,238.3 13% 49,693.6 29% 25,684.0 14% 18 - MININFRA 30,711.9 28% 47,098.4 31% 32,050.5 19% 52,125.0 29% 19 - MIGEPROF data missing n.a. 4,407.3 3% 0.0 0% 0.0 0% 20 - MIFOTRA data missing n.a. 138.6 0% 1,550.2 1% 1,919.0 1% 22 - MINITERRE 5,157.0 5% 14,767.2 10% 20,214.2 12% 25,903.0 14% 23 - MINALOC 4,847.4 4% 19,788.1 13% 16,903.0 10% 19,048.1 10% TOTAL DB Actual 110,507.8 n.a. 153,131.1 100% 170,264.7 100% 181,869.5 100% Source: CEPEX/MINECOFIN (in 2007 real terms). Note: DB = Development Budget. *2007 Revised Development Budget Law. In 2006, the allocation priorities of the total development budget changed. MINISANTE received the largest amount of development funding with 29 percent of the total, followed by MININFRA with 19 percent. And while MINALOC's allocation share decreased from 13 percent in 2005 to 10 percent in 2006, MINITERRE and MINAGRI' allocations increased slightly to 12 percent and 10 percent, respectively. MINEDUC's allocation dipped from 8 percent to 7 percent. The largest change of allocation happened within MINECOFIN, where the allocation portion decreased by half, from 8 percent in 2005 to 4 percent in 2006. In 2007, some of these trends were reversed again as MININFRA spent 29 percent of the total development budget compared with 19 percent in 2006. The ministries with the next largest development allocations in 2007 were MINISANTE and MINITERRE, each with 14 percent. Actual overall development spending by source of funding cannot be examined in detail in 2004 and 2005 based on the way MINECOFIN reported data in this period. Such data are available only by an examination of original budget data. However, based on 2006 and 2007 data, table 4.22 provides a breakdown of source of funding, again keeping in mind that actual data for 2006 include some offbudget projects as well. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 83 Table 4.22. Overall Actual Development Spending by Source of Funding 2004-07 Growth REAL Terms % of % of % of % of % (in million Actual Total Actual Total Actual Total Actual Total 2006- RwF) 2004 DB 2005 DB 2006 DB 2007* DB 2007 Total CP n.a. n.a. n.a. n.a. 11,187.9 7% 7,237.1 4% ­35% Total Don n.a. n.a. n.a. n.a. 85,781.2 50% 88,040.0 48% 3% Total FI n.a. n.a. n.a. n.a. 29,971.6 18% 52,062.4 29% 74% Total Loan n.a. n.a. n.a. n.a. 43,323.9 25% 34,530.0 19% ­20% Total CP & FI n.a. n.a. n.a. n.a. 41,159.5 24% 59,299.5 33% 44% Total Don & n.a. n.a. n.a. n.a. 129,105.1 76% 122,570.0 67% ­5% Loan TOTAL DB: n.a. n.a. n.a. n.a. 170,264.7 100% 181,869.5 100% 7% Source: CEPEX/MINECOFIN (in 2007 real prices). Note: DB = Development Budget; CP = Counterpart funding; FI = Internally Financed. *2007 Revised Development Budget Law. As table 4.22 illustrates, in 2006 50 percent of the development budget came from donors. In 2007, this figure was down to 48 percent. The second largest source of funds in 2006 was loans, with a 25 percent contribution. Internally financed (FI) funding increased from 18 percent in 2006 to 29 percent in 2007, while counterpart (CP) funding decreased from 7 percent to 4 percent. Since specific executed data by source of funding for 2004 and 2005 are not readily available, it may be helpful to look at the original budget data to identify any trends in allocation by source of funding in the overall period. Table 4.23 illustrates the same information as table 4.22, but with original development budget data across all four years. Table 4.23. Overall Budgeted Development Spending by Source of Funding 2004­07 Growth Orig % of Orig % of Orig % of Orig % of % Budget Total Budget Total Budget Total Budget Total 2004- REAL Terms 2004 DB 2005 DB 2006 DB 2007 DB 2007 Total CP 3,524.3 3% 5,885.5 4% 8,373.6 7% 7,307.1 4% 107% Total Don 55,727.8 49% 65,717.8 49% 59,938.5 48% 88,040.0 48% 58% Total FI 23,026.5 20% 29,296.2 22% 23,474.6 19% 51,992.4 29% 126% Total Loan 30,415.5 27% 33,454.3 25% 33,224.4 27% 34,529.9 19% 14% Total CP & FI 26,550.7 24% 35,181.7 26% 31,848.2 25% 59,299.5 33% 123% Total Don & 86,143.4 76% 99,172.2 74% 93,162.9 75% 122,569.9 67% 42% Loan TOTAL DB: 112,694.1 100% 134,353.9 100% 125,011.2 100% 181,869.4 100% 61% Source: MINECOFIN (in 2007 real prices). Note: DB = Development Budget; CP = Counterpart funding; FI = Internally Financed. As table 4.23 shows, little or no change occurred in the original budget sources between 2004 and 2005 and only a slight change between 2005 and 2006. In 2006, CP funding increased to 7 percent from 4 percent, and FI funds decreased from 22 percent 84 World Bank Working Paper to 19 percent, while donor funding and loans remained stable at around 48 percent and 27 percent, respectively. In 2007, however, there was a slight change in the source of funding. This change occurred mostly between FI funding and loans. While FI funds increased from 19 percent in 2006 to 29 percent in 2007, loans decreased from 27 percent to 19 percent. Donor funding remained consistent at 48 percent, CP funding was back down at only 4 percent. Furthermore, table 4.23 shows that the largest percentage increase among the four sources of funding as anticipated in the original budget occurred in the FI funds, with an increase between 2004 and 2007 of 126 percent. Simultaneously, CP funding, while still the smallest source of funding, had increased 107 percent by 2007. Donor funding increased by 58 percent, and loans increased by 14 percent. Comparing external sources of funding (donors and loans) and internal sources of funding (CP and FI), internal funding increased by 123 percent, while external funding increased by 42 percent. Overall, the development budget increased 61 percent between 2004 and 2007, and external financing contributing twothirds of the total development budget in 2007. This is a decrease from 76 percent, 74 percent and 75 percent in 2004, 2005, and 2006 respectively. Trends in Development Expenditures by Line Ministry The portion of total actual development spending attributed to each of the main ministries contributing to the five priority sectors (education, health, agriculture, infrastructure, and land and resources) remained stable in some sectors and fluctuated in others (see table 4.24). MINEDUC's allocation of the development budget went from 5 percent to 11 percent between 2004 and 2007, while MINISANTE saw a sharper increase in development expenditure allocation, from 8 percent to 14 percent, in the same period. MINAGRI retained a stable development expenditure allocation of about 7 percent to 10 percent, and MININFRA only a slight increase from 28 percent in 2004 to 29 percent in 2007. MINITERRE's share of total development spending increased from 5 percent in 2004 to 14 percent in 2007. In 2006, MINISANTE received the largest allocation of development funding (29 percent), followed by MININFRA (19 percent), MINITERRE (12 percent), MINAGRI (10 percent), and MINEDUC (7 percent). In 2007, MININFRA was responsible for the greatest portion of the development spending (29 percent), followed by MINISANTE and MINITERRE (14 percent each). Table 4.24. Actual Development Budget for Five Priority Ministries 2004­07 % of % of % of % of total total total total Increase Increase REAL Terms (in Actual Dev Actual Dev Actual Dev Actual Dev RwF 2004- % 2004- million RwF) DB 2004 Spend DB 2005 Spend DB 2006 Spend DB 2007* Spend 07 07 Education 5,854 7% 12,653 10% 11,864 8% 19,862 11% 14,008 239% Health 9,052 11% 19,238 16% 49,694 33% 25,684 14% 16,632 184% Agriculture 9,977 12% 14,468 12% 16,517 11% 13,517 7% 3,540 35% Infrastructure 30,712 38% 47,098 39% 32,050 21% 52,125 29% 21,413 70% Land & Natural 5,157 6% 14,767 12% 20,214 13% 25,903 14% 20,746 402% Resources TOTAL ALL SECTORS 110,508 100% 153,131 100% 170,265 100% 181,870 100% 71,362 65% Source: MINECOFIN (in 2007 real prices). *2007 Revised Development Budget. ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 85 .5 .25 further illu Figure 4. and table 4. cation of the o ustrate the alloc overall actual development budget among the five ministr ries. ctual Developm Figure 4.5. Ac Trends 2004­07 ment Spending T 7 In million RwF OFIN (in 2007 real prices). Source: MINECO verall Actual De Table 4.25. Ov penditures 2004 evelopment Exp 4­07 In Real Terms Actual DB Actual DB A Actua DB al B Actual DB Growth % in million RwF 2004 2005 2006 2007* 2004-2007 Education 5,854 12,653 64 11,86 19,862 239% Health 9,052 19,238 94 49,69 25,684 184% Agriculture 9,977 14,468 16,517 13,517 35% Infrastructure 30,712 47,098 32,05 50 52,125 70% Land & Natural Res sources 5,157 14,767 20,214 25,903 402% Total Priority Sectors 60,752 108,225 130,34 40 137,091 126% Total DB All Sectors s 110,508 153,131 65 107,26 181,870 65% g % of Total DB going to Priority: 55% 71% 7% 77 75% n.a. Source: MINECOOFIN (in 2007 real prices). Development Budg *2007 Revised D get. be e RA As can b seen in table 4.25, MININFR has remain ned the largest development budget spend der in the period eption of MINIS d, with the exce SANTE in 2006 6. Overall, the ministries accou five priority m unted for 55 perrcent, 71 percennt, 77 percent, annd 75 percent of the overalll nationwide ex xecuted developpment budget in n 2004, 2005, 20006, and 2007, respectively. Figure 4. shows the trend within ea .6 t ctor, comparing actual and ach priority sec g original deveelopment budget allocations. Series 1 (blue) represents oriiginal budget data, and Seri esents actual budget data. ies 2 (red) repre 86 World Bank Working Paper Figure 4.6. Original Development Budget versus Actual Development Expenditures for Five Priority Sectors 2004­07 Source: MINECOFIN (in 2007 real prices). In terms of actual development spending, MINITERRE steadily increased their development budgets, while MININFRA's budget fell between 2005 and 2006 and MINISANTE's and MINAGRI's budgets dropped in 2007. MINEDUC showed almost flat actual development spending between 2005 and 2006. MINEDUC, MINISANTE, MINITERRE, and MINAGRI executed above their development budgets in 2006. MINISANTE had the highest overexecution (in 2006) of all the priority sectors. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 87 MININFRA and MINEDUC executed closest to their original budgets, while MINITERRE and MINAGRI outspent their original budgets in 2005 and 2006. Since revised 2007 development budget data are used to estimate 2007 actual development spending, no conclusion about over or underexecution can be made for 2007. Source of funding for the actual development budget does not exist for 2004 and 2005. Table 4.26 shows the sources for original budget data, broken down by year, for 2004­07. Table 4.26. Sources of Original Development Expenditure Allocation 2004­07 FI % of CP % of DON % Loan % Total Total of Total of Total Source of Funding Counter- TOTAL Sector Sector Sector Sector Original Budget Only Internally part Donor DEVEL. DB DB DB DB 2004­07 Financed Funds Funds Loans FUNDING Funding Funding Funding Funding Budgeted (Original) 2007 Education 12,693 388 3,070 3,711 19,862 64% 2% 15% 19% Health 904 219 23,141 1,420 25,684 4% 1% 90% 6% Agriculture 1,919 957 6,101 4,540 13,517 14% 7% 45% 34% Infrastructure 10,866 2,606 22,535 16,118 52,125 21% 5% 43% 31% Land & Natural 7,467 1,735 13,104 3,597 25,903 29% 7% 51% 14% Resources Budgeted (Original) 2006 Education 3,021 916 4,255 3,035 11,227 27% 8% 38% 27% Health 561 399 17,734 1,721 20,415 3% 2% 87% 8% Agriculture 0 934 5,740 3,044 9,718 0% 10% 59% 31% Infrastructure 8,154 2,710 8,614 15,120 34,599 24% 8% 25% 44% Land & Natural 786 2,364 6,685 3,477 13,312 6% 18% 50% 26% Resources Budgeted (Original) 2005 Education 3,354 444 3,572 3,133 10,503 32% 4% 34% 30% Health 1,777 195 19,497 0 21,469 8% 1% 91% 0% Agriculture 444 498 2,680 4,354 7,976 6% 6% 34% 55% Infrastructure 12,754 2,373 15,126 16,001 46,253 28% 5% 33% 35% Land & Natural 1,586 1,472 5,356 2,835 11,249 14% 13% 48% 25% Resources Budgeted (Original) 2004 Education 691 359 6,714 4,476 12,239 6% 3% 55% 37% Health 284 62 9,987 0 10,333 3% 1% 97% 0% Agriculture * * * * * n.a. n.a. n.a. n.a. Infrastructure 4,355 1,135 10,820 8,638 24,948 17% 5% 43% 35% Land & Natural 346 631 7,609 3,676 12,261 3% 5% 62% 30% Resources Source: MINECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for MINAGRI (agriculture) are omitted. In 2007, among the five priority sectors, education received the largest portion of internally financed development sources at 64 percent, followed by land and natural resources at 29 percent and infrastructure at 21 percent. In terms of donor funding in 2007, health sourced 90 percent of its total development budget from donor funds. Land and natural resources sourced 51 percent of its overall sector development budget from donors in 2007. Agriculture and 88 World Bank Working Paper infrastructure received 45 percent and 43 percent, respectively of their overall development spending from donors. These two latter sectors also received the largest funds from loans at 34 percent and 31 percent, respectively, in 2007. While education sourced 15 percent of the sector's overall development funds from donor contributions, health sourced only 6 percent of its development spending from loans. Financing from counterpart funds remained minimal across all five sectors. Put in a different way, data for 2007 show that health relied most heavily on donors to finance its development activities (90 percent of its 2007 total sector development spending), while education received the smallest portion from donors (15 percent). Education instead relied mostly on FI (64 percent of its overall development budget), followed by loan funding (19 percent). Agriculture and infrastructure had a more even distribution of sources. Land and natural resources got 80 percent of its development financing from a combination of FI and donors. Looking at the same original budget data across all years, a change in the sources of funding by the five priority sectors is evident. Table 4.27 provides the same data as table 4.26 but with the source of funding broken down by priority sector. Table 4.27. Original Development Expenditure Allocation for Priority Ministries by Funding Sources 2004­07 FI % of CP % of DON % Loan % Total Total of Total of Total Source of Funding - Counter- TOTAL Sector Sector Sector Sector Budget Trends by Internally part Donor DEVEL. DB DB DB DB Priority Ministry Financed Funds Funds Loans FUNGING Funding Funding Funding Funding Education Budgeted (Original) 2007 12,693 388 3,070 3,711 19,862 64% 2% 15% 19% Budgeted (Original) 2006 3,021 916 4,255 3,035 11,227 27% 8% 38% 27% Budgeted (Original) 2005 3,354 444 3,572 3,133 10,503 32% 4% 34% 30% Budgeted (Original) 2004 691 359 6,714 4,476 12,239 6% 3% 55% 37% Health Budgeted (Original) 2007 904 219 23,141 1,420 25,684 4% 1% 90% 6% Budgeted (Original) 2006 561 399 17,734 1,721 20,415 3% 2% 87% 8% Budgeted (Original) 2005 1,777 195 19,497 0 21,469 8% 1% 91% 0% Budgeted (Original) 2004 284 62 9,987 0 10,333 3% 1% 97% 0% Agriculture* Budgeted (Original) 2007 1,919 957 6,101 4,540 13,517 14% 7% 45% 34% Budgeted (Original) 2006 0 934 5,740 3,044 9,718 0% 10% 59% 31% Budgeted (Original) 2005 444 498 2,680 4,354 7,976 6% 6% 34% 55% Budgeted (Original) 2004 * * * * * n.a. n.a. n.a. n.a. Infrastructure Budgeted (Original) 2007 10,866 2,606 22,535 16,118 52,125 21% 5% 43% 31% Budgeted (Original) 2006 8,154 2,710 8,614 15,120 34,599 24% 8% 25% 44% Budgeted (Original) 2005 12,754 2,373 15,126 16,001 46,253 28% 5% 33% 35% Budgeted (Original) 2004 4,355 1,135 10,820 8,638 24,948 17% 5% 43% 35% Land & Natural Resources Budgeted (Original) 2007 7,467 1,735 13,104 3,597 25,903 29% 7% 51% 14% Budgeted (Original) 2006 786 2,364 6,685 3,477 13,312 6% 18% 50% 26% Budgeted (Original) 2005 1,586 1,472 5,356 2,835 11,249 14% 13% 48% 25% Budgeted (Original) 2004 346 631 7,609 3,676 12,261 3% 5% 62% 30% Source: MINECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for MINAGRI (agriculture) are omitted. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 89 Table 4.27, presenting the data per ministry and year, allows a closer examination of the trends in original development budgets for 2004­07. For example, while health continued to receive a significant part of its development budget from donors, the education sector saw its portion of funding from donors go down from 55 percent in 2004 to 15 percent in 2007. Instead, education has seen a significant increase in funding from FI, up from 6 percent in 2004 to 64 percent in 2007. This is in sharp contrast with health, which maintained its sourcing of FI at the 3 percent to 8 percent range. While land and natural resources and agriculture have increased their portions of FI funding from low singledigits in 2004 to 29 percent and 14 percent, respectively, of their total 2007 sector development funding, infrastructure more consistently received FI between 17 percent and 28 percent (21 percent in 2007). Furthermore, in the period, only health saw an increase in the portion of its development budget sourced by loans. Overall, the biggest change in sourcing of funds appears to have occurred in education, with its sharp increase of FI and decrease in donor financing. The following tables and graphs will shed additional light on the source of funding as allocated in the original development budgets for the five priority sectors. Table 4.28 and figure 4.7 show allocation among priority sectors of FI funds. Table 4.28. Comparing Internally Financed Development Original Budget Data 2004­07 Internally Internally Internally Internally REAL Terms Financed Financed Financed Financed Growth % FI in million RwF 2004 2005 2006 2007 2004­07 Education 691 3,354 3,021 12,693 1736% Health 284 1,777 561 904 218% Agriculture * 444 0 1,919 n.a. Infrastructure 4,355 12,754 8,154 10,866 149% Land & Resources 346 1,586 786 7,467 2060% Total Priority Sectors 5,677 19,915 12,522 33,849 496% Source: MINECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for MINAGRI (agriculture) are omitted. Figure 4.7. Trends in Development Budget Internally Financed Funding 2004­07 14,000 12,000 10,000 Million RwF 8,000 6,000 4,000 2,000 0 Education Health Agriculture Infrastructure Land & Resources Internally Financed 2004 Internally Financed 2005 Internally Financed 2006 Internally Financed 2007 Source: MINECOFIN (in 2007 real prices). 90 World Bank Working Paper As is clearly shown in the graphics above, education significantly increased its source of funding from FI in 2007 compared with previous years, from RwF 691 million in 2004 to RwF 12.7 billion in 2007, representing 64 percent of its 2007 total source of funding. Similarly, land and resources also saw a significant increase in its share of FI funding over the period, although this sector continued to receive its largest financing from donors. Land and resources FI funding increased from RwF 346 million to RwF 7.5 billion in the same period, which was a 2,060 percent increase, but still only represented 29 percent of total funding in 2007 for this sector. While FI for the infrastructure sector fluctuated somewhat in the period, infrastructure received the most money in absolute terms from FI funding over the last three years of the period. Agriculture and health continued to receive minimal FI. Overall, the priority sectors saw a 496 percent increase in FI financing from 2004 to 2007. Table 4.29 shows allocation among priority sectors based on CP funds. Although CP funding remains the smallest source of funding for the priority sectors, this source of funding nonetheless increased overall in the period by 170 percent, with the largest increase, 255 percent, in the health sector. Table 4.29. Comparing Counterpart Development Budgets 2004­07 Counter-part Counter-part Counter-part Funding Funding Funding Counter-part Growth % CP REAL Terms in million RwF 2004 2005 2006 Funding 2007 2004­07 Education 359 444 916 388 8% Health 62 195 399 219 255% Agriculture * 498 934 957 n.a. Infrastructure 1,135 2,373 2,710 2,606 130% Land & Resources 631 1,472 2,364 1,735 175% Total Priority Sectors 2,186 4,982 7,323 5,905 170% Source: MINECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for MINAGRI (agriculture) are omitted. Figure 4.8. Trends in Development Budget Counterpart Funding 2004­07 3,000 2,500 Million RwF 2,000 1,500 1,000 500 0 Education Health Agriculture Infrastructure Land & Resources Counter part Funding 2004 Counter part Funding 2005 Counter part Funding 2006 Counter part Funding 2007 Source: MINECOFIN (in 2007 real prices). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 91 For CP funds, the trends illustrated in figure 4.8 are not as drastic as for some other sources of funds. The main sectors receiving CP funding were infrastructure and land and natural resources, growing their CP funding by 130 percent and 175 percent, respectively, between 2004 and 2007. Agriculture, too, showed an upward trend, but its portion of CP funding remained significantly less than infrastructure and land and resources in absolute terms. Health had a small portion of CP funding but saw an increase of 255 percent in the period. A closer look at donor funding between 2004 and 2007 shows that education in particular experienced a significant change in donor funding. As noted earlier, education experienced a significant increase in FI funding, which was offset by its significant decrease in donor funding. Overall, donor funding increased by 93 percent. Table 4.30 and figure 4.9 illustrate the trends in donor funding for the five priority sectors. Table 4.30. Comparing Donor Development Budgets 2004­07 Donor Donor Donor Funding Funding Funding Donor Growth % REAL Terms in million RwF 2004 2005 2006 Funding 2007 DON 2004­07 Education 6,714 3,572 4,255 3,070 ­54% Health 9,987 19,497 17,734 23,141 132% Agriculture * 2,680 5,740 6,101 n.a. Infrastructure 10,820 15,126 8,614 22,535 108% Land & Resources 7,609 5,356 6,685 13,104 72% Total Priority Sectors 35,129 46,231 43,027 67,951 93% Source: MINECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for MINAGRI (agriculture) are omitted. Figure 4.9. Trends in Development Budget Donor Funding 2004­07 25,000 20,000 15,000 Million RwF 10,000 5,000 0 Education Health Agriculture Infrastructure Land & Resources Donor Funding 2004 Donor Funding 2005 Donor Funding 2006 Donor Funding 2007 Source: MINECOFIN (in 2007 real prices). 92 World Bank Working Paper As figure 4.9 shows, health received the greatest portion of donor funding for its development budget, with a significant jump in its 2007 budget allocation. But the greatest increase in donor development funding occurred in the infrastructure and the land and natural resource sectors in 2007; their donor funding more than doubled between 2006 and 2007, putting donor funding for infrastructure in 2007 not far behind health in absolute terms. Overall, donor funding through the period for health, infrastructure, and land and natural resources grew by 132 percent, 108 percent, and 72 percent, respectively. As noted earlier, education saw a decrease in donor funding, and in absolute terms, its figures continued to be significantly lower than the three largest donorreceiving sectors. Agriculture enjoyed an increase in donor funds, but remained significantly below other priority sectors with exception of education. Table 4.31 and figure 4.10 illustrate development funding sourced by loans for all five priority sectors. Table 4.31. Comparing Loans Development Budgets 2004­07 REAL Terms Growth % in million RwF Loans 2004 Loans 2005 Loans 2006 Loans 2007 LOAN 2004­07 Education 4,476 3,133 3,035 3,711 ­17% Health 0 0 1,721 1,420 n.a. Agriculture * 4,354 3,044 4,540 n.a. Infrastructure 8,638 16,001 15,120 16,118 87% Land & Resources 3,676 2,835 3,477 3,597 ­2% Total Priority Sectors 16,789 26,323 26,397 29,386 75% Source: MINECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for MINAGRI (agriculture) are omitted. Figure 4.10. Trends in Development Budget Loan Funding 2004­07 18,000 16,000 14,000 12,000 Million RwF 10,000 8,000 6,000 4,000 2,000 0 Education Health Agriculture Infrastructure Land & Resources Loans Loans Loans Loans 2004 2005 2006 2007 Source: MINECOFIN (in 2007 real prices). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 93 Infrastructure was the largest borrower of funds for development measured in absolute terms, an increase in loans between 2004 and 2007 of 87 percent. In the period, as illustrated in earlier tables, infrastructure had sourced between 31 percent and 44 percent of its overall sector funding from loans. While agriculture and to some degree land and natural resources and education also received a nonnegligible amount of funding from loans, health received by far the smallest amount of funding from loans (6 percent in 2006 and 8 percent in 2007). Education loans decreased by 17 percent, while land and natural resource loans decreased by 2 percent. Source funding for executed budget data is available only through CEPEX annual reports for 2006 and 2007. However, because of some challenges in the 2007 reported data (which included expenditures that did not go through MINECOFIN), an accurate comparison to budget allocations is difficult. Consequently, for the purposes of this comparison, revised budget data are used for 2007. Table 4.32 provides a highlight of the amount in the executed development budget that was spent in priority sectors versus overall development spending. Table 4.32. Comparing Sources of Actual Development Budget Funding 2006­07 REAL Terms Internally Counterpart Donor in million RwF Financed Funds Funds Loans TOTAL DB 2006 Priority Sectors 16,968 9,575 70,982 32,815 130,340 Total DB All Sectors 29,972 11,188 85,781 43,324 170,265 Portion of Total in % 57% 86% 83% 76% 77% 2007* Priority Sectors 33,919 5,835 67,951 29,386 137,091 Total DB All Sectors 52,062 7,237 88,040 34,530 181,870 Portion of Total in % 65% 81% 77% 85% 75% Source: MINECOFIN (in 2007 real prices). Note: DB = Development Budget. *2007 Revised Development Budget. As table 4.32 shows, the five priority sectors received an overall 75 percent of total executed development spending in 2007. Loans to the priority sectors accounted for 85 percent of all loans for development spending received in 2007, the highest proportion of sources going to the priority sectors. Next was CP funds, of which 81 percent went to the priority sectors. Donor funding for the five sectors accounted for 77 percent of all actual development spending from donors, while the priority development spending sourced by FI funding made up 65 percent of all FI funding in 2007. Table 4.33 illustrates the breakdown of sources of funds for the development budgets of each of the five priority sectors in 2006 and 2007 (using revised budget data for 2007). Source data on revised and executed funding were not available for 2004 and 2005. 94 World Bank Working Paper Table 4.33. Sources of Development Budget Funding for the Five Priority Sectors 2006­07 FI % of CP % of DON % Loan % Source of Funding-- Counter- TOTAL Total Total of Total of Total Comparing 2006 & Internally part Donor DEVEL. Dev Dev Dev Dev 2007 Actual Only Financed Funds Funds Loans FUNDING Funding Funding Funding Funding Revised Budget 2007 Education 12,693 388 3,070 3,711 19,862 64% 2% 15% 19% Health 904 219 23,141 1,420 25,684 4% 1% 90% 6% Agriculture 1,919 957 6,101 4,540 13,517 14% 7% 45% 34% Infrastructure 10,866 2,606 22,535 16,118 52,125 21% 5% 43% 31% Land & Natural 7,537 1,665 13,104 3,597 25,903 29% 6% 51% 14% Resources Executed 2006 Education 2,897 1,774 3,005 4,189 11,864 24% 15% 25% 35% Health 563 622 48,004 504 49,694 1% 1% 97% 1% Agriculture 0 1,790 4,466 10,261 16,517 0% 11% 27% 62% Infrastructure 10,951 2,269 6,152 12,679 32,050 34% 7% 19% 40% Land & Natural 2,558 3,120 9,355 5,182 20,214 13% 15% 46% 26% Resources Source: MINECOFIN (in 2007 real prices). Note: Revised and executed source funding data not available for 2004 and 2005. As table 4.33 shows, in 2007 the health sector received the greatest percentage of its total actual development budget from donors--90 percent of its total development funds. That translated into RwF 23.1 billion in absolute terms. Infrastructure and agriculture received 43 percent and 45 percent, respectively, from donors. Education donor funding dropped to 15 percent of its total development budget in 2007, compared with 25 percent in 2006. In terms of executed internally sourced financing in 2007, land and natural resources and education received the highest percentage of their development budgets from FI funding at 29 percent (RwF 7.5 billion) and 64 percent (RwF 12.7 billion), respectively. In the health sector, actual FI financing was minimal at 4 percent, while infrastructure and agriculture received 21 percent and 14 percent, respectively, from this type of domestic funding. Furthermore, agriculture and infrastructure were the two sectors with high portions of funding from loans, at 34 percent and 31 percent of their respective total development funds for executed development expenditures in 2007, down significantly from 62 percent and 40 percent, respectively, in 2006. In absolute terms, agriculture received RwF 4.5 billion in loans and infrastructure received RwF 16.1 billion in loans in 2007. Taking a closer look at original budget data for each priority sector, the allocation between total internal funding (CP and FI) and total external funding (donors and loans) from 2004 to 2007 can be identified. In the education sector, the amount of internal funding significantly increased as a portion of its total original budget for development. As table 4.34 and figure 4.11 show, sourcing went from 91 percent external and 9 percent internal in 2004 to 34 percent external and 66 percent internal in 2007. ng Reconstruction to Re Budgetin for Effectiveness in Rwanda: From R eform 95 nternal and Exte Table 4.34. In ent r ernal Developme Funding for MINEDUC 2004­07 REAL Terms CP&FI as % of Don& Loan as in million RwF CP&FI DON & LOAN Total Edu % of Total Edu Edu 2004 1,050 11,189 9% 91% Edu 2005 3,798 6,705 36% 64% Edu 2006 3,937 7,290 35% 65% Edu 2007 13,081 6,781 66% 34% OFIN (in 2007 real prices). Source: MINECO ternal Developm Figure 4.11. Internal and Ext or 04­07 ment Funding fo MINEDUC 200 In million RwF OFIN (in 2007 real prices). Source: MINECO sion between in The divis nternal and exteernal funding in ctor remained n the health sec nt in 2004­07 (se fairly constan d figure 4.12). ee table 4.35 and 96 nk World Ban Working Paper ernal and Extern Developmen Funding for M Table 4.35. Inte nal nt MINISANTE 2004­07 REAL Terms CP&FI as % of Don& Loan as in million RwF CP&FI DON & LOAN Tot Health % of Tot Health Health 2004 346 9,987 3% 97% Health 2005 1,972 19,497 9% 91% Health 2006 960 19,455 5% 95% Health 2007 1,123 24,561 4% 96% FIN (in 2007 real pr Source: MINECOF rices). ternal and Exter Figure 4.12. Int ent 04­07 rnal Developme Funding for MINISANTE 200 In million RwF FIN (in 2007 real pr Source: MINECOF rices). External fu unding in the health sector increased cons e sistently in the period in absolute terms, , but the portion evelopment fund n of its total de ding coming fro om external rs ), sources (donor in particular) remained at 91­97 percent. Internal financ cing in this sector was min nimal. Agriculture received a majority of its fu e unding from ext as ternal sources a well, but with less conce ternal funding t entration in ext h ble than the health sector (see tab 4.36 and figure 4.13). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 97 Table 4.36. Internal and External Development Funding for MINAGRI 2004­07 REAL Terms CP&FI as % of Don& Loan as in million RwF CP&FI DON & LOAN Tot Agri % of Tot Agri Agri 2004 * * n.a. n.a. Agri 2005 942 7,034 12% 88% Agri 2006 934 8,783 10% 90% Agri 2007 2,876 10,641 21% 79% Source: M INECOFIN (in 2007 real prices). * Note: Because of a problem in the 2004 Original Budget Law, data for agriculture are omitted. Figure 4.13. Internal and External Development Funding for MINAGRI 2005­07 12,000 10,000 in million RwF 8,000 6,000 4,000 2,000 0 Agri 2005 Agri 2006 Agri 2007 CP&FI DON & LOAN Source: MINECOFIN (in 2007 real prices). Note: Because of a problem in the 2004 budget law regarding agriculture, 2004 data are not shown. As is shown above, agriculture's external funding as a portion of its total original development budget fluctuated somewhat over the period, between 79 percent and 90 percent, although in absolute terms it saw an increase of its external funds in absolute terms, from RwF 7.0 billion in 2005 to RwF 10.6 billion in 2007. From 2004 to 2007, both internal and external funding for the infrastructure development budget increased significantly (see table 4.37 and figure 4.14). External sources remained the dominant source of funding during the period, accounting for 67 percent to 78 percent of the original development budget. 98 nk World Ban Working Paper ernal and Extern Developmen Funding for M Table 4.37. Inte nal nt ­07 MININFRA 2004­ REAL Terms CP&FI as % of Don& Loan as in million RwF CP&FI DON & LOAN Tot Infra % of Tot Infra Infra 2004 5,490 19,458 22% 78% Infra 2005 15,127 31,127 33% 67% Infra 2006 10,865 23,734 31% 69% Infra 2007 13,472 38,653 26% 74% FIN (in 2007 real pr Source: MINECOF rices). ternal and Exter Figure 4.14. Int ent 4­07 rnal Developme Funding for MININFRA 2004 In million RwF FIN (in 2007 real pr Source: MINECOF rices). Land and n natural resource crease in interna es experienced a significant inc al financing over the periodd, from 8 percennt of its original development budget in 2004 to o 36 percent ble 4.38 and figu in 2007 (see tab ure 4.15). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 99 Table 4.38. Internal and External Development Funding for MINITERRE 2004­07 REAL Terms CP&FI as% of Don & Loan as in million RwF CP&FI DON &LOAN Tot Land&Res % of Land&Res Land&Res 2004 977 11,284 8% 92% Land&Res 2005 3,058 8,191 27% 73% Land&Res 2006 3,149 10,162 24% 76% Land&Res 2007 9,202 16,701 36% 64% Source: MINECOFIN (in 2007 real prices). Figure 4.15. Internal and External Development Funding for MINITERRE 2004­07 18,000 16,000 14,000 12,000 10,000 in million RwF 8,000 6,000 4,000 2,000 0 Land&Res 2004 Land&Res 2005 Land&Res 2006 Land&Res 2007 CP&FI DON & LOAN Source: MINECOFIN (in 2007 real prices). Overall, external financing in the land and natural resources original development budget increased in absolute terms from RwF 11.3 billion in 2004 to RwF 16.7 billion in 2007. Internal financing increased significantly more, from RwF 977 million to RwF 9.2 billion, representing about onethird of land and natural resources' total original development funding in 2007. For the remaining, nonpriority ministries, their actual development expenditures combined accounted for 29 percent of Rwanda's overall development expenditures in 2005 and 23 percent in 2006, or in nominal terms terms RwF 35.4 billion in 2005 and RwF 35.6 billion in 2006. In real terms (2007 prices), this was RwF 44.9 billion in 2005 and RwF 39.9 billion in 2006, increasing to RwF 44.8 billion by 2007. In 2007, these ministries' development spending represented 25 percent of the country's total development spending, consistent with the previous year. Although some specific data for 2004 executed development spending for nonpriority sectors are missing, data show that priority sectors were allocated 55 percent of total overall development spending that year, indicating that in 2004 nonpriority sectors captured the remaining 45 percent of actual development spending. Hence, from 2004, there was a significant 100 World Bank Working Paper reallocation in development spending, with a decreasing portion going to nonpriority sectors--from 60 percent in 2004 to 25 percent in 2007. Table 4.39 summarizes the development spending for these nonpriority ministries. Table 4.39. Overall Actual Development Expenditure Trends in Nonpriority Ministries 2004­07 RwF) Development Budget % of total Dev . Spending % of total Dev. Spending % of total Dev. Spending % of total Dev. Spending Real Terms (in million Increase RwF 2005-07 Increase % 2005-07 Actual 2007* Actual 2004 Actual 2005 Actual 2006 01 - PARLIAMENT data missing n.a. 28 0% 0 0% 0 0% ­28 ­100% 04 - PRIMATURE 4 0% 0 0% 1,492 1% 2,454 1% 2,454 n.a. 05 - COURS.SUP data missing n.a. 2,414 2% 795 0% 674 0% ­1,740 ­72% 06 - MINADEF data missing n.a. 803 1% 713 0% 1,060 1% 257 32% 07 -MININTER 259 0% 1,548 1% 2,157 1% 2,484 1% 936 60% 08 - MINAFFET data missing n.a. 0 0% 0 0% 1,115 1% 1,115 n.a. 10 - MINICOM 532 0% 713 0% 3,774 2% 5,357 3% 4,644 651% 12 - MINECOFIN data missing n.a. 12,555 8% 7,131 4% 5,799 3% ­6,756 ­54% 13 - MINIJUST data missing n.a. 2,403 2% 4,411 3% 3,555 2% 1,152 48% 15 - MIJESPOC data missing n.a. 107 0% 999 1% 1,313 1% 1,206 1122% 19 - MIGEPROF data missing n.a. 4,407 3% 0 0% 0 0% ­4,407 ­100% 20 - MIFOTRA data missing n.a. 139 0% 1,550 1% 1,919 1% 1,780 1285% 23 - MINALOC 4,847 4% 19,788 13% 16,903 10% 19,048 10% ­740 ­4% TOTAL Non- Priority n.a. 45% 44,906 29% 39,925 23% 44,778 25% ­128 0% Sectors Total Overall DB 110,508 100% 153,131 100% 170,265 100% 181,870 100% 28,738 19% Source: MINECOFIN (in 2007 real prices). *2007 Revised Development Budget. As table 4.39 indicates, MINALOC was by far the largest nonpriority sector development spender, with 13 percent, 10 percent, and 10 percent of total development spending in 2005, 2006, and 2007, respectively. It is noteworthy however, that MINALOC is responsible for social protection agenda in the country (and this sector, although just recently viewed as a sector in its own right, is reviewed in further detail in chapter 5). This was up from a mere 4 percent in 2004. MINALOC was followed by MINECOFIN, the Ministry of Commerce, Industry, Investment Promotion, Tourism, and Cooperatives (MINICOM), and MINIJUST, the only other ministries with development spending in excess of RwF 3 billion in 2007. Although MINECOFIN saw its portion of the total development budget decrease from 8 percent to 3 percent between 2005 and 2007, MINICOM saw its portion increase from less than 1 percent in 2005 to 3 percent in 2007. MIGEPROF also had an executed development budget in 2005 in excess of RwF 3 billion; however, in 2006, this ministry was moved under PRIMATURE. This restructuring was not fully reflected in a corresponding increase in PRIMATURE's development spending, which was higher in 2006 than in 2005, but at only onethird of MIGEPROF's pervious development budget. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 101 Table 4.40 highlights the executed (actual) development budgets between 2004 and 2007 for the four largest (in terms of their development spending) nonpriority sectors. Table 4.40. Overview of Actual Development Budgets of Four Major Nonpriority Ministries 2004­07 In Real Terms Actual DB Actual DB Actual DB Actual DB Growth % in million RwF 2004 2005 2006 2007* 2005-07 23 - MINALOC 4,847 19,788 16,903 19,048 ­4% 12 - MINECOFIN data missing 12,555 7,131 5,799 ­54% 13 - MINIJUST data missing 2,403 4,411 3,555 48% 10 - MINICOM 532 713 3,774 5,357 651% Total 4 Largest Non-Priority n.a. 35,459 32,220 33,759 ­5% Portion of Total Non Priority DB n.a. 79% 81% 75% n.a. of which MINALOC n.a. 44% 42% 43% n.a. Source: CEPEX/MINECOFIN (in 2007 real prices). Note: DB = Development Budget *2007 Revised Development Budget As table 4.40 shows, of the actual development spending by all of the nonpriority ministries, MINALOC accounted for 44 percent in 2005, 42 percent in 2006, and 43 percent in 2007. Measured in real terms, there was a significant increase in MINALOC's actual development spending between 2004 and 2007, from RwF 4.8 billion to RwF 19.0 billion. MINECOFIN saw its spending decrease significantly between 2005 and 2007, while MINICOM experienced a significant increase from a mere RwF 532 million in 2004 to RwF 5.4 billion in 2007. Similarly, MINIJUST saw a significant increase in the period as well, with its development spending increasing by 48 percent. Further details about development budget programs and projects in both priority sectors and nonpriority sectors, for actual, revised, and original budget data, can be found in Data Sets 10­18. Notes 1 Some inconsistencies between chapter 3 and section 4 of this report exist although, generally, they indicate the same trends across years. Whenever possible, these differences are explained in the text. For further details about data provided by MINECOFIN, please refer to the referenced data sets of this report. 2 The government adopted a National Decentralization policy in May 2000 to achieve three main goals: (i) good governance, (ii) propoor service delivery, and (iii) sustainable development. Implementation has occurred in three phases. Phase one (200003) established democratic and community development structures and enhanced their capacities through legal, institutional, and policy reforms. These covered roles and responsibilities, financing of services, and mechanisms to ensure accountability. Phase two (200408) consolidated and expanded decentralization by emphasizing local service delivery through wellintegrated, accountable networks that empower communities to participate more fully in planning and managing local affairs. Phase three (post2008) will continue to improve, support and sustain the efforts of phases one and two. In 2006, as part of the decentralization process, local governments were restructured 102 World Bank Working Paper from previous 12 provinces into 5 provinces, 30 Districts, 416 Sectors and 2,148 cells, covering an estimated total of 15,000 villages across the country with implications for the structure of the Organic Budget Law. In addition, during this period, MINECOFIN worked at strengthening its overall budgetary processes through various budget reforms. Therefore, analysis of public expenditure transfers from central government to provinces and districts over this period presents some challenges and requires some assumption when comparing expenditures year overyear. 3 Based on MINECOFIN data. Note that the data are not consistent with all sources provided by the government of Rwanda. For example, the fiscal table provided in donors' package, JBSRV April 2007, shows slightly different figures. 4 Based on MINECOFIN data. Note that the data are not consistent with all sources provided by the government of Rwanda. For example, the fiscal table provided in donors' package, JBSRV April 2007, shows slightly different figures. 5 Note that MINECOFIN sometimes reports these figures in different formats, for example, sometimes including domestic capital only, sometimes total capital, sometimes excluding debt or other classification. It is also not clear based on reporting how MINECOFIN classifies these economic expenditures; hence, it is not always clear what is included in each category. The assumption here is that classifications are consistent across the period. 6 Based on classifications of districts' programs in the budget. For details, please refer to data sets 6, 7, 8, and 9. CHAPTER 5 Policy Overview, Resource Allocation, and Results T his chapter provides a detailed summary of the most recently completed analytical work assessing the policy, resource allocation, and results in all key sectors. The core of this chapter is derived from findings from public expenditure reviews (PERs) that cover the following priority sectors: agriculture, education, health, social protection, and water and sanitation. (Water and sanitation services [WSS] come under the land and natural resources line ministry, but here they are considered as an infrastructurerelated priority sector.) Infrastructure, and in particular the energy and transport subsectors, has yet not conducted a PER; however, a snapshot of these sectors is provided in this chapter. The objective of this chapter is to ensure that the overall indepth analysis conducted on behalf of these key ministries (a) provides an enhanced understanding of the country's overall public expenditure, (b) helps put each indepthsector analysis into the context of the overall budget allocation considerations, and (c) enhances the overall analysis by line ministry as presented in this report in chapters 3 and 4. Some discrepancies between this chapter 5 report data and the data in other chapters of this report can often be explained and, to the extent possible, are noted in the endnotes or in the overall text. However, at times underlying assumptions in the analytical approaches provide differences in data and for sectorspecific results. This report refers to both the extensive data sets in this report (Excel spreadsheets that provide the basis for the bulk of the analysis in this report) and the source documentations used in the PER reports (which can be requested from the specific ministries). Finally in this chapter, a brief snapshot of all nonpriority sectors is provided to allow a quick overview of all sectors' 2004­07 recurrent and development spending.1 As this chapter focuses on an overview of individual sector policy, resource allocations, and results, it could prove helpful to first put the sectorspecific reviews into the context of Rwanda's overall progress to date on poverty and on its MDGs. Below follows a summary of Rwanda's poverty profile, demographic profile, and the status of Rwanda's pursuit to reach its MDGs. First, data on the geographic distribution (urban/rural) show that poverty levels in rural areas are much higher than in urban areas. Poverty in Rwanda remains disproportionately a rural phenomenon, as nearly 92 of the poor live in rural areas. 103 104 World Bank Working Paper Second, the share of the different quintile groups by geographical distribution also illustrates inequality. On the one hand, 73 percent of the city of Kigali population is in the richest quintile group, and other urban households are generally disproportionately located in the higher quintile groups. On the other hand, rural households are overrepresented in lower quintile groups. Third, the proportion of the population of the Southern and Western provinces that is in the lowest quintile is high compared with other provinces. Conversely, in the eastern province, there are fewer households in the lowest quintile group and more in the middle and upper quintiles. More than 60 percent of households in the city of Kigali province are in the richest quintile. Table 5.1. Poverty Indicators and Geographical Distribution Poverty Indicators Income quintile Poverty Headcount Share of Poorest 2nd 3rd 4th Richest (share of population) the Poor Quintile Quintile Quintile Quintile Quintile By Stratum Kigali 13.0% 1.7% 1.9% 5.1% 6.5% 13.5% 72.9% Other Urban 41.0% 6.7% 12.1% 15.0% 16.4% 20.3% 36.2% Rural 62.5% 91.6% 22.5% 21.9% 21.6% 20.6% 13.5% By Province City of Kigali 20.2% 3.4% 3.8% 8.3% 8.5% 163.0% 63.1% Southern 67.3% 30.2% 26.2% 24.9% 18.8% 16.7% 13.3% Province Western 62.0% 26.3% 23.1% 21.6% 20.3% 20.5% 14.6% Province Northern 62.7% 20.3% 22.5% 21.1% 22.4% 20.1% 13.9% Province Eastern 50.4% 19.7% 14.4% 16.9% 24.0% 24.9% 19.9% Province National 56.90% Source: EICV II (20052006). Rwanda's demographic situation has been characterized by rapid population growth: it has increased steadily and rapidly from more than 2,000,000 in 1952, to 7,666,000 in 1996 and to 8,128,553 in 2002. In 2007, the population of Rwanda was estimated at approximately 9 million (Rwanda General Population and Housing Census 2002). The population is predominantly young, with 67 percent of all Rwandese under the age of 20. In terms of gender, females are the majority (52 percent). The fertility rate was estimated at 6.1 children per woman for the country as a whole in 2005, but there were clear differences by geographic distribution. Women in urban areas had a lower fertility rate (4.9) than those in rural areas (6.3). Recently, Rwanda reviewed its progress toward its MDGs to attempt to determine what is likely and unlikely to be achieved. Despite much progress to date in key priority areas, an assessment at the midway point between the MDG adoption in 2000 and the 2015 target date shows that Rwanda, like many other SubSaharan African countries, is struggling to achieve a number of goals. Although there were significant Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 105 gains and Rwanda has already achieved some of the MDGs--gender equality in primary school enrollment and reducing the level of HIV/AIDS, for example--there has not been sufficient progress in reducing extreme poverty. Some success, however, can be noted. For example, the country is on track on MDGs for preventing the rise of malaria incidence and promoting gender equality. However, reducing extreme poverty, mortality rates of children under age five, and stunting due to child malnutrition remain extremely challenging. The government has determined that three of the seven MDGs are achievable under current conditions. Below are highlights of the status of each MDG. MDG 1: Eradicate extreme poverty and hunger. Although initiatives have begun, the percentage of people living under the poverty line is still significantly high at 56.9 percent in 2006, down from 60.4 percent in 2000. It is unlikely that the country will reduce the percentage to the target of 30.2 percent by 2015. To achieve this target, Rwanda would have to reduce poverty by 26.7 percentage points between 2006 and 2015 against the 3.5 percentage point reduction achieved between 2000 and 2006. Malnutrition among children under five remains relatively high in Rwanda: in 2005, more than one in every five children (22.5 percent) suffered from moderate or severe malnutrition compared with 24 percent in 2000, and 4.4 percent of children suffered from severe malnutrition, based on the weight forage indicator, against targets of 2 percent by 2015. MDG 2: Achieve universal primary education. There was progress toward universal primary education, with net enrollment increasing from 72 percent in 2000 to 95 percent in 2006. Completion rates have also increased from 22 percent to 51 percent during the same period. The country is on track to achieve this goal. MDG 3: Promote gender equality. The gender gap in primary education was eliminated. Success was registered in reducing the gender gap in literacy, from 10 percent in 2000 to 0.2 percent in 2006. Females holding over 50 percent of seats in Parliament2, above the target of 50 percent by 2015. On this goal, the country has reached its goal. MDG 4: Reduce child mortality. Under5 mortality rates dropped from 152 per 1,000 live births in 2005 to 103 per 1,000 in 2007. While an improvement yearoveryear, significant impact is still needed to move toward the objective of a twothirds reduction by 2015. This goal is not likely to be achieved. On a positive note, however, child immunization against measles (percentage of 11­23 montholds) was 90.4 percent in 2007. MDG 5: Improve maternal health. Maternal mortality rate per 100,000 births fell from 1,071 in 2000 to 750 in 2006. Although there has been progress and the government believes the goal is 106 World Bank Working Paper achievable, there is still a challenge to attain a further reduction of 482 in order to meet the target of 268 per 100,000 births by 2015. The likelihood that this goal will not be met is high. MDG 6: Combat AIDS, malaria, and other diseases. Significant progress was made in the area of HIV/AIDS. HIV/AIDS prevalence was reduced from about 9 percent in 2000 to 3 percent in 2006. Malariarelated mortality has also been reduced from 51 percent in 2000 to 26 percent in 2006. This goal is achievable. MDG 7: Ensure environmental sustainability. The status of most indicators for this goal is that it is still too early to tell. Data are available for one indicator: access to safe, clean drinking water. The percentage of people with access to safe water remained unchanged around 64 percent between 2000 and 2006, against the target of 82 percent by 2015. Most likely, this goal will not be achieved. It is in this overall context that the following detailed summary of the most recently completed analytical works assessing policy reforms, resource allocation, and results in all key sectors should be considered. Priority Sectors Agriculture Agriculture Sector Overview Several ministries are involved in agriculturerelated activities. In addition to MINAGRI (which was reorganized in 2005 to improve its core functions) and its semiautonomous agencies (Rwanda Agriculture Development Authority [RADA], Rwanda Animal Resources Development Authority [RARDA], and Agricultural Research Institute of Rwanda [ISAR]), other ministries with agriculturerelated activities include MINECOFIN, MINITERRE, MINICOM (Rwanda Coffee Development Authority [OCIRCoffee], Rwanda Tea Development Authority [OCIR Tea], RIEPA [in charge of investments and export promotion and now under the Rwanda Development Board], Rwanda Bureau of Standards [RBS] [in charge of certification and quality control]), and MINALOC. The agricultural policy framework is set forth in three main policies: the agricultural Poverty Reduction Strategic Plan (PRSP), the National Agriculture Policy (NAP), and the Strategic Plan for Transformation of Agriculture (SPTA). At the time of the latest PER of the agricultural sector, the agriculture sector supported four main programs and 16 subprograms, shown in table 5.2. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 107 Table 5.2. Agriculture Programs and Subprograms Intensification and Development of Sustainable Production Systems Subprogram #1: Sustainable Management of Natural Resources Subprogram #2: Development of Animal Sector Subprogram #3: Marshland Development Subprogram #4: Irrigation Development Subprogram #5: Supply and Utilization of Inputs Subprogram #6: Food Security and Vulnerability Management Support to Professionalization of Producers Subprogram #1: Promoting Farmers' Organizations and Capacities Subprogram #2: Restructuring of Local Service Provisions Subprogram #3: Rural Finance Systems and Credit Development Promotion of Commodity Chains and Development of Agribusiness Subprogram #1: Promotion and Development of Commodity Chains Subprogram #2: Improvement of Competitiveness of Products Subprogram #3: Rural Infrastructure Development Support to Producers Subprogram #4: Promotion of Entrepreneurship Institutional Development Subprogram #1: Legal and Regulatory Framework Subprogram #2: Reforms and Institutional Support to Public Services Subprogram #3: Coordination, Monitoring, and Evaluation of Sector Source: Agriculture PER Report, 2007. The two largest activities with the largest funding are crops production and livestock production. Crop production in Rwanda benefits from its varying agroclimatic zones that are good for producing a wide range of crops. A total of 1.6 million hectares of farm land are planted with five types of crops. Table 5.3 shows the distribution of the land by types of crops. Table 5.4 shows the relative contribution of each of the four main types of crops to total agricultural output. Table 5.3. Farm Land Distribution by Types of Crop Type of Crops: Distribution Roots & tubers 26% Legumes 25% Bananas 22% Cereals 21% Fruits & Vegetables 5% Source: Agriculture PER Report, 2007. 108 World Bank Working Paper Table 5.4. Proportion of Total Agricultural Output Contributed by Four Major Food Groups 2000­05 2000 2001 2002 2003 2004 2005 Average Cereals 4.0% 4.9% 4.3% 4.3% 4.7% 5.7% 4.7% Legumes 4.3% 4.5% 4.1% 4.2% 3.6% 3.4% 4.0% Bananas 38.1% 31.7% 39.2% 35.4% 36.6% 34.9% 36.0% Roots & Tubers 49.9% 54.5% 49.0% 45.6% 44.0% 43.1% 47.7% Source: Agriculture PER Report, 2007, page 14. Original Source: MINAGRI/NEPAD 2006. The four crop classifications, according to the NAP, are ExportOriented Crop (coffee, tea, pyrethrum) Regional Export and ImportReducing Crops (Irish potatoes, rice, wheat, maize, soybean) Traditional Food Crops (banana, cassava, sweet potatoes, beans) New HighPotential Chains (passion fruit and macadamia) As part of the government's commoditychain focus, special emphasis is given to promotion of cereals, in particular rice development, through key programs such as Project Recovery of Emergency Agricultural Production3 (PRAUPA), the Rural Sector Support Project (RSSP), and farmers' organizations such as Union of Rice Cooperatives in Rwanda (UCORIRWA)4. The National Rice Development Program estimated that US$590 million in funding is needed between 2007 and 2016.5 Maize production is low compared with household and processing industry needs. In 2005, the government launched a national banana and cassava seed multiplication program, which was deemed successful at the 2007 review. Livestock production benefits from good natural conditions but the small amount of land, poor standards of husbandry, and local species' low productivity levels constrain potentials. Table 5.5 shows the evolution of national herds and flocks in Rwanda between 2000 and 2005. Table 5.5. Evolution of National Livestock Population 2000­05 2000 2001 2002 2003 2004 2005 Cattle 755,123 814,124 960,450 991,697 1,006,572 1,079,206 Goats 756,502 916,753 919,785 1,270,903 1,263,962 266,355 Sheep 232,724 266,539 300,600 371,766 686,837 689,556 Pigs 177,220 197,081 207,783 211,918 326,652 456,041 Poultry 2,043,077 1,277,706 1,055,644 2,432,449 2,482,124 2,943,703 Rabbits 338,616 495,290 488,629 498,401 643,927 565,696 Source: Agriculture PER Report, 2007, page 15. Original Source: World Bank 2007. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 109 The three livestock classifications are ExportOriented Products (hides and skins) DomesticOriented Markets and Export Potential/Import Reducing (meat, milk, poultry, eggs) Fish Production The government, with the help of NGOs, has focused on increased milk production, which has proven successful. For example, the One Cow for Every Poor Family program, under RARDA, is part of this effort. Still, the country faces milk shortages (shown by the low milk consumption per capita of 12 liters per year), and more funding is required. A limited market for milk commodities also limits the milk industry. Subsidies exist in the agricultural sector in various forms.6 For example, The government is encouraging use of fertilizers by buying inputs in bulk and then distributing them to rural cooperatives for sale at subsidized prices. Funds are allocated to MINAGRI's development budget from government and STABEX (an EU body that makes compensatory payments when export prices fluctuate) receipts, and payments by cooperatives and communities are deposited into a revolving fund in Rwanda Development Bank (BRD) to be used by private sector fertilizer importers. The estimated total subsidy in 2007 was RwF 1.2 billion. The government plans to reduce this subsidy over the next five years. Favorable tax treatment is given to the sector, as commercial agriculture is classified as a priority sector. Table 5.6 shows the available tax subsidies. Table 5.6. Tax Subsidies in the Commercial Agriculture Sector · Tax-free investment allowance of 50 percent · Tax-free imports of farm equipment · Tax exemption on fertilizers (since 2000) · VAT exemption of all inputs and outputs · Export levies on all coffee exports (up to 3 percent of the free-on-board [f.o.b.] value) Source: Agriculture PER Report, 2007. However, agriculture contributes very little to taxes as most farmers are outside the formal economy and the majority of the poor earn their livelihoods from agriculture (Foreign Investment Services 2006). Agricultural inputs are important for overall output. In Rwanda, past poor performance in agriculture, as measured by low productivity, is partly a consequence of very limited use of improved seeds and fertilizers. Increased use of inputs are restricted by Financial factors (high price versus low purchasing power among farmers) Inadequate access to credit Weak researchextension links 110 World Bank Working Paper Limited private sector supply and distribution systems Agricultural markets in Rwanda face the following restrictions:7 Market integration is poor for both crop and livestock. Sixtysix percent of food production is for subsistence use. Only 34 percent of food production reaches markets. Only 23 percent of livestock is commercialized and only 4 percent processed. Performance in the agricultural sector is further constrained by the following challenges: Massive soil erosion Rainfed agriculture Weak agricultural research and extension systems Low productivity Low level of value addition Insufficient and lack of access to financing Weak farmers' organization A food security review in Rwanda will help to put the importance of the agriculture sector in perspective, not only from a GDP perspective, but also from a basic poverty reduction view. A food security review would assess availability and accessibility of food and the stability of the food supply. Per capita nutritional needs in Rwanda are8 · Energy: 2,100 kcal per day · Proteins: 59 g per day · Lipids: 40 g per day According to FAO minimum per capita annual animal products consumption recommendations are · Recommended: Milk 220 liters per year · Recommended: Meat 50 kg per year · Rwanda's Actual: Milk 12 liters per year (5.5 percent of recommendation) · Rwanda's Actual: Meat 4.8 kg per year (9.6 percent of recommendation) Food insecurity has improved since 2000, but seasonal and regional food shortages resulting from poor rainfall, lack of seeds and other inputs, and disease outbreaks continue to occur. Furthermore, a clear deficit between recommended minimum nutritional standards and actual nutrition consumption exists. In 2006, MINAGRI contracted with a team of international and local consultants9 to conduct a detailed review of the agriculture sector covering 2000­07. Because the data made available on actual expenditures was inconsistent, the resulting PER report10 of the agriculture sector was based primarily on budgeted data. In short, key conclusions of the agriculture PER report can be summarized as follows: The agriculture sector · contributed more than onethird of the GDP Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 111 · provided employment to more than 80 percent of the working population · accounted for more than 70 percent of merchandise exports · helped to meet food requirements for the bulk of the population (83 percent living in rural areas) MINAGRI had one of the lowest budget allocations of all ministries (9th out of 13 in 2007) and its 2007 allocation was lower, in real terms, than in 2000. Actual expenditure showed, in general, a high level of execution of the budget. MINAGRI's budget as proportion of agricultural GDP is low, declined noticeably from 2004 to 2007, and ranked far below average for developing countries (including SubSaharan Africa). MINAGRI's recurrent budget funding in districts did not appear to be guided by any consideration for addressing relative poverty levels. MINAGRI's increased allocation to districts showed its commitment to decentralization efforts. Development partners played a key role in sector funding (about 50 percent of the total budget and 86 percent of the development budget); however, the government provided an increasing amount of the funding. Offbudget support was significant: in 2007, offbudget support exceeded on budget support by 6 percent. Total funds provided to MINAGRI (including related activities in other sectors) were still below the Comprehensive African Agricultural Development Program's (CAADP) commitments by 10 percent (even when offbudget is added). The balance of recurrent and development budgets was healthy, with recurrent spending accounting for 25 percent of the total, and less than 50 percent on average of the recurrent budget being used for wages and salaries. Distribution of funding between PSTA programs and subprograms seemed to be optimized. Few direct subsidies were provided in the sector. The most significant current example of subsidy was the fertilizer program. Further decentralization will require increased involvement in the process of budget and financial management by the districts. Putting the agriculture sector in the macroeconomic context, the PER identified the following trends: The agriculture sector provided 37 percent of Rwanda's GDP in 2006 (of which food crops were 85 percent and livestock 5 percent), while services contributed 41 percent of GDP and the industrial sector contributed 15 percent. In recent years, agriculture's contribution to GDP has fluctuated, but no trends are apparent: · 38 percent in 1999 · 35 percent in 2002 · 39 percent in 2004 Vision 2020 aims at changing the GDP contributions to 42 percent services, 26 percent industrial, and 33 percent agriculture. Specifically for the agriculture sector, 112 World Bank Working Paper Vision 2020 aims to decrease the portion of the population dependent on agriculture employment from 80 percent to 50 percent, and increase the share of farmland using modern technologies from 3 percent to 50 percent. Agriculture Sector Public Expenditure Allocation Review Overall recurrent and development trends in MINAGRI's budget allocations (not actual expenditures) over the years 2000­07 are illustrated table 5.7: Table 5.7. MINAGRI Budget Allocations 2000­07 Increase in % 2000-07 W N-W RT DB Overall In Real Terms (in mill RwF) 22.0% 99.5% 54.5% 7.6% 16.3% Recurrent Share of % Non- Overall YEAR Wages Wage** TOTAL Develop TOTAL W N-W RT DB Overall 2000 1,667.1 1,199.9 2,867.0 12,559.2 15,426.2 10.8% 7.8% 18.6% 81.4% 100% 2001 1,728.4 2,029.4 3,757.8 19,422.0 23,179.7 7.5% 8.8% 16.2% 83.8% 100% 2002 1,650.4 2,378.1 4,028.5 20,865.2 24,893.7 6.6% 9.6% 16.2% 83.8% 100% 2003 1,565.7 2,141.5 3,707.2 11,616.6 15,323.8 10.2% 14.0% 24.2% 75.8% 100% 2004 1,250.5 3,409.8 4,660.3 13,709.1 18,369.4 6.8% 18.6% 25.4% 74.6% 100% 2005 1,432.1 4,544.8 5,976.9 10,105.2 16,082.1 8.9% 28.3% 37.2% 62.8% 100% 2006 1,755.9 3,416.1 5,172.0 9,717.6 14,889.6 11.8% 22.9% 34.7% 65.3% 100% 2007 2,034.7 2,394.2 4,429.0 13,517.0 17,945.9 11.3% 13.3% 24.7% 75.3% 100% Source: Agriculture PER Report, 2007, page 33. Original Source: Analysis of data contained in MINECOFIN, 2006(a), Republic of Rwanda 1999­2005 and annex 6.1. Note: W=Wages; NW=Nonwages; RT=Recurrent; DB=Development Budget. *Includes wages, salaries, allowances and other benefits and employer contributions including social protection (SSF and Healthcare); budget codes 331334, 111112, 115116, 121 and 123. **Includes telecommunications, office equipment and supplies, rent, car hire, building maintenance. From table 5.7, the following trends measured in real terms in MINAGRI's budget allocation over the period can be identified: Wages increased between 2000 and 2007, from RwF 1.7 billion to RwF 2 billion, representing a 22 percent increase in eight years. Nonwages in the period 2000­07 increased from RwF 1.2 billion to RwF 2.4 billion, representing a 99.5 percent increase in eight years. Total recurrent expenditures, including wages and nonwages, increased by 54 percent in the period. The development budget doubled from RwF 12.6 billion to RwF 13.5 billion, a 7.6 percent increase between 2000 and 2007. Distribution in the recurrent and development budgets measured in nominal terms changed. Recurrent expenditure increased by 187.5 percent, and development expenditure increased by 100 percent over the eightyear period, which contributed to an overall 116.5 percent increase in MINAGRI's budget allocation, or an average 14.6 percent increase per year. Yet, the allocation of these funds changed over time. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 113 Recurrent expenditures for MINAGRI in 2000 were 18.6 percent of total MINAGRI expenditures, peaked at 37.2 percent in 2005, and stood at 24.7 percent in 2007. Development expenditures for MINAGRI in 2000 were 81.4 percent of total MINAGRI expenditures, with the lowest budget allocation in 2005 and 2006 at 62 percent and 65 percent, respectively, and stood at 75.3 percent in 2007. Table 5.8 compares MINAGRI spending and total public spending. Table 5.8. MINAGRI and National Budget Allocations 2000­07 MINAGRI National MINAGRI/ Agr/Nat'l Agr/Nat'l YEAR (Recur&Dev) (Recur&Dev) Nat'l (%) Recur (%) Dev (%) 2000 15,426.2 314,854.6 4.9% 1.5% 10.2% 2001 23,179.7 334,793.8 6.9% 1.6% 18.9% 2002 24,893.7 492,658.0 5.1% 1.3% 11.4% 2003 15,323.8 394,395.6 3.9% 1.3% 11.4% 2004 18,369.5 462,583.0 4.0% 1.4% 11.4% 2005 16,082.1 475,257.3 3.4% 1.8% 7.2% 2006 14,889.7 454,296.9 3.3% 1.7% 6.8% 2007 17,946.0 506,745.1 3.5% 1.4% 7.4% Average: 4.4% 1.5% 10.6% Source: Agriculture PER Report, 2007, page 34. Original Source: Analysis of data contained in MINECOFIN, 2006(a), Republic of Rwanda 19992005 and annex 6.1. Table 5.8 highlights the following: MINAGRI's combined proportion of the national budget fluctuated in the eightyear period, from as low as 3.3 percent in 2006 to as high as 6.9 percent in 2001 (including both recurrent and development). MINAGRI's recurrent allocation in 2007 of 1.4 percent of the total recurrent budget was around the average for the period of 1.5 percent. Over the years, the budget proportion allocated to MINAGRI declined. MINAGRI's portion of the development budget in the 2000­07 period decreased from 10.2 percent in 2000 to only 7.4 percent in 2007, showing an overall decrease in the share of the country's development budget allocated to MINAGRI. In each of the last three years of the period, 2005­07, MINAGRI's allocations were below the period's average of 10 percent. 114 World Bank Working Paper Only MINAGRI's proportion of the recurrent allocation was somewhat consistent over the period, with a peak of 1.8 percent in 2005 and a low of 1.3 percent in 2002 and 2003. In 2007, the recurrent allocation was 1.4 percent, close to the average for the period of 1.5 percent. The proportion of overall expenditure allocation in the agriculture sector alters only slightly when including other ministries' spending in the sector.11 Also, as table 5.9 shows, when district allocations to agriculture are added for 2007, the overall budget allocation to MINAGRI increases only slightly to 3.9 percent as compared with 3.5 percent without district allocations, as shown in table 5.8. Table 5.9. Budget Allocations to MINAGRI and Other Key Ministries 2007 In Real Terms (in mill RwF) Ministry Recurr. Develop. Total % Share Presidency 8,602.3 0.0 8,602.3 1.7% Defense 32,923.9 1,060.0 33,983.9 6.7% Interior 9,612.7 2,484.0 12,096.7 2.4% Foreign Affairs 6,919.8 1,115.0 8,034.8 1.6% Agriculture 4,429.0 13,517.0 17,945.9 3.5% Commerce 4,170.0 5,357.0 9,527.0 1.9% Finance 72,348.1 5,799.0 78,147.1 15.4% Education 34,649.4 19,862.4 54,511.8 10.8% Health 14,006.9 25,684.0 39,690.9 7.8% Infrastructure 25,405.1 52,125.0 77,530.1 15.3% Lands 2,916.9 25,903.0 28,819.9 5.7% Local Government 16,969.4 19,048.1 36,017.5 7.1% DISTRICTS 64,639.7 0.0 64,639.7 12.8% Districts (of which Agri) 2,005.4 0.0 2,005.4 0.4% Agriculture (Total): 6,434.4 13,517.0 19,951.4 3.9% TOTAL: 324,875.6 181,869.5 506,745.1 100% Source: Agriculture PER Report, 2007, page 34. Original Source: Analysis of Budget Law (MINECOFIN, 2006 (a)). The picture of lower budget allocation priority to MINAGRI was in contrast to the government's stated priority of agriculture as a main engine for growth. It also showed a sharp contrast to allocations of other social sectors, in particular health and education. Health and education in particular experienced strong growth in budget allocations in real terms12 from 2004 to 2007, demonstrating the prioritization stated in the PRSPs (see table 5.10). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 115 Table 5.10. Budget Allocations: Agriculture and Social Sectors Compared 2004­0713 In Real Terms (in constant 2001 RwF in millions) Growth Ministry 2004 2005 2006 2007 2004­07 Health 11,973 19,892 18,761 23,348 95% Education 22,343 23,957 27,523 32,066 44% Agriculture 10,737 9,448 8,449 11,736 9% Source: Agriculture PER Report, 2007, page 35. Original Source: Analysis of Budget Laws (MINECOFIN 2006(a) and Republic of Rwanda 200305; Ministry of Finance and Economic Planning, 2007; and own projections. Health and education increased in constant 2001 terms by 95 percent and 44 percent, respectively, while agriculture increased by only 9 percent. Looking across all sectors in the 2007 budget (figure 5.1), agriculture emerged as one of the smaller sectors based on total budget allocation (recurrent and development budgets combined) and the smallest priority sector. Figure 5.1. Total 2007 Budget Allocations by Ministry 90,000 80,000 70,000 60,000 50,000 In million RwF 40,000 30,000 20,000 10,000 0 Ministry Source: MINECOFIN 2007. Budget allocations by ministries for 2007 recurrent expenses only, as illustrated in figure 5.2, show that MINECOFIN, MINEDUC, MINADEF, MININFRA, MINALOC, and MINISANTE represented the largest sectors, with a combined RwF 196.3 billion in 2007 or a 76 percent of the recurrent budget allocation. 116 World Bank Working Paper Figure 5.2. Total 2007 Recurrent Budget Allocations by Ministry 80,000 70,000 60,000 50,000 In million RwF 40,000 30,000 20,000 10,000 0 Ministry Source: MINECOFIN 2007. Recurrent 2007 budget allocation to ministries shows that education and health received a combined RwF 48.7 billion or 19 percent of the total recurrent budget--13 percent and 5 percent, respectively--compared with 2 percent for agriculture. Table 5.11 compares MINAGRI's original budget with actual expenditures for 2003­05 for both recurrent and development expenses. Table 5.11. MINAGRI Budget and Actual Expenditure 2003­05 In Real Terms (in RwF millions) 2003 2004 2005 2003 2004 2005 Budget Actual Budget Actual Budget Actual % % % Recurrent 3,707 2,300 4,660 4,338 5,977 6,008 62% 93% 101% Development 11,616 8,173 13,710 9,976 10,105 14,468 70% 73% 143% TOTAL 15,323 10,474 18,369 14,314 16,083 20,476 68% 78% 127% Source: Agriculture PER Report, 2007, page 36. Original Source: Analysis of Budget Laws, Republic of Rwanda 2002­05. Actual expenditures for MINAGRI differed significantly from budgeted expenditures, with overall execution rates of 68 percent in 2003, 78 percent in 2004, and 127 percent in 2005. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 117 The high execution rate in 2005, which exceeded the budget by RwF 4.4billion is explained by the Rural Sector Support Project's (RSSP) overexecution and is attributed to improved efficiency in procuring goods and services.14 MINAGRI also exceeded its overall original budget in 2006, as shown in table 5.12. Table 5.12. 2006 MINAGRI Execution Rates In Real Terms (in RwF millions) 2006 Budget 2006 Actual Execution Rate Recurrent 5,172 3,846 74% Development 9,718 16,519 170% TOTAL 14,890 20,365 137% Source: MINECOFIN 2006(a). Recurrent expenditures in 2006 were only 74 percent executed because of lack of capacity, according to MINAGRI. Development expenditures as reported by CEPEX for 2006 show an overexecution of 170 percent. Agriculture expenditures contributed an average of 3 percent of the agriculture GDP between 2000 and 2005, but only an estimated 2.3 percent in 2006 (see table 5.13). Table 5.13. Agriculture Budget as Proportion of Agriculture GDP 2000­06 In Real Terms (in RwF millions) MINAGRI Agriculture Expend/ AgriContrib YEAR Expend. GDP AgrGDP % Nat'l GDP* 2000 15,426 467,949 3.3% 37% 2001 23,180 501,600 4.6% 37% 2002 26,006 529,091 4.9% 36% 2003 16,353 574,784 2.8% 39% 2004 19,448 610,748 3.2% 39% 2005 17,153 654,754 2.1% 39% 2006** 14,890 651,176 2.3% 37% Average: 3% Source: Agriculture PER Report, 2007, page 37. Original Source: MINECOFIN, 2006(a); Republic of Rwanda 19992004; NISR 2007; and MINECOFIN 2007. *Includes funds allocated to agriculture under districts budget. **Provisional. Although the 2006 agriculture budget as portion of agriculture GDP contributions is only an estimate, the 2.3 percent is significantly lower than previous years' (ranging between 2.8 percent and 4.9 percent) and lower than the period's average. Rwanda's agriculture budget allocations as a percentage of agriculture GDP is smaller than other developing countries. Although Rwanda's rate of 3 percent is in line with its neighbor Uganda, a 2002 study (Fan and Rao, 2003; Fan and Saurkar, 2006) of 118 World Bank Working Paper 44 developing countries showed that for 17 of SubSaharan African countries, the figure was 6.7 percent. Even in 2000­02, when Rwanda's agriculture allocation was a higher percentage of agriculture GDP than it was in 2006, Rwanda's rate was lower than the 6.7 percent reported in the 2002 study. Overall, this study showed that the average agriculture budget allocation as a proportion of agriculture GDP for all 44 developing countries was 10.3 percent.15 Sector funding to districts occurred for the first time in 2007 when MINECOFIN directly transferred a total of RwF 2.0 billion. Table 5.14 describes some of the characteristics of the district allocations of agriculture funds. Table 5.14. Overall Picture of District Allocations for Agriculture in 2007 · There was an unequal allocation of funds between districts, ranging from RwF 44 per person (Nyamasheke district) to RwF 647 per person (in Nyamagabe district). · Funding for agricultural activities as a proportion of the total budgets of individual districts varied significantly, ranging from 0.5 percent (Karongi district) to 10 percent (Kirehe district), with a national average of 3.1 percent. · Justifications for these differences in district allocations are not clear. Source: Agriculture PER Report, 2007. Further evidence of lack of equity in allocations of funds to districts is illustrated in table 5.15. Table 5.15. Proportion of Agriculture Recurrent Budget and Prevalence of Poverty by Province Share of Agri By Province Poverty Head Count Share of the Poor Recurr. Budget Southern 67% 30% 29% Western 62% 26% 16% Northern 63% 20% 28% Eastern 50% 20% 21% Kigali 20% 3% 7% National 60% 100% 100% Source: Agriculture PER Report, 2007, page 39. Original Source: NIRS, 2006, and analysis of data contained in Annex 6.1. The Western Province received the lowest recurrent budget allocation, 16 percent, despite having the second highest share of poor among all the provinces and the third highest prevalence of poverty head count, at 62 percent, higher than the national average of 60 percent.16 Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 119 Table 5.16. Earmarked Budget Lines in Districts 2007 · Livestock vaccines (allocated through RADA budget) · One Cow Per Family · Hillside Irrigation and Erosion Control · Total allocation for all districts for three above programs was RwF 2.3 billion. · The methodology used for allocating these funds among districts is unclear. Source: Agriculture PER Report, 2007. As a proportion of total funds allocated to recurrent budget activities in the agriculture sector, the 30 districts accounted for about 31 percent in 2007, compared with only 13 percent in the previous year and only 11 percent in 2005. Recurrent funding of districts' forestry activities in 2007 accounted for almost 24 percent of MINITERRE's total recurrent budget and for about 3 percent of its total budget. In general, MINAGRI's decentralization was further along than, for example, decentralization of MINITERRE's activities. Overall donor assistance accounted for 55 percent of MINAGRI's 2007 overall budget, compared with 59 percent in 2006 and the average of 80 percent in the 2000­03 period. Donor assistance for MINAGRI's development budget averaged about 87 percent in the 2000­07 period, although the government increased its internal contributions in the same period, especially in 2005 and 2007, as illustrated in table 5.17. Table 5.17. Source of Funding of MINAGRI's Development Budget 2000­07 Government of Development TOTAL Development Dev Partners/ YEAR Rwanda Partners Budget Total Dev (%) 2000 819 11,740 12,559 93% 2001 701 18,720 19,422 96% 2002 1,466 19,399 20,865 93% 2003 743 8,147 8,889 92% 2004 1,025 5,466 6,492 84% 2005 2,944 7,161 10,105 71% 2006 934 8,783 9,718 90% 2007 3,559 9,958 13,517 74% AVERAGE 87% Source: Agriculture PER Report, 2007, page 40. Original Source: Raw Data Contained in Budget Laws (MINECOFNIN 2006 (a) and Republic of Rwanda 19992005). Development partner contributions were a much smaller portion of the total development budget in 2007, as measured in real terms. Looking at real terms development partners contributions, the RwF 9.9 billion budgeted for 2007 was slightly more than half of the RwF 19.4 billion committed in the 2002 budget. 120 World Bank Working Paper A limited number of aid agencies provide the bulk of MINAGRI's development resources. The agencies listed in table 5.18 provided between 50 percent and 75 percent of all development budget resources for the sector in 2002­07. Table 5.18. Main Development Agencies in the Agriculture Sector · African Development Bank · European Commission · World Bank · Luxembourg Source: Agriculture PER Report, 2007. Many aid agencies indirectly contributed to the recurrent expenditures by general budget support. These agencies are shown in table 5.19. Table 5.19. General Budget Support Partners · African Development Bank · DFID · European Commission · Sida · World Bank Source: Agriculture PER Report, 2007. Internally sourced financing for MINAGRI comes from CP funds and FI funds. Table 5.20 shows the trends for MINAGRI's internally sourced financing for its development budget between 2003 and 2007. Table 5.20. Government Funding of MINAGRI's Development Budget 2003­07 Internal Counterpart TOTAL GoR Proportion of FI YEAR Financing (FI) Financing (CP) Contribution to Total GoR (%) 2003 269 474 743 36% 2004 484 541 1,025 47% 2005 2,574 371 2,944 87% 2006* 0 934 934 0% 2007 1,919 1,640 3,559 54% Source: Agriculture PER Report, 2007, page 41. Original Source: Raw data contained in budget laws (MINECOFIN, 2006(a)) and Republic of Rwanda 1999­2005). *No FI funds budgeted for MINAGRI 2006 in budget law (year of full decentralization). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 121 Allocations between FI and CP for MINAGRI fluctuated significantly over the five years in table 5.20. For 2006, no FI was reported as full decentralization came into effect and changes to reporting occurred. In 2007, the distribution between CP and FI appeared to have balanced out, with 54 percent of total government funding classified as FI. Data on offbudget support from development partners are incomplete. However, identified offbudget funding for development in agriculture represented 68 percent of total onbudget development funds in 2006 and 78 percent in 2007, as shown in table 5.21. Table 5.21. Comparing On- and Off-Budget Support from Donors in 2006 and 2007 In real terms (in million RwF On-Budget Off-Budget TOTAL OVERALL Off-Budget Fund Fund On & Off On Budget % of Total YEAR Budget Estimated Budget MINAGRI On-Budget 2006 8,783 6,583 15,366 9,718 68% 2007 9,958 10,527 20,485 13,517 78% Source: Agriculture PER team consultancy interviews with development partner representatives (table 6.10 Agriculture PER, 2007); and MINECOFIN 2006 and 2007. In 2007, donor offbudget contributions even exceeded donor onbudget contributions by about 6 percent. Table 5.22 presents some of the offbudget projects in the agriculture sector. The list of projects is by no means allinclusive. Offbudget sources represented a significant portion of funds in the sector, RwF 5.9 billion in 2006 and RwF 10.5 billion in 2007. Several other government ministries contributed to the agriculture sector in addition to offbudget donors. Complete data for other ministries' and agencies' contributions to the sector do not exist.17 Overall agriculture spending: Rwanda's overall budget for agriculture increased by about 30 percent during 2000­07, although in real terms, the increase was less significant at 3.6 percent per year.18 Agriculture's recurrent budget almost doubled in real terms between 2000 and 2007, an annual increase of about 7.6 percent. The total national recurrent budget also increased in the same period, but agriculture's share of the recurrent budget in the same period barely changed, from 1.5 percent to 1.4 percent.19 The average annual increase in the development budget in real terms was about 2 percent per year over the 2000­07 period, and as a proportion of the total national development budget, agriculture's share decreased significantly, down to 7.4 percent in 2007 from 10.6 percent in 2000.20 122 World Bank Working Paper Table 5.22. Principal Agriculture-Related Off-Budget Interventions by Development Partners 2006­07 Approximate Budget In Nominal Terms (in RwF millions*) *Exchange Rates DONOR INTERVENTION FUNDED 2006 2007 (period average): CANADA Fonds Women NGOs, 100 100 Euro 1.00=US$1.11 Canadiennes co-ops, crops, US$1.00=C$ 1.15 Genre et Dev. livestock, fish, US$1.00 = RwF 550.00 processing PAGOR Demand-driven 290 rural interventions NETHERLANDS Support to Soil 480 1221 HIMO (help conservations, Age) marchlands, fisheries Support to Terracing & 855 HIMO (German marsh-land Agro-Action) development ROPARWA Farmer orgs. 670 646 strength Loan Guarantee Agri-business 382 638 credit CAPMER Agribusiness 160 233 ISAR Agricultural 252 426 research FIFA Farmers' 644 organizations USAID** Essential Oils Geranium&petunia 165 165 Agri-business Loan Guarantee 132 132 credit SPREAD Agri-bus support 660 660 coffee JAPAN Gov Sustain 771 660 Plan formulation & Agri&Rural pilot water projects Development Support to Poor ^^Assumes 25% of Fertilizers 151^^ H.H. projected imports to be consumed in 2007 Various Seed, food FAO 700^ 693 ^Actual figure not Interventions security available, this is an Food-4- estimate WFP Terracing&erosion 749 3,657 Work/Train. TOTAL: 5,865 10,527 Source: Agriculture PER Report, 2007. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 123 Table 5.23. Agriculture Recurrent and Development Budget 2000­07 In 2001 prices (in RwF millions) YEAR Recurrent Develop. TOTAL 2000 1,557 6,822 8,380 2001 2,075 10,725 12,801 2002 2,226 11,527 13,753 2003 2,048 6,416 8,464 2004 2,573 7,568 10,141 2005 3,292 5,566 8,858 2006 2,935 5,514 8,449 2007 2,605 7,951 10,556 Source: Agriculture PER Report, 2007, page 46. Original Source: R aw data in budget laws (MINECOFIN 2006a; Republic of Rwanda 19992005); NIRS 2007 and MINECOFIN 2007. An economic classification overview within MINAGRI is provided below: Recurrent Economic Classification · The recurrent budget allocation increased slightly in absolute terms over the eightyear period, representing about 25 percent of the total MINAGRI budget in 2007. · The proportion of the recurrent as percentage of the total budget, however, decreased in the last year of the period, down from 37 percent in 2005 and 34 percent in 2006.21 · In contrast with many other developing countries where recurrent expenditures often are dominated by wages and too little expenditure remains for operation and maintenance activities, Rwanda kept a rather balanced recurrent allocation between wages and nonwages--although the proportion fluctuated from about 58 percent allocated to wages in 2000 to 46 percent in 2007.22 Development Economic Classification · MINAGRI's development budget fluctuated significantly between 2003 and 2007 (see table 5.24), which in part was attributed to the "projectized nature" of the development budget and the high involvement of external donors.23 · The government increased its contribution as a portion of the total development budget from about RwF 400 million (about 4 percent in 2000 and 7 in 2001) to about RwF 3.6 billion (26 percent of total) in 2007. · The development budget funds some recurrent expenditure, as shown in table 5.24. 124 World Bank Working Paper Table 5.24. Development Budget Allocations to Recurrent and Capital Cost YEAR* Recurrent Cost Capital Cost TOTAL Recurr/Tot % 2003 4,505 4,385 8,889 51% 2004 3,081 3,410 6,492 47% 2005 6,652 3,454 10,105 66% 2006 6,343 3,375 9,718 65% 2007 7,178 6,339 13,517 53% Average: 56% Source: Agriculture PER Report, 2007, page 48. Original Source: Analysis of Budget Laws (MINECOFIN 2006a; Republic of Rwanda 200205). *Published budget only started being broken down to this level in 2003. An average of 56 percent of the development budget was allocated for funding of recurrent expenditure, including salaries. This raises the question of sustainability of longerterm projects and MINAGRI's ability to absorb these recurrent costs, as the ministry is expected to take over these expenditures as projects come to an end. On average, only 3 percent of the development budget was devoted to training to help build capacity; in no year has the training budget exceeded 5 percent. That is a low figure considering the importance of capacity building to ensure sustainability and quality of projects and institutions. The total cost of projects allocated to credit and credit guarantees represents a rather high portion of the total cost of projects, averaging 10 percent between 2003 and 2007. There is no regular monitoring of the use and impact of these credit lines. A functional classification overview shows that MINAGRI has transformed from a functional structure of departments to one of programs and subprograms. Table 5.25 shows MINAGRI's organization and recurrent budget allocations to each department prior to 2006. Table 5.25. Recurrent Budget Allocations to MINAGRI Departments 2000­05 Department: 2000 2001 2002 2003 2004 2005 Minister's Office 74% 5% 6% 4% 2% 3% Agricultural Production 11% 36% 48% 42% 61% 66% Livestock 7% 24% 21% 229% 23% 17% Forestry 2% 4% 7% 10% 1% 1% Soil & Water Conservation 3% 7% 3% 4% 3% 4% Extension 2% 2% 7% 7% 6% 4% Planning & Statistics 1% 2% 4% 2% 2% 1% HRD/Public Relations 1% 19% 4% 2% 3% 4% Biotechnology 1% Source: Agriculture PER Report, 2007, page 49. Original Source: Analysis of budget laws (Republic of Rwanda 19992005). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 125 Crops and livestock activities received the majority of funds. Almost no data are available on the breakdown of expenditures by function. However, given that almost all funding for agricultural research (which represents approximately onequarter of all recurrent expenditure of the Ministry) is listed separately in the budget laws, it is possible to give an indication of the agricultural research intensity. This is a measure that shows public expenditure on research as a proportion of agricultural GDP. Agricultural research as a proportion of agriculture GDP was in 2006 to be about 0.2 percent, compared with 0.5 percent in Uganda and 0.72 percent in SubSaharan Africa (Akroyd and Smith 2006), indicating that agricultural research in Rwanda is well below the regional average. Agricultural extension is somewhat harder to analyze as some funds are allocated to RADA and RARDA. Nonetheless, the total budget to these agencies in 2007 was RwF 2.1 billion or about 0.37 percent of total agriculture GDP. Programs and subprograms (listed in table 5.2) have defined MINAGRI's annual budget since 2006. Table 5.26 shows the budget allocations to these programs in 2006 and 2007.24 Table 5.26. Budget Allocations to MINAGRI by Program 2006 and 2007 Program Program 1 Program 2 Program 3 Program 4 TOTAL Recurrent 2,215 1,471 470 1,016 5,172 Development 3,927 2,023 3,768 0 9,718 2006 TOTAL 6,142 3,494 4,238 1,016 14,890 Proportion % 41% 23% 28% 7% 100% Recurrent 1,650 1,597 462 720 4,429 Development 5,792 512 5,887 1,326 13,517 2007 TOTAL 7,442 2,109 6,349 2,046 17,946 Proportion % 41% 12% 35% 11% 100% Source: Agriculture PER Report, 2007, page 50. Original Source: Analysis of budget laws (MINECOFIN 2006a and Republic of Rwanda 2005). Program 1 (Intensification and Development of Sustainable Production Systems), with its six subprograms, was the largest program, with 41 percent of total MINAGRI recurrent and development budgeted spending for both 2006 and 2007. Program 4 (Institutional Development) was the smallest with only 11% of the total funding. Programs 2 (Support to Professionalization of Producers) and 3 (Promotion of Commodity Chains and Development of Agribusiness) showed a slight reallocation between 2006 and 2007 as Program 2 declined from 23 percent of MINAGRI's budget in 2006 to 12 percent in 2007, while Program 3 increased its share from 28 percent to 35 percent in the period. However, Program 3 has a significantly higher proportion of development spending, in line with the much larger Program 1. 126 World Bank Working Paper Overall, the total spending increased in real terms from RwF 14.9 billion to RwF 17.9 billion, an increase of 20 percent. Furthermore, the PER report has found these program allocations to be in line with PSTA sector priorities.25 Agriculture Sector Results Before 2004, the agriculture sector did not have a clear strategy or vision. However, with guidance through sector dialogue, a Strategic Plan for Agricultural Transformation (PSTA) was adopted in 2005 with four priority program areas: (a) intensification and development of sustainable production systems, (b) support to professionalization of farmers, (c) promotion of commodity chains and development of agribusiness, and (d) institutional development. A sector MTEF was developed in 2005 and 2006 that was aligned with the PSTA; the PRSP matrix was also updated to reflect PSTA priorities. In 2005 a land policy and land law, guaranteeing tenurial security and potentially permitting the development of a land market, were adopted and an Agriculture Guarantee Fund was established. Table 5.27 lists the key sector reforms undertaken between 2004 and 2007 to enhance private sector growth and foster economic growth, of which the agriculture sector plays a key role. Table 5.27. Agriculture Sector Reform Measures 2004­07 Theme/Sector Reform Measures Undertaken Year Two more tea factories privatized 2007 National microfinance policy adopted 2006 New microfinance law drafted and submitted to Parliament 2006 Rwanda Agriculture Development Authority (RADA) established 2006 Rwanda Animal Resources Development Authority (RARDA) established 2006 Rwandatel, RwandaEx, and two tea factories privatized 2005 Export promotion strategy developed (incl. Rwanda Investment & Export Promotion 2005 Private Sector Agency) Development & Land law published in the Official Gazette 2005 Economic Growth Agriculture sector MTEF aligned with Sector Strategic Plan (PSTA) 2005 Regulatory framework for microfinance adopted 2005 Investment code revised to align with revised tax code 2004 Commercial chambers established 2004 National Agricultural Policy and PSTA developed 2004 Premium pricing of tea leaves for producers introduced 2004 Agriculture Guarantee Fund established 2004 Source: Agriculture PER Report, 2007. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 127 In 2006, key implementing agencies--RADA and RARDA--were established, and the agriculture research institute was restructured to improve its representation in different agricultural zones and to develop a technology transfer unit. In 2006, earmarked transfers to local government began, in line with the national decentralization and territorial reform process and also in line with PSTA national priorities. Funds were transferred to districts for terracing and animal disease control operations, with technical oversight from RADA and RARDA. This expanded in the 2007 budget with earmarked grants for animal disease control, extension, erosion control, rural infrastructure, water harvesting and hillside irrigation, and food security. Furthermore, in 2007 almost 5,000 hectares had been terraced, with the onecow program enjoying great success. More marshland was placed in protective management with plans to increase protected areas. There was distribution of seeds and an increase in import of fertilizer. The government's Poverty Reduction Strategy Paper (PRSP) of 2002 emphasized intensification and commercialization through increased use of traditional and modem inputs, particularly fertilizer, and the importance of public actions in the area of research and extension and increased public spending to provide services to the sector. However, the PSTA, which was to be the framework for policies and actions, was not finalized until 2005. The predominance of very smallscale producers across the country has been a key challenge to ensuring access to inputs and agricultural services, particularly in the context of weak capacity of the sector ministry and poor coordination among the main actors. Limited progress was made in the use of modern inputs in the sector. Fertilizer usage: In 2005, the total volume of fertilizer imports increased 90.6 percent over 2004 as part of the drive to boost agricultural productivity. However, farmers continued to eschew the use of fertilizer, and this low usage was attributed to a lack of knowledge about fertilizers, high prices, inadequate supplies, and lack of credit. With fertilizer usage at 4kg/ha, Rwanda was far behind other developing countries with average usage of 300­400kg/ha (MINAGRI/RADA 2006 b). Consequently, the government liberalized the market for fertilizer imports to encourage private sector participation. A 2006 Strategy for Increasing Fertilizer Use was prepared to set specific targets and clarify objectives for improved agricultural inputs.26 This led to an increase in the number of fertilizer importers, but because of low demand by farmers, the quantity of fertilizer imports peaked at 7,349 tons in 2005 and remained far below the target of 63,000 tons. Furthermore, the underlying microlevel constraints linking demand and supply persisted. As a result, the big commercial farmers benefited more than smallholders from a liberalized fertilizer market. Nevertheless, there are signs that commoditybased cooperatives are now playing increasing roles to address the microlevel constraints, by improving access of smallholders to fertilizers with bulk purchases at wholesale prices, credit guarantees for farmers, and associated extension services. Also, the government's flagship Vision 2020 strategy addresses several of the issues of agricultural production, including the issue of improved access to credit. Seed usage: Seed production was found to be significantly below demand during the last PER, a mere 0.7 percent of total seed demand in 2006 (see table 5.28). 128 World Bank Working Paper Table 5.28. Seed Production vs. Seed Needs 2005­06 2005 2006 Seed Seed Seed Seed Average Annual Production Production Production Production vs Seed Type Seed Needs (in Tons) vs Needs (%) (in Tons) Needs (%) Maize 3,900 271 6.9% 188 4.8% Rice 400 119 29.8% 52 13.1% Wheat 1,500 107 7.1% 77 5.1% Beans 18,900 224 1.2% 154 0.8% Soya 2,160 74 3.4% 69 3.2% Irish Potato 230,000 2,464 1.1% 1,220 0.5% TOTAL: 256,860 3,259 1.3% 1,760 0.7% Source: Agriculture PER Report, 2007, page 18. Original Source: From information provided by the Seed Unit, RADA, 2007 The seed sector received continued support from the Belgian Cooperation, although now with significant restraints because of limited capacity and weak private sector involvement. The Seed Support Project 2004­2009 is working to strengthen the capacity of cooperatives and the private sector. Food Security: Food insecurity has improved since 2000, but seasonal and regional food shortages from poor rainfall, lack of seeds and other inputs, and disease outbreaks continue to occur. Furthermore, a clear deficit between recommended minimum nutritional standards and actual nutrition consumed exists. Domestic production contributes about 90 percent of the nutritional portion measured by food availability. The gap is met by imports because the government strictly regulates and attempts to reduce food aid. Tables 5.29­5.31 provide a snapshot of the latest food security data. Table 5.29. Nutritional Needs Satisfaction between 2001 and 2005 · Energy 83% · Proteins 73% · Lipids 30% Source: Agriculture PER Report, 2007. Table 5.30. Food Security Profile for Rural Population, 2006 · 28% food insecure · 24% vulnerable · 26% moderately food secure · 22% food secure Source: Agriculture PER Report, 2007. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 129 Table 5.31. Highest Food Insecurity Zones in Rwanda 1. Bugesera 2. Crete Congo Nile 3. Kivu Lakeshore 4. Eastern Curve 5. Southern Plateau Source: Agriculture PER Report, 2007. Private sector development: Agriculture private sector development continues to be weak, and private investment faces major constraints such as high risk and lack of access to external markets. This weakness persists despite the government's efforts to try to strengthen the private sector, including initiatives such as the Agriculture Guarantee Fund to subsidize lines of credit in the sector. Private investments were identified as an important future contributor of growth in the sector. The 2002 National Investment Strategy announced the government's intent to increase public investment from RwF 8.3 billion in 2002 to RwF 15 billion in 2006, and then to gradually decrease it to about RwF 10 billion in 2010--with the intent for the private sector to play an increasingly greater role. However, private sector investments in agriculture remained significantly lower than in other sectors, as illustrated Table 5.32. Table 5.32. Development of Total Bank Loans in the Agriculture Sector 2002­05 Sector Dec 2002 Dec 2003 Dec 2004 August 2005 Crop, Livestock, Fish, Forestry 2.1 3.0 5.5 5.1 (%) 1.7% 2.4% 4.4% 4.0% Commerce, Rest., Hotels 50.0 55.2 44.3 49.7 (%) 40.4% 42.6% 35.6% 39.2% Construction 25.1 26.3 30.6 30.1 (%) 20.3% 20.4% 24.7% 23.3% Industry/Manufacturing 16.6 18.3 18.2 14.5 (%) 13.5% 14.2% 14.7% 11.4% Other Sectors 29.7 26.5 25.4 21.5 (%) 24.1% 20.5% 20.5% 21.7% TOTAL: 123.5 129.4 124.0 120.8 (%) 100% 100% 100% 100% Source: Agriculture PER Report, 2007, page 27. Original Source: Compiled from NBR 2005 Annual Report, June 2006. Furthermore, while total loans in agriculture sector increased from RwF 2.1 billion in 2002 to RwF 5.5 billion in 2004 and to RwF 5.1 billion in 2005, total bank loans in the agricultural sector (excluding agroprocessing) represented only a mere 4 percent of all bank loans, compared with loans to the commerce and construction sectors, which accounted for more than 62 percent of all bank loans in 2002­05.27 130 World Bank Working Paper Agriculturebased products accounted for 71 percent of Rwanda's total export revenues in 2006, according to the most recent PER. For example, export earnings from coffee and tea are shown in table 5.33. Table 5.33. Trends in Coffee and Tea Production and in Exports 2000­05 2000-05 2000 2001 2002 2003 2004 2005 Ave. Rev COFFEE Total Production (tons) 16,098 18,267 19,426 14,175 29,000 18,000 Coffee Exports US$ (mill) 22.4 $ 19.4$ 19.2$ 15.0 $ 32.0$ 38.0$ $ 24.33 % of Export Contribution 32.5% 26.6% 29.1% 23.8% 36.0% n.a. TEA (black tea) Total Production (tons) 14,481 17,817 14,948 15,483 13,998 16,899 Tea Export US$ (mill) 18.0 $ 20.5$ 18.1$ 18.8 $ 23.2$ 26.5$ $ 20.85 % of Export Contribution 36.0% n.a. 34.0% 35.0% n.a. n.a. Source: Agriculture PER Report, 2007, page 16. Export revenues from coffee rose from US$22.4 million to US$38.0 million between 2000 and 2005, representing on average 30 percent of total export earnings. Coffee market success can be attributed largely to the PEARL Project, supported by the U.S. Agency for International Development (USAID) in the Southern Province. The project was later extended by OCIR Café/STABEX and the Cash Crop Project (PDCRE) other projects and to private investors. Export revenue from tea generated the second greatest amount of export earnings, an average of US$20.9 million between 2000 and 2005, with export revenue rising from US$18.0 million in 2000 to US$26.6 million in 2005. Furthermore, to encourage export diversification, the government created the Rwanda Horticulture Development Authority and installed cold storage facilities at Kigali Airport with support from the Dutch government (US$637,500). Agriculture Sector Challenges Expenditure for agricultural research is low, less than onethird of the SubSaharan African average. Additionally, agriculture has not been prioritized within the national budget, with annual allocations being stagnant (at best) in real terms between 2000 and 2007, according to PER findings. Furthermore, development partners noted in September 2007 that an effective monitoring and evaluation system was still lacking one year after it was recommended. Projectlevel information is important because specific impacts must be known to justify the scaling up of funding. Moreover, a well functioning agriculture system is fundamental to a propoor development strategy. It is important to note what is causing low yields so that the problem can be rectified. Also, as a major portion of the development budget is financed by development partners, actual expenditures are highly unpredictable.28 Although agriculture is considered a priority sector and an engine for economic growth, the trends in the sector's recurrent and development budget allocations indicate there is less emphasis on the sector. PER Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 131 recommendations included greater focus on efficiency within the sector and a proposal to conduct expenditure tracking surveys to evaluate efficiency of sector allocation and use. Furthermore, MINAGRI should consider undertaking more systematic measurements of expenditure efficiencies by setting up a comprehensive management information system that links districts' performance to central targets in the sector. Education Education Sector Overview Education in Rwanda has undergone a great number of changes in the last few years, a process guided by both policy and strategic direction. The most significant change was the alteration from six years of primary education and six years of secondary education to nine years of basic education and three years of upper secondary, providing full education for the 7­15 age group. Furthermore, as part of the decentralization of government across all sectors, education administration was decentralized to newly created district administrations, to schools and community institutions, and to semiautonomous government agencies. In addition, the primary education capitation grant today provides important funding for teacher bonuses and hiring supplementary teachers. According to the second household living condition survey (EICV II), literacy in Rwanda is rather high and has improved over the past 10 years. Seventyseven percent of 15­24 yearolds are literate, and 65 percent of the whole population over age 15 is literate (see table 5.34). Table 5.34. Literacy Rates in Rwanda Male Female Total Literacy--People Aged 15+, by Gender & Location (%) Kigali Urban 90% 84% 86% Other Urban 76% 67% 79% Rural 69% 57% 63% Rwanda 72% 60% 65% Literacy--People Aged 15-24, by Gender & Location (%) Kigali Urban 89% 88% 88% Other Urban 78% 84% 82% Rural 76% 75% 75% Rwanda 77% 77% 77% Literacy--People Aged 15+ (%) Illiterate 26% 37% 32% Can read but not write 3% 3% 3% Can read and write 71% 60% 65% Total 100% 100% 100% Kigali Urban Other Urban Rural Illiterate 12% 26% 35% Can read but not write 2% 3% 3% Can read and write 86% 71% 62% Total 100% 100% 100% Source: Education PER Report, July 2007, page 15. 132 World Bank Working Paper A strong indicator of improvement in recent years is that the 15­24 age group shows almost no difference in literacy rates between male and female students, while in older age groups male literacy rates are clearly higher than female rates. However, 23 percent of the 15­24 age bracket remains illiterate. Furthermore, literacy rates are considerably higher in Kigali than in the other regions of the country. For example, 86 percent of people living in Kigali can read, while only 62 percent of the rural population can read or write. Thirtyfive percent of the rural population is illiterate. Table 5.34 summarizes the most recent data on literacy rates. As of 2007, development partners were funding about 50 percent of Rwanda's overall government budget, and as a result aid effectiveness continues to be vital to ensure resources are effectively used and contribute to the strategic goals of all sectors. Yet, aid predictability remains uncertain, and the education sector financing gap is projected to grow further from 2009 and on. As part of its Aid Policy in 2006, the government has therefore expressed its preference for GBS from its partners. As of 2007, donors providing the largest amounts to education in Rwanda are DFID, AfDB, the Netherlands, and the World Food Programme (WFP). Furthermore, Rwanda is part of a FastTrack Initiative (FTI) to ensure funding of US$26 million in 2007/2008 and US$44 million in 2008/2009. Education sector strategies are guided mainly by the policies listed in table 5.35. Table 5.35. Education Sector Policies Education Sector Strategic Plan 2006­10 (ESSP) Based on the Education Sector Policy Paper (ESP) 2002 Long-Term Strategy and Financial Framework 2006­15 For further details, see annex 11.1 in the original Education (LTSFF) PER report, July 2007. Subsector documentation Provides additional guidance to the main policy papers Source: Education PER Report, July 2007. Over the past few years, MINEDUC has used a computer simulation model based on standard techniques for planning its budget allocations and meeting its policy objectives. Although a valuable tool, the model has become complex and difficult to update. As MINEDUC continues to face difficulty estimating budget needs, as shown later in this report, the model needs to be improved to fully capitalize on resources and capacity, and to successfully execute its policy objectives. Rwanda's ministry of education, MINEDUC, conducted a public expenditure review (PER) of the education sector in July 2007 (Education PER 2007). Some of the key conclusions of this PER report, covering 2001­07, are summarized below. Education public spending rose from 3.4 percent of GDP at the beginning of the period to about 5 percent in 2007, representing about 26 percent of government recurrent spending. The increase in spending on education as a percentage of GDP was partially mirrored in budgets per student since 2001 when considered in real rather than nominal terms. Nearly half of recurrent education spending was allocated to districts for salaries, capitation grants, school feeding, subsidies for secondary schools, and technical education. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 133 Just over half of the 2007 capitation grant was intended to support teachers, through bonuses for existing teachers (an amount equivalent to an average salary increase of 30 percent) and as funds for schools to hire supplementary teachers. However, more guidance to schools on the use of these resources is needed. Variations in education expenditure show the following tends: Household spending on education was between 3 percent of their total spending (poorest households) and 5 percent (richest households). Household expenditure varied considerably across Rwanda, from far less than RwF 4,000 in the lowest quintile in the Northern Province to more than RwF 250,000 in the highest quintile in Kigali. Rwanda continued to spend relatively much more on secondary and especially tertiary education than on primary education. Evaluation of school funding and its management practices revealed the following: The Capitation Grant Minister's Directive with guidelines on the use of capitation grants was received in 2007 by more than 9 out of 10 primary schools and about 8 in 10 secondary schools. Directive requirements include having in place school bank accounts, parent teacher associations (PTAs) and school management committees (SMCs) and prepare action plans and budgets, and financial reports. In primary schools the principal actors in budget preparation are the school director, the management committee, and the teachers. The average performance bonus that schools intended to pay in 2007 was RwF 12,747 per month per teacher, which was slightly more than the recommended RwF 12,287 per month. Performance contracts were signed by 4 out of 5 primary teachers and 7 out of 10 secondary principals (by the time of the survey). The National Examination Council (CNER) organizes three public examinations, and its current spending cannot be analyzed to show how much it costs to run each of these three exams. This information would enable officials to use resources more efficiently. Following its national education strategy, Rwanda is well on the way to meeting the MDG of universal primary education as can be seen by the rapid expansion within the education sector: Primary enrollment rates are high, with a large proportion of the relevant age group in school. Completion rates remain disappointing: only 40 percent of those who enter Grade 1 will complete Grade 6. Tronc commun29 enrollment has been growing at 12 percent a year since 2002. Tronc commun publicly supported classes have more than doubled since 2002 as new, local tronc commun schools opened. Upper secondary education has grown quickly from its very low base, mostly in private schools. Tertiary education enrollment quadrupled between 2000 and 2006, as new public and private learning institutions opened. 134 World Bank Working Paper Gender equity in enrollment is good, particularly at the primary level, although girls' learning achievement appears lower than boys' at all levels. Enrollment equity is less good at higher levels. However, many challenges remain to meeting all strategic goals, including the following: Enrollment in primary school is unbalanced, meaning large numbers enter an overcrowded Grade 1, but grade repetition and dropout rates are high. Implementation changes in primary education are urgently needed. Transition rates from primary to tronc commun are growing, while those from tronc commun to secondary upper have fallen slightly. Potential teachers have limited knowledge of school system employment, making it difficult for schools to be fully operational. Improved teacher recruitment will be critical for the sector to meet pressing future needs. Textbooks are lacking at all levels. (World Bank funds and FTI resources in 2007 were filling a portion of the gap.) Youth Vocation Training Center enrollment rates are small at the 53 existing training centers intended to provide basic, practical training. Tertiary education (first year) enrollment in 2004 was half of that in the final year of secondary in 2003 (54 percent for men and 41 percent for women). Education Sector Public Expenditure Allocation The recurrent budget for education rose as a portion of GDP from 3.4 percent in 2001 to 4.8 percent in 2007.30 Generally, a minimum of 5 percent of GDP is considered the minimum for quality education in a country. Between 2001 and 2006, recurrent education spending fluctuated between 20 percent and 25 percent of total government recurrent expenditures. However, in this regard Rwanda lags behind international best practice figures of 25 percent to 30 percent of total government recurrent expenditures. Table 5.36 compares education spending as a percent of GDP in selected neighboring countries. Table 5.36. Comparative Education Expenditures Rwanda Burundi Kenya Malawi Uganda Education as % of GDP 3% 5% 7% 6% 5% Education as % of Gov Spend 13% 13% 29% 25% 18% % Expenditure on Primary 45% 44% 63% 63% 62% % Expenditure on Secondary 20% 32% 25% 10% 24% % Expenditure on Higher 30% 24% 11% n.a. 12% Source: Education PER Report, July 2007, page 25. Original Source: World Education Indicators 2005. As can be seen, despite Rwanda's clear commitment to basic education and its intrasectoral reallocation changes that put less emphasis on tertiary education, the country still lags behind several of its neighbors in terms of education spending allocations. Despite recent efforts, Rwanda allocates 30 percent of its recurrent budget Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 135 on higher education, significantly higher than all other countries listed above. In primary education spending, Rwanda lags well behind its neighbors. It is in the above context that public expenditures in the education sector in Rwanda should be evaluated. Table 5.37 summarizes this recurrent education spending trend. Table 5.37. Education Sector Recurrent Budget 2001­07 In Real Terms (in millions RwF) Growth Growth % 2001- % 2006- 2001 2002 2003 2004 2005 2006 2007 07 07 Primary 19,607.7 31,283.9 21,615.8 22,485.8 25,368.7 27,019.5 31,186.0 59% 15% ow Pre-Primary -- -- -- -- -- 207.7 243.0 -- 17% ow Primary -- -- -- -- -- 26,811.9 30,943.0 -- 15% Secondary 7,214.4 13,419.9 9,358.0 12,418.2 12,087.0 11,842.9 17,613.0 144% 49% ow Tronc Commun -- -- -- -- -- 7,251.0 10,378.0 -- 43% ow Upper Secondary -- -- -- -- -- 3,228.2 4,225.0 -- 31% ow Teacher Education -- -- -- -- -- 400.7 1,087.0 -- 171% ow Technical -- -- -- -- -- 538.8 918.0 -- 70% ow Vocational -- -- -- -- -- 424.3 1,005.0 -- 137% Higher 16,196.1 23,317.0 18,578.4 17,644.9 17,343.3 18,981.7 21,896.0 35% 15% ow Higher Learning Instit. -- -- -- 15,089.6 9,210.0 10,168.2 15,056.0 -- 48% ow Student Fin. Agency -- -- -- 2,555.3 8,133.3 7,810.0 6,840.0 -- --12% (SFAR) Non-Formal 0.0 0.0 42.3 522.7 1,065.2 533.2 516.0 n.a. --3% Adult Literacy -- -- -- -- -- 533.2 516.0 -- --3% Institutional Support 2,772.4 2,401.3 2,428.7 2,448.8 2,666.2 2,581.6 3,245.0 17% 26% Total Educ Recurrent 45,790.5 70,422.2 52,023.2 55,520.4 58,530.5 60,958.9 74,456.0 63% 22% Bud. Science, Tech, Research 498.0 862.1 976.5 999.7 862.1 -- -- -- -- TOT Ministry & 46,288.5 71,284.3 52,999.7 56,520.1 59,392.6 60,958.9 74,456.0 61% 22% Teacher Salary Source: Education PER Report, July 2007, page 19. Original Source: Various MINEDUC and MINECOFIN documents. Note: This table includes higher education salaries that are listed under the development budget law. Recurrent spending on primary education increased 59 percent between 2001 and 2007, according to budget data, while secondary education has increased more than 144 percent over the same period. During this same period higher education and institutional support increased 35 percent and 17 percent, respectively, in the same period. Overall, this translates to an increase in the entire education recurrent budget of 191 percent in nominal terms. However, as the PER report shows, in real terms31 the growth in education allocations was less significant. Table 5.37 includes more than just MINEDUC spending on education. For example, central budget district spending such as capitation grants is included. In addition, expatriate salaries currently listed in the development budget under higher education projects are included in these totals to provide a better picture of overall recurrent expenditures. Furthermore, some data are missing in the table. This reflects the lack of consistent data and insufficient data (such as subprogram spending) over the years, which makes a trend analysis of the sector spotty. This also partially explains why the data on the education sector in other parts of this report show slightly different figures than are presented in the Education PER report. However, this is an overall problem across all sectors, not just the education sector. 136 World Bank Working Paper Rwanda continues to allocate a relatively large portion of its education spending on higher education. The absolute size of the higher education budget compared with secondary education and perstudent support is also much higher as parental fees for higher education continue to be complemented by state subsidies. However, the relative share between the different education levels has been changing, especially between secondary and tertiary education, and the government has restricted the relative growth of the higher education budget allocation since 2003. In this period (20032007), secondary education increased from 16 percent to 24 percent as a portion of total recurrent spending. Nonetheless, the budget portion of tertiary education has lingered around 30 percent since 2003. By economic classification of the education spending from 2003 to 2007, salaries and transfers to students have fallen while goods and services have risen. This is particularly true for 2007, where FTI funds and capitation grants have contributed to a larger source of funds for goods and services spending. Table 5.38 shows recurrent spending in the sector by economic classification: Table 5.38. Education Recurrent Budget by Economic Classification Economic Classification (%) 2003 2004 2005 2006 2007 Average Salaries & Benefits 70% 68% 65% 62% 48% 63% Goods & Services 16% 19% 24% 28% 43% 26% Transfers to Students 14% 14% 11% 9% 9% 11% TOTAL 100% 100% 100% 100% 100% 100% Source: Education PER Report, July 2007, page 123. Note: Salaries of teachers, both school and expatriate higher education, are included If the salary component of the capitation grant is redistributed, salaries in 2007 would account for 56 percent of the education recurrent budget and goods and services for 35 percent, instead of 48 percent and 43 percent, respectively. Salaries and benefits in the 2007 budget accounted for around 50 percent of recurrent spending, down from 70 percent in 2003, while goods and services increased from 16 percent to 43 percent in the same period. Meanwhile, transfers to students decreased slightly, showing an average over the period of 11 percent. Of the transfers to students in higher education, both internal and external have fallen in relative terms to less than 10 percent of recurrent expenditure. Experience with budgetary spending: MINEDUC has improved its overall recurrent budget execution, as seen in table 5.39. MINEDUC overspent its recurrent budget by 14 percent in 2002 and underspent by 13 percent in 2003, but has since improved its budget execution significantly. From 2004 to 2006, the variation between budgeted recurrent expenditures and actual (executed) expenditures was 1 percent­2 percent or less, implying improvements in mediumterm planning and possibly better coordination with MINECOFIN in the timing of the release of funds. However, varying execution rates can be found between the different subsectors in education. For example, primary and secondary education improved execution in 2004 and 2005 but missed targets rather significantly in 2006 with overexecution of 14 percent and underexecution of 35 percent in primary and secondary education, respectively. This variation in 2006 could possibly be explained by the 2006 decentralization efforts. Also, Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 137 part of this variation was explained by MINEDUC's difficulty estimating teacher salaries, which are then reflected on districts budgets. Furthermore, looking at execution rates in other sectors, there appears to be no overall trend emerging. However, average execution rates have varied as much as 16 percent in 2003 to 5 percent in 2005. The pattern of significant variations in execution indicates that greater measures are needed to improve budget planning based on realistic assumptions, which--in light of increased decentralization--would require further collaboration with districts. Table 5.39 shows the education recurrent budget allocation, actual budget execution, and variations in real terms 2001­06. Table 5.39. Education Recurrent Budget and Budget Execution 2001­06 BUDGET 2002 2003 2004 2005 2006 Pre-Primary & Primary 31,283.9 21,615.8 22,485.8 25,368.7 27,153.1 Secondary 13,419.9 9,358.0 12,418.2 12,087.0 11,305.3 Higher 23,317.0 18,578.4 17,644.9 17,343.3 18,981.7 Science, Tech, & Research 862.1 976.5 999.7 862.1 1,101.1 Non-Formal 0.0 42.3 522.7 1,065.2 810.4 Institutional Support 2,401.3 2,428.7 2,448.8 2,666.2 4,631.2 TOTAL: 71,284.3 52,999.7 56,520.1 59,392.6 63,982.8 EXECUTION Pre-Primary & Primary 27,095.4 26,382.5 22,870.2 25,797.8 23,322.2 Secondary 10,778.3 10,193.7 11,259.5 11,316.3 15,261.9 Higher 20,887.1 20,331.0 19,591.8 17,946.4 18,508.0 Science, Tech, & Research 728.6 732.4 925.0 709.7 1,055.1 Non-Formal 0.0 21.9 206.0 877.3 929.4 Institutional Support 1,951.2 2,015.6 2,540.1 2,229.5 4,631.2 TOTAL: 61,440.6 59,677.0 57,392.6 58,877.1 63,707.8 VARIATION Pre-Primary & Primary 4,188.5 ­4,766.7 ­384.4 ­429.1 3,830.9 Secondary 2,641.6 ­835.7 1,158.7 770.7 ­3,956.6 Higher 2,429.9 ­1,752.7 ­1,946.9 ­603.1 473.7 Science, Tech, & Research 133.5 244.1 74.7 152.4 46.0 Non-Formal 0.0 20.3 316.6 187.9 ­119.0 Institutional Support 450.1 413.1 ­91.3 436.8 0.0 TOTAL: 9,843.7 ­6,677.4 ­872.5 515.5 275.0 VARIATION % Pre-Primary & Primary 13% ­22% ­2% ­2% 14% Secondary 20% ­9% 9% 6% ­35% Higher 10% ­9% ­11% ­3% 2% Science, Tech, & Research 15% 25% 7% 18% 4% Non-Formal -- 48% 61% 18% ­15% Institutional Support 19% 17% ­4% 16% 0% TOTAL: 14% ­13% ­2% 1% 0% Average Variation Rate 16% 8% 10% 9% ­5% Source: Education PER Report, July 2007, page 26. Original Source: MINECOFIN and MINEDUC figures from budget books etc. In Rwanda, there are three teacher levels that qualify for different levels of salaries and bonuses. A2 refers to those teachers with primary teaching qualifications only; A1 refers to those teachers with a teaching certificate following two years of post secondary study; and A0 refers to those with a degreelevel certification. Table 5.40 shows teacher salary budget execution by districts in 2006. 138 Table 5.40. Teacher Salary Execution by District 2006 World Bank Working Paper Source: Education PER Report, July 2007, page 27. Original Source: 2006 budget Execution report. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 139 As can be seen in table 5.40, primary teacher salaries were overestimated, with only an 80 percent execution rate, while secondary teacher salaries where underestimated, with a 154 percent execution rate. Because teacher salaries constitute such a large portion of overall recurrent spending in the education sector, these variations in district budgets affect the overall subsector execution rate. However, overall recurrent execution was rather close to target in 2004­06 and this is consistent with the combined primary and secondary teacher salary budget execution for the districts in 2006 of 100 percent. Table 5.41 compares execution of teacher salaries in the districts for 2006 and planned allocations for 2007. Currently, teacher salaries are part of district budgets but are input through MINEDUC and implemented by MIFOTRA. Table 5.41. Teacher Salary Actual District Budget 2006 and Original District Budget 2007 In Real Terms (in millions RwF) Primary Teacher Salary Secondary Teacher Salary Primary & Secondary Salary 2006 2007 2006 2007 2006 2007 District Actual Budget Diff Actual Budget Diff Actual Budget Diff Bugesera 461.3 449.0 ­12.3 255.9 83.0 172.9 717.2 532.0 185.2 Burera 559.0 758.0 199.0 336.7 164.0 172.7 895.7 922.0 ­26.3 Gakenke 533.2 203.0 ­330.2 364.8 96.0 268.8 898.0 299.0 599.0 Gasabo 424.3 578.0 153.7 253.7 121.0 132.7 678.0 699.0 ­21.0 Gatsibo 388.4 514.0 125.6 230.1 86.0 144.1 618.5 600.0 18.5 Gicumbi 575.8 924.0 348.2 449.0 399.0 50.0 1,024.8 1,323.0 ­298.2 Gisagara 434.4 424.0 ­10.4 310.9 202.0 108.9 745.3 626.0 119.3 Huye 580.3 508.0 ­72.3 417.5 311.0 106.5 997.9 819.0 178.9 Kamonyi 462.4 574.0 111.6 275.0 77.0 198.0 737.4 651.0 86.4 Karongi 548.9 772.0 223.1 409.7 346.0 63.7 958.6 1,118.0 ­159.4 Kayonza 341.2 264.0 ­77.2 281.7 217.0 64.7 623.0 481.0 142.0 Kicukiro 346.8 297.0 ­49.8 199.8 85.0 114.8 546.6 382.0 164.6 Kiehe 370.4 86.0 ­284.4 234.6 52.0 182.6 605.0 138.0 467.0 Muhanga 673.5 562.0 ­111.5 465.8 307.0 158.8 1,139.3 869.0 270.3 Musanze 597.1 935.0 337.9 413.1 227.0 186.1 1,010.2 1,162.0 ­151.8 Ngoma 431.0 652.0 221.0 386.1 395.0 ­8.9 817.1 1,047.0 ­229.9 Ngoroero 571.3 469.0 ­102.3 356.9 172.0 184.9 928.3 641.0 287.3 Nyabihu 552.2 305.0 ­247.2 385.0 225.0 160.0 937.2 530.0 407.2 Nyagatare 559.0 582.0 23.0 403.0 284.0 119.0 961.9 866.0 95.9 Nyamagabe 529.8 681.0 151.2 370.4 275.0 95.4 900.2 956.0 ­55.8 Nyamasheke 694.8 744.0 49.2 474.8 329.0 145.8 1,169.6 1,073.0 96.6 Nyanza 478.2 404.0 ­74.2 342.3 237.0 105.3 820.5 641.0 179.5 Nyarugenge 484.9 465.0 ­19.9 294.1 272.0 22.1 779.0 737.0 42.0 Nyaruguru 449.0 265.0 ­184.0 270.5 115.0 155.5 719.5 380.0 339.5 Rubavu 633.1 541.0 ­92.1 426.5 279.0 147.5 1,059.6 820.0 239.6 Ruhango 334.5 457.0 122.5 287.3 302.0 ­14.7 621.8 759.0 ­137.2 Rulindo 501.7 132.0 ­369.7 336.7 41.0 295.7 838.5 173.0 665.5 Rusizi 606.1 406.0 ­200.1 396.2 198.0 198.2 1,002.3 604.0 398.3 Rutsiro 557.9 566.0 8.1 359.2 198.0 161.2 917.0 764.0 153.0 Rwamagana 415.3 340.0 ­75.3 322.1 270.0 52.1 737.4 610.0 127.4 RWANDA 15,095.8 14,857.0 ­238.8 10,309.7 6,365.0 3,944.7 25,405.4 21,222.0 4,183.4 Source: Education PER Report, July 2007, page 29. Original Source: 2006 budget execution report and 2007 budget law. 140 World Bank Working Paper The conclusion by the PER report conducted in 2007 on the issue of execution variations and the difficulty of estimating teacher salaries is that the method for teacher salary allocation must be improved. Although the report noted that the method for teacher salary allocations was new in 2006, the method in 2007 was based on workforce and salary limitations (3 percent) and did not consider what districts had the greatest needs. A review of the MIFOTRA payroll system indicates further why primary teacher salaries were lower than budget targets while secondary teacher salaries were above. Table 5.42 shows average gross salaries paid by MIFOTRA to teachers at different levels in March 2007 (on an annual basis). Although primary teachers are rather well paid at 3.9 times Rwanda's per capita GDP (5.1 times when bonuses are included), compared with bestpractice standards of 3.5­4.0 times, secondary teachers are paid between 2.5 and 3.3 times as much as primary teachers. Table 5.42. Average Teacher Salaries 2007 In thousand RwF RwF* Multiply of Ratio Ratio to A2 w/ Teacher Level Annual GDP/cap to A2 Bonus A2 Teachers 516.7 3.9 1.0 A2 Teachers with Bonus 664.2 5.1 1.3 1.0 A1 Teachers 1,388.6 10.6 2.7 2.1 A0 Teachers 1,972.3 15.0 3.8 3.0 Avg A1, A0 (all qualified 2nd) 1,684.0 12.8 3.3 2.5 Source: Education PER Report, July 2007, page 30. Decentralized expenditures: Decentralized expenditures refer to district budgets and to government financial support to schools. Budgets for districts include major funding for salaries to teachers, capitation grants, and school feeding as shown in table 5.43 for the budget year 2007. Teachers are paid directly through the government payroll into their bank accounts, and this amount shows up in districts' budgets. Other expenses are allocated from the Treasury. The 2007 budget allocation to districts was approaching 50 percent of the sector's recurrent funding. Capitation grants for primary schools contain two components: learning support and teacher support. For example, the capitation grant is supposed to help improve teacher quality by providing performancebased bonuses of RwF 12,287 per teacher per month and RwF 560 per pupil for hiring contract teachers to address higher teacher student ratios and reduce the burden of double shifts for teachers. However, the PER report in 2007 noted that the mechanism for how these amounts are set should be reviewed. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 141 Table 5.43. Education Budgets for Districts 2007 In Real Terms (in million RwF) Primary Tronc Commun Upper Secondary Total Capitation School Capitation School Edu District Teachers Grant Teachers Feeding Grant Teachers Feeding Technical Budget Bugesera 449.0 288.0 58.0 21.0 29.0 25.0 9.0 1.0 880.0 Burera 758.0 338.0 115.0 32.0 45.0 49.0 14.0 1.0 1,352.0 Gakenke 203.0 299.0 67.0 64.0 88.0 29.0 27.0 1.0 778.0 Gasabo 578.0 229.0 85.0 86.0 119.0 36.0 37.0 1.0 1,171.0 Gatsibo 514.0 316.0 60.0 45.0 62.0 26.0 19.0 1.0 1,043.0 Gicumbi 924.0 334.0 279.0 55.0 76.0 120.0 24.0 1.0 1,813.0 Gisagara 424.0 217.0 141.0 60.0 82.0 61.0 26.0 1.0 1,012.0 Huye 508.0 251.0 218.0 72.0 99.0 93.0 31.0 1.0 1,273.0 Kamonyi 574.0 267.0 54.0 40.0 55.0 23.0 17.0 1.0 1,031.0 Karongi 772.0 305.0 242.0 62.0 85.0 104.0 27.0 1.0 1,598.0 Kayonza 264.0 231.0 152.0 50.0 68.0 65.0 21.0 1.0 852.0 Kicukiro 297.0 148.0 60.0 45.0 62.0 26.0 19.0 1.0 658.0 Kiehe 86.0 260.0 36.0 46.0 64.0 16.0 20.0 1.0 529.0 Muhanga 562.0 271.0 215.0 70.0 96.0 92.0 30.0 1.0 1,337.0 Musanze 935.0 331.0 159.0 56.0 77.0 68.0 24.0 1.0 1,651.0 Ngoma 652.0 246.0 277.0 37.0 52.0 119.0 16.0 1.0 1,400.0 Ngoroero 469.0 303.0 121.0 66.0 91.0 52.0 28.0 1.0 1,131.0 Nyabihu 305.0 337.0 157.0 68.0 93.0 67.0 29.0 1.0 1,057.0 Nyagatare 582.0 306.0 199.0 57.0 79.0 85.0 25.0 1.0 1,334.0 Nyamagabe 681.0 296.0 193.0 75.0 104.0 83.0 32.0 1.0 1,465.0 Nyamasheke 744.0 367.0 230.0 97.0 134.0 99.0 42.0 1.0 1,714.0 Nyanza 404.0 232.0 166.0 35.0 49.0 71.0 15.0 1.0 973.0 Nyarugenge 465.0 159.0 190.0 47.0 64.0 82.0 20.0 1.0 1,028.0 Nyaruguru 265.0 249.0 80.0 58.0 80.0 34.0 25.0 1.0 792.0 Rubavu 541.0 296.0 195.0 52.0 71.0 84.0 22.0 1.0 1,262.0 Ruhango 457.0 253.0 212.0 97.0 134.0 91.0 42.0 1.0 1,287.0 Rulindo 132.0 260.0 29.0 89.0 122.0 12.0 38.0 1.0 683.0 Rusizi 406.0 317.0 138.0 47.0 65.0 59.0 20.0 1.0 1,053.0 Rutsiro 566.0 267.0 139.0 40.0 56.0 59.0 17.0 1.0 1,145.0 Rwamagana 340.0 229.0 189.0 53.0 73.0 81.0 23.0 1.0 989.0 TOTAL 14,857.0 8,202.0 4,456.0 1,722.0 2,374.0 1,911.0 739.0 30.0 34,291.0 EDUCATION ONLY: TOTAL DISTRICT BUDGET: 61,835.8 Source: Education PER Report, July 2007, page 41. Donor contributions: The donor community is working together to support one comprehensive education plan, and a wellestablished coordination mechanism is operating under the leadership of DFID. A significant development in the sector includes Rwanda's acceptance into the FTI (see table 5.44). 142 World Bank Working Paper Table 5.44. Education Budget, FTI, and LTSFF Allocations 2006 and 2007 In Real Terms (in million RwF) 2006 2007 FTI With FTI Recurrent Budget (RB) Budget Budget Proposal Proposal LTSFF % LTSFF Pre-Primary 207.7 89.0 154.0 243.0 234.0 104% Primary 26,811.9 26,656.0 4,287.0 30,943.0 36,065.0 86% Tronc Commun 7,251.0 9,834.0 544.0 10,378.0 8,952.0 116% Upper Secondary 3,228.2 4,140.0 85.0 4,225.0 4,400.0 96% Teacher Education 400.7 478.0 609.0 1,087.0 817.0 133% Technical Education 538.8 786.0 132.0 918.0 1,034.0 89% Vocational Education 424.3 865.0 140.0 1,005.0 565.0 178% Higher Education 17,977.1 19,476.0 0.0 19,476.0 20,377.0 96% Adult Literacy 533.2 416.0 100.0 516.0 516.0 100% Institutional Support 2,581.6 3,245.0 0.0 3,245.0 3,322.0 98% TOTAL 59,954.3 65,985.0 6,051.0 72,036.0 76,282.0 94% Development Budget (DB) Pre-Primary -- -- -- -- -- -- Primary -- 2,195.0 4,069.0 6,264.0 7,987.0 78% Tronc Commun -- 584.0 1,846.0 2,430.0 4,609.0 53% Upper Secondary -- 515.0 532.0 1,047.0 1,333.0 79% Teacher Education -- 150.0 1,350.0 1,500.0 1,500.0 100% Technical Education -- 319.0 201.0 520.0 520.0 100% Vocational Education -- -- -- 125.0 125.0 100% Higher Education construction -- 2,724.0 0.0 2,724.0 1,156.0 236% Higher Educ. Ex-pat Salaries -- 2,420.0 0.0 2,420.0 1,628.0 149% Adult Literacy -- -- -- -- -- -- Institutional Support -- -- -- -- -- -- Other Projects/Tech. Assist. -- 3,052.0 0.0 3,052.0 2,850.0 107% TOTAL -- 11,959.0 8,123.0 20,082.0 21,708.0 93% TOTAL RB & DB 59,954.3 77,944.0 14,174.0 92,118.0 97,990.0 Source: Education PER Report, July 2007, page 36. Part of donor efforts includes attempts to ensure that all donor funding is captured to facilitate resource planning. But further improvements in coordination, dialogue, and efficiencies are needed and ought to include offbudget projects to a greater extent. A significant step to address donor coordination and proper and timely funding for the sector is Rwanda's acceptance into the FTI to help the sector meet its MDG. The FTI is to fund additional support in 2007­10 of US$200 million. The Catalytic Fund Strategy Committee accepted a request for US$44 million additional funds for 2008­09, which will help MINEDUC reach the many LTSFF targets for its programs. However, significant shortfalls in funding still existed at the time of the PER Review in July 2007. Semiautonomous government agencies: About onethird of the higher education budget goes to the semiautonomous agencies in higher education, of which 99 percent goes to the Student Financing Agency of Rwanda (SFAR), an agency set up after a 2003 Cabinet decision. The remaining subsector funds for higher education are spent on support for the six major governmentfunded higher learning institutions. Funds for these semiautonomous agencies are listed in table 5.45. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 143 Table 5.45. Higher Education Recurrent Budget 2007 In Real Terms (in million RwF) Goods & Academic Student Agency Salaries Services Support Support TOTAL Institutions 11,888.0 2,018.0 1,045.0 0.0 14,951.0 % 80% 13% 7% 0% 100% National Univ. of Rwanda (NUR) 3,437.0 551.0 323.0 -- 4,311.0 Kigali Instit. of Science & Tech. (KIST) 2,354.0 411.0 250.0 -- 3,015.0 Kigali Institute of Education (KIE) 1,138.0 216.0 125.0 -- 1,479.0 Kigali Health Institute (KHI) 1,007.0 191.0 77.0 -- 1,275.0 Higher Instit. of Agri/Animal Husb.(ISAE) 902.0 285.0 204.0 -- 1,391.0 School of Finance & Banking (SFB) 630.0 364.0 66.0 -- 1,060.0 Ex-pat Salaries 2,420.0 -- -- -- 2,420.0 Support 173.0 114.0 0.0 6,633.0 6,920.0 % 3% 2% 0% 96% 100% Planning, Policy, and Capacity Building 1.0 20.0 -- -- 21.0 Nat. Council for Higher Edu. (NCHE) 46.0 12.0 -- -- 58.0 Student Finan. Agency of Rw.(SFAR) 126.0 82.0 -- 6,633.0 6,841.0 Sub-Sector Total 12,061.0 2,132.0 1,045.0 6,633.0 21,871.0 % 55% 10% 5% 30% 100% Source: Education PER Report, July 2007, page 38. Original Source: MINECOFIN 2007 Budget. In addition to the agencies listed above, CNER manages three main national examinations in Grades 6, 9, and 12. As previously noted, only limited data on expenditures for this agency exists. Funding for CNER comes from both MINEDUC and fees (see table 5. 46). Table 5.46. CNER Income 2001­06 In Real Terms (in million RwF) Projected Income 2001 2002 2003 2004 2005 2006 2007 Expenditure 720.7 1,014.7 1,037.5 1,124.2 1,043.6 1,187.5 1,480.0 Fees 132.2 343.3 425.6 282.1 384.7 336.7 0.0 Source: Education PER Report, July 2007, page 39. Original Source: CNER. Unit Cost: Unit cost is a valuable measure of both the absolute and relative expenditure per student within the education system. Although an exact unit cost measure is not possible, an estimate of unit cost provides valuable insight into the spending patterns in education. Table 5.47 shows the estimated recurrent government unit expenditure for 2001­06. 144 World Bank Working Paper Table 5.47. Apparent Unit Government Recurrent Expenditure by Level 2001­06 In Real Terms Sub-Sector Sector Budget in millions RwF 2001 2002 2003 2004 2005 2006 Pre-Primary & Primary 19,607.7 31,283.9 21,615.8 22,485.8 25,368.7 27,018.4 Secondary 7,214.4 13,419.9 9,358.0 12,418.2 12,087.0 11,844.1 Higher 16,196.1 23,317.0 18,578.4 17,644.9 17,343.3 17,977.1 Total Enrollment in Thousands Primary 1,475.6 1,534.5 1,636.6 1,752.6 1,857.8 2,020.0 Tronc Commun 99.9 114.2 131.4 142.2 156.3 Upper Secondary 57.6 64.9 72,124.0 76.3 82.5 Higher Education 28.0 37.2 Publicly Funded Enrollment in Thousands Primary 1,475.6 1,534.5 1,636.6 1,752.6 1,857.8 2,020.0 Tronc Commun 51.2 57.1 66.5 77.9 88.9 102.7 Upper Secondary 28.5 31.7 34.1 36.1 36.5 37.6 Higher Education 10.4 10.2 12.2 14.5 13.4 16.2 Apparent Unit Cost RwF Primary 13,287.9 20,387.4 13,207.7 12,830.3 13,655.0 13,376.2 Secondary 90,538.0 151,077.1 92,963.8 108,935.0 96,424.2 84,402.3 Higher 1,564,158.4 2,280,159.9 1,521,397.4 1,220,622.8 1,296,444.9 1,110,154.3 Ratio of Apparent Cost to Primary Primary 1.0 1.0 1.0 1.0 1.0 1.0 Secondary 7.0 7.0 7.0 8.0 7.0 6.0 Higher 118.0 112.0 115.0 95.0 95.0 83.0 Source: Education PER Report, July 2007, page 31. Original Source: Figures from MINECOFIN and MINEDUC. Table 5. 47 shows all public or government expenditure and students supported by the government. As stated earlier, Rwanda has unusually high government support for tertiary education while rather low support for primary education. In 2002, the budget for each publicly funded higher education student was 112 times higher than the amount funded for each primary student, meaning a budget for 10,226 students of higher education equaled the budget for 1.2 million primary school students. Even though the ratio came down from 112 times in 2002, the ratio remained significant at 83 times the cost for a primary student in 2006. As indicated in table 5.47, estimated government recurrent unit spending per student in 2006 (in real terms) was about RwF 13,376, RwF 84,402, and RwF 1,110,154 for primary, secondary, and higher education students, respectively. The higher education budget in 2006 equaled the primary education budget that supported 1.3 million primary students. In 2006, the year of full decentralization efforts, secondary education was split between tronc commun and upper secondary education. Table 5.48 estimates nominal and real unit expenditures for that year. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 145 Table 5.48. Nominal and Real Unit Education Expenditures 2006 Levels of Secondary Education 2006 Nominal 2001 Real Tronc Commun 62,906.0 46,267.0 Upper Secondary 108,724.0 79,966.0 Source: Education PER Report, July 2007, page 32. As can be seen in table 5.48, more funds are allocated to the upper secondary cycle. However, this is a common phenomenon, partially explained by more inefficiencies because of a smaller number of students combined with a more specialized curriculum, which contribute to increased costs in teacher and material specializations. However, considering inflation in unit cost for each subsector provides a slightly different picture. Between 2001 and 2006 in real terms (2001 prices), as shown in table 5.49, the budget per student for primary education increased, but not for tertiary education. In real terms with further FTI funding in 2007 forward, this trend is expected to continue. Table 5.49. Education Enrollment and Real Unit Expenditure Change 2001­06 Education Level 2001 2002 2003 2004 2005 2006 Enrollment Growth Primary 100% 104% 111% 119% 126% 137% Secondary 100% 111% 126% 143% 157% 176% Higher 100% 99% 118% 140% 129% 156% Real Expenditure (in real terms) Primary 13,288 20,387 13,208 12,830 13,655 13,376 Secondary 90,538 151,077 92,964 108,935 96,424 84,402 Higher 1,564,158 2,280,160 1,521,397 1,220,623 1,296,445 1,110,154 Ratio to 2001 Expenditure Primary 100% 153% 99% 97% 103% 101% Secondary 100% 167% 103% 120% 107% 93% Higher 100% 146% 97% 78% 83% 71% Source: Education PER Report, July 2007, page 34. Original Source: Figures from MINECOFIN. Household expenditures on education: The following section summarizes household expenditures on education. Table 5.50 shows that public schools play an important role in the first years of schooling, but at higher grade levels, private schooling is more common. 146 World Bank Working Paper Table 5.50. Sample* Households by School Type and Level** Type of School Primary Trunc Commun Second Cycle TOTAL Public 6,505 95% 216 65% 126 59% 6,847 Private 369 5% 114 35% 88 41% 571 TOTAL 6,874 100% 330 100% 214 100% 7,418 Source: Education PER Report, July 2007, page 42. Note: Public Schools include government supported schools in private ownership. ** The Education PER team allocated all individuals for whom there was expenditure data to a school level based on age, highest grade completed and age. There were 7,418 individuals [about a quarter of the total sample] who were in school in the twelve months prior to the survey and for whom the data was sufficient [see Table 2.17 in PER]. Only five percent of pupils were in fully private schooling at primary level but this rose to 35 percent at tronc commun and 41 percent at the second cycle of secondary. The average expenditure for each household with any education spending was about RwF 44,000 per year. Average expenditures ranged from RwF 5,000 in the poorest households to RwF 150,000 in the richest households (see table 5.51). Table 5.51. Mean Total Education Expenditure by Household Per Household Per Person Per Funded Person RwF Quintile Amount Ratio* Amount Ratio* Amount Ratio* Lowest 5,086.0 1 788.0 1 2,057.0 1 Second 13,132.0 2 1,954.0 2 4,779.0 2 Third 19,513.0 4 3,078.0 4 7,147.0 3 Forth 32,885.0 6 5,188.0 7 11,551.0 6 Highest 149,774.0 29 24,847.0 32 55,222.0 27 ALL 44,078.0 8.4 7,171.0 9.2 16,151.2 7.8 Source: Education PER Report, July 2007, page 43. Original Source: Preliminary Poverty Data 2007. Includes only households with education expenditure. *As education expenditure per member of the household [to see whether household size was influential] or by the number of persons on whom the education expenditure was spent [to see whether there were a larger proportion of children in education]. Total consumption expenditure for households that had any education expenditure averaged just under RwF 1 million, as illustrated in table 5.52. On average, the richest households spent 14 times as much as the poorest quintile of households. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 147 Table 5.52. Mean Total Household Consumption Expenditure by Quintile Income RwF Educ Number Quintile Amount Ratio %* Households Lowest 197,874 1 3% 854 Second 355,144 2 4% 851 Third 478,623 2 4% 844 Forth 725,629 4 5% 844 Highest 2,853,290 14 5% 997 ALL 922,112 4.6 5% 4,390 Source: Education PER Report, July 2007, page 43. Original Source: Preliminary Poverty Data 2007. Note: Preliminary Poverty data includes only households with education expenditure. *Portion of total household expenditure spent on education. Median household expenditure per primary student in 2005/2006 was RwF 1,845, with minimal change since 2000/2001 (see table 5.53). No major differences between male and female student costs were found in the data. Table 5.53. Median Education Expenditure per Student in Last 12 Months of Primary Education 2000/2001 2005/2006 RwF Quintile Male Female All Male Female All Lowest 975 928 950 1,233 1,136 1,174 Second 1,335 1,417 1,412 1,516 1,729 1,616 Third 1,775 1,856 1,817 1,827 1,810 1,817 Forth 2,363 2,261 2,300 2,050 2,161 2,117 Highest 3,558 3,491 3,505 3,653 3,327 3,466 ALL 1,817 1,795 1,798 1,844 1,854 1,845 Source: Education PER Report, July 2007, page 44. Original Source: Preliminary Poverty Update Report 2007. Notes: All values are expressed in January 2006 prices using CPI. Median education expenditure for secondary schooling is shown in table 5.54, broken down by the three types of secondary schooling in Rwanda: full government schooling, fully subsidized schools (through other agencies, churches, and so forth), and fully private schools. As can be noted, tronc commun and upper secondary median education expenditure rise progressively through these types of schooling. 148 World Bank Working Paper Table 5.54. Median Education Expenditure per Type of School Tronc Commun Secondary Upper RwF Free Free Quintile Public Subsid Private Total Public Subsid Private Total Lowest 18,000 13,500 5,100 16,100 29,000 125,000 65,500 60,750 Second 40,075 44,750 5,300 40,525 83,800 86,900 93,000 84,900 Third 41,600 55,500 89,400 51,000 74,500 87,000 79,700 80,700 Forth 55,250 72,500 60,375 58,400 75,990 98,200 103,450 88,650 Highest 73,650 84,250 94,500 81,300 79,000 82,500 140,000 100,000 ALL 55,900 71,060 86,900 65,500 76,990 88,000 120,820 90,000 Source: Education PER Report, July 2007, page 45. Original Source: Preliminary Poverty Data 2007. Includes only households with education expenditure Median education expenditures per student by age across all levels of schooling are shown in table 5.55. Up to midteens, individual expenditure is quite low, with 50 percent of expenditures less than RwF 2,500 per year. Overall household spending on education is in general dominated by three types of expenditure: registration fees, uniforms, books and school materials. As more basic education becomes free, the average cost for the younger age group can be maintained relatively low. Table 5.55. Median Education Expenditure per Student by Age Category Primary Tronc Commun Secondary Cycle Post-Secondary Age 7­12 Age 13­15 Age 16­18 Age 18­25 Quintile Male Female All Male Female All Male Female All Male Female All Lowest 905 970 930 1,690 1,440 1,520 2,510 1,575 1,940 2,900 5,000 4,100 Second 1,315 1,690 1,530 2,105 2,000 2,030 3,100 3,090 3,090 40,150 61,800 42,400 Third 1,610 1,700 1,675 2,500 2,425 2,480 3,860 3,500 3,775 57,500 62,000 57,525 Forth 2,000 2,110 2,060 2,800 3,335 3,010 6,100 6,720 6,500 61,500 69,000 63,850 Highest 5,400 4,710 5,000 5,820 6,195 6,165 55,000 36,550 45,500 89,500 113,750 97,400 ALL 1,768 1,870 1,800 2,560 2,500 2,520 5,000 4,900 5,000 64,000 75,975 70,000 Source: Education PER Report, July 2007, page 46. Original Source: Preliminary Poverty Data 2007. Includes only households with education expenditure. Education Sector Results Rwanda made substantial achievements in the education sector. The move toward a resultsoriented budget and MTEF was accompanied by reforms in the education sector. A number of significant actions were taken, including the development of an ESSP aligned to the resultsoriented sector MTEF, the initiation of capitation grants for primary schools, and provision of grants to promote access for the most vulnerable segments of the population. Furthermore, capitation grants were transferred successfully from the central government to primary schools to cover the loss in fee income from the introduction of the government's policy of feefree primary education; the size of the capitation grant increased from RwF 300 per student in 2004 to RwF 2,500 per student in 2006 (and RwF 5,300 per student in 2007). An Organic Education Law was passed and the National Council for Higher Education and Higher Education Law were adopted, with provisions aligned with the sector financing Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 149 strategy and MTEF. Furthermore, District Education Funds were adopted to support particularly needy students. The sector financing strategy was also updated to give it a 10year perspective and put more emphasis on quality issues, including the reduction of pupilteacher ratios. On the strength of an LTSFF, developed with technical assistance, Rwanda applied for funds to the Education for All FTI, resulting in an award of US$70 million for 2007­09 to help fill the sector financing gap. Table 5.56 lists the key reform measures undertaken in the education sector between 2004 and 2007. Table 5.56. Reform Measures Undertaken in Education 2004­07 Theme/Sector Reform Measures Undertaken Year Education National Council of Higher Education established 2006 Independent Student Financing Agency of Rwanda established 2006 Long-Term Strategy and Financing Framework for education developed 2006 Higher Education Law passed 2005 Organic Education Law passed 2004 Education Sector Strategic Plan (ESSP) and sector MTEF developed 2004 Capitation grants established 2004 District Education Funds adopted to support particularly needy students 2004 Source: Education PER Report, July 2007. Today, primary and trunc commun education are free in Rwanda. Primary education includes six years of schooling, from Grade 1 to Grade 6 for ages 7­12. Furthermore, tronc commun (ages 13­15) is publicly supported, but 34 percent of this level's enrollment is in private schools without any public funding. In the past, tronc commun was mostly boarding school, but it is rapidly expanding since tronc commun became feefree. However, upper secondary remains boardingschool based, and enrollment is growing more slowly than tronc commun and higher education. Consequently, the government is promoting regularity and gender balance in higher education. The establishment of a National Council for Higher Education is complete, and an evaluation of the status of higher education in Rwanda was undertaken in 2007, evaluating ways to promote female enrollment in technical training, standardize educator salaries, and ensure consistency in the provision of quality education. Furthermore, the government is providing institutional support to MINEDUC, the Rwanda National Education Council, and the National Curriculum Development Center. Support to students in the form of financial assistance loans has benefited 12,800 students. Nonetheless, enrollment for both trunc commun and upper secondary is still low compared with international standards. The same is true for higher education, but tertiary enrollment is experiencing a rapid increase. Today, Rwanda has 12 private and 6 public higher education institutions, and students in the 6 public higher education institutions are not charged a fee; their living expenses are paid by the state and are classified as loans. Many Rwandese students currently study abroad with expenses paid by the state.32 Rwanda is on track to meet its MDG 2 goal of achieving universal primary education. There has been progress toward universal primary education, with net 150 World Bank Working Paper enrollment increasing from 72 percent in 2000 to 95 percent in 2006. Net primary school enrollment reached 95.8 percent in 2007, exceeding the PRSP target for 2008. The 2007 net enrollment of females of 96.8 percent was higher than that of 94.7 percent for males. In secondary education, enrollment rates were around 10 percent in 2006 (10.6 percent for males and 9.5 percent for females), increasing to 13.1 percent in 2007. The recent introduction of feefree tronc commun in schools in the public sector should help to increase enrollment in secondary education and address the disparities in enrollment between the poor and the rich households. Finally, the gross enrollment rate at the tertiary level (approximately 3 percent) is comparable to the regional average. Completion rates also increased, from 22 percent in 2000 to 51 percent in 2006, and remained stable in 2007. Table 5.57 summarizes the development of the education system from 2002 to 2006. Table 5.57. Enrollment by Level 2002­06 In thousands students Growth % Level 2002 2003 2004 2005 2006 2002­06 Primary Enrolment 1,534.5 1,636.6 1,752.6 1,857.8 2,020.0 32% Female % 50% 50% 51% 51% 51% -- Gross enrollment rates 115 122 130 136 145 -- Tronc Commun 99.9 114.2 131.4 142.2 156.3 57% Female % 49% 48% 48% 48% 48% -- Gross enrollment rates 15 17 20 22 24 -- Upper Secondary 57.6 64.9 72.1 76.3 82.5 43% Female % 50% 49% 48% 47% 47% -- Gross enrollment rates 9 10 11 12 13 -- Higher Education 15.9 20.4 25.2 27.8 37.1 133% Female % 34% 37% 39% 40% 42% -- Source: Education PER Report, July 2007, page 14. Original Source: MINEDUC Statistics, Population Census, PER Calculations. As table 5.57 shows, Rwanda has shown rapid increase in enrollment and is on track to meet its MDG enrollment objectives. Primary education enrollment increased 32 percent between 2002 and 2006, while tronc commun increased 57 percent in the same period. The highest growth rate was in higher education, where enrollment went up by 133 percent, while upper secondary increased by only 43 percent. While overall enrollment grew rapidly, female enrollments at primary and tronc commun levels remained consistent, female enrollment in upper secondary saw a slight decrease, and female enrollment higher education went from 34 percent to 42 percent between 2002 and 2006. Additionally, the first half of 2007 saw an increase in use of technical equipment, especially in upper secondary schools, where computers were distributed and connected to the Internet. Although budget execution in the first and second quarters of 2007 was only 32 percent, these key programs had very high delivery. Because of the timelines involved in tendering processes, budget execution is expected to jump dramatically in the immediate future as infrastructure projects are realized. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 151 Education Sector Geographic Equity Household expenditure on education varies considerably across Rwanda. These geographical differences are illustrated in the tables that follow. Mean annual education expenditure per household by region, shown in table 5.58, shows high variations between regions, from less than RwF 4,000 in the lowest quintile in the Southern Province to more than RwF 250,000 in the highest quintile in Kigali. Table 5.58. Mean Annual Education Expenditure per Household by Region in RwF City of Eastern Western Southern Northern Total Quintile Kigali Province Province Province Province Rwanda Lowest 7,839 6,754 5,405 3,943 5,016 5,086 Second 16,919 9,791 12,629 16,987 9,119 13,132 Third 20,564 19,082 20,294 25,912 11,240 19,513 Forth 44,509 36,996 34,631 25,138 24,915 32,885 Highest 253,917 73,893 69,325 80,259 56,477 149,774 ALL 181,906 30,888 25,068 24,366 18,213 47,534 Source: Education PER Report, July 2007, page 49. Original Source: Preliminary Poverty Data 2007, includes only households with education. Mean annual education expenditure per funded individual by region shows similar disparities across regions. The relative expenditures are very similar both between quintiles in the same region and between regions. Nationally, the richest households spend more than 25 times as much as the poorest households per child supported in education. Overall, the mean annual education expenditure per individual for Rwanda is RwF 17,430, compared with RwF 91,833 for the highest quintile in Kigali or RwF 65,794 for all quintiles in the city of Kigali. Meanwhile, the lowest quintiles across all regions except the city of Kigali spend no more than RwF 2,993 per funded individual, with Western and Southern provinces spending less than RwF 2,000 per funded child (see table 5.59). Table 5.59. Mean Annual Education Expenditure per Funded Individual by Region in RwF City of Eastern Western Southern Northern Total Quintile Kigali Province Province Province Province Rwanda Lowest 3,330 2,793 1,958 1,739 2,074 2,057 Second 6,007 3,715 4,575 5,931 3,708 4,779 Third 8,670 6,857 7,172 9,486 4,308 7,147 Forth 15,435 12,252 11,975 11,147 7,597 11,551 Highest 91,833 33,046 25,298 28,546 20,216 55,222 ALL 65,794 12,137 8,968 9,132 6,477 17,430 Source: Education PER Report, July 2007, page 49. Original Source: Preliminary Poverty Data 2007, includes only households with education. 152 World Bank Working Paper The mean number of persons per household funded for education expenditures is shown in table 5.60 for each region. There appears to be very little variations across regions and even within regions across quintiles. Most households appeared to have about two members that incurred educational expenditures. The only outlier in the data was again the city of Kigali, where on average 2.6 persons had educational support in each household. Table 5.60. Mean Number of Persons per Household Funded for Education, by Region City of Eastern Western Southern Northern Total Quintile Kigali Province Province Province Province Rwanda Lowest 2.1 2.1 2.1 1.9 2.1 2.0 Second 2.3 2.1 2.2 2.1 2.1 2.1 Third 2.1 2.1 2.2 2.2 2.0 2.1 Forth 2.3 2.3 2.2 1.8 2.0 2.1 Highest 2.8 2.3 2.2 2.2 2.4 2.5 ALL 2.6 2.2 2.1 2.0 2.1 2.2 Source: Education PER Report, July 2007, page 50. Original Source: Preliminary Poverty Data 2007, includes only households with education. Mean individual education expenditure by age and location indicates a strong association between location of household and the average education expenditure, as illustrated in table 5.61. For example, for all persons ages 7­25, the average individual education expenditure in urban locations was almost five times that in rural areas. Furthermore, for each age category, the relative difference between rural and urban household expenditure per individual is large but increases after basic education. Table 5.61. Mean Individual Education Expenditure by Age and Location In RwF Age Category Urban Rural Rwanda Primary Age 5,200 1,600 1,800 Tronc Commun 6,810 2,210 2,520 Secondary Cycle Age 33,400 3,780 5,000 Post Secondary Cycle Age 88,000 60,000 70,000 ALL 9,425 2,000 24,000 Source: Education PER Report, July 2007, page 50. Original Source: Preliminary Poverty Data 2007, incl. only households with edu. Mean individual education expenditure by quintile and location shows that there was a clear increase in expenditure from lowest to highest quintile, and for each quintile, mean urban expenditure was higher than for rural households. Overall in Rwanda, the mean individual educational expense for both urban and rural locations was RwF 2,450, compared with the mean of RwF 10,150 for all urban quintiles and RwF 2,000 for all rural quintiles (see table 5.62). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 153 Table 5.62. Mean Individual Education Expenditure by Quintile and Location in RwF Age Category Urban Rural Rwanda Lowest 1,250 1,220 1,228 Second 2,100 1,800 1,800 Third 2,750 2,020 2,100 Forth 4,700 2,500 2,800 Highest 400,000 3,380 12,850 ALL 10,150 2,000 2,450 Source: Education PER Report, July 2007, page 50. Original Source: Preliminary Poverty Data 2007, incl. only households with edu. Education Sector Challenges The current status of education as it compares to the EDPRS education sector priorities highlights a number of challenges. The primary student­teacher ratio at 71:1 in 2006 was high and even continued to rise to 74:1 in 2007, coupled with high repetition and dropout rates of 18.8 percent and 14 percent, respectively. Access for girls at the levels of higher education, especially technical, remains unequal to that of boys (though in general, more girls attend higher education than boys). Of all secondary school leavers, only 3.2 percent attain higher education, and those that do often receive education by poorly qualified professors. Furthermore, the quality of primary education remains low. To accommodate the increasing demand and improve quality, increasing the supply of welltrained teachers at the primary level will be critical. Also, concern has been raised about discrepancies between project figures and actual expenditure, reportedly as a result of a long tendering process. To address this concern, the government is looking to improve consideration for procurement timelines when planning, so that budget execution is better aligned with planned activities. To address outstanding issues in the education sector, the government during the 2006 Joint Budget Support Review in September set out to develop subsector strategies with annual action plans, implement a costsharing strategy for higher education, prioritize the teaching of science and technology with an emphasis on closing the gender gap among enrollments, strengthen the sectorwide strategy and support subsector priorities and action plans, and reinforce its commitment to attaining the MDGs in education. Health Health Sector Overview The Rwanda Health Sector Strategic Plan (HSSP) is the implementation arm of the EDPRS. The HSSPI provides an overarching mediumterm framework (2005­09) for health sector support aiming at reducing poverty and improving the health status of the population. The HSSP operationalizes the health sector policy and the EDPRS; it was developed with development partners under the umbrella of a SWAP and defines the 154 World Bank Working Paper health sector's objectives for 2005­09. Seven major goals that were translated into the ordinary budget MTEF structure for 2007­09 were laid out in the health sector policy and HSSP (see table 5.63). Table 5.63. Policy Objectives and Expected Impact on Poverty and Growth of the HSSP-I (2005­09) Policy Objective Impact on Poverty and Growth 1) To ensure the availability of human · Focus on rural placement of human resources benefits resources the comparatively poorer rural population. 2) To ensure the availability of and access to · Drugs pricing policy with subsidies for the poor quality drugs, vaccines, and consumables improves access to drugs. · Vaccination campaigns benefit children from poor households. 3) To improve the geographical accessibility · Improved facilities for emergency transport from rural to health services health centers benefit the comparatively poorer rural population. 4) To improve financial accessibility to health · Mutuelle scheme removes financial barriers to health services care. 5) To improve the quality of and demand for · Integrated management of childhood illness and services and strengthen the control of maternal health programs directly target vulnerable disease groups. · Malaria programs have a secondary productivity impact on the population. 6) To improve national referral hospitals and · Multiresistant tuberculosis treatment benefits poor research and treatment institutions parts of the population. · Improved performance of national referral hospitals reduces demand for treatment courses abroad. 7) To reinforce institutional capacity · Health sector reform is aimed at improving efficiency through contracting. Source: HSSP 1 (2005­09). By the end of 2007 it was apparent that the HSSPI needed updating because important new programs had been initiated since the plan was adopted, most activities had been implemented, and many 2009 targets had already been reached or would be reached in 2008. In September 2007, the government of Rwanda finalized the EDPRS 2008­12 and, to align the HSSP to the EDPRS, MINISANTE decided that the HSSPI would be internally and externally evaluated in 2008, and that the HSSPII would subsequently be developed one year earlier than originally envisioned. Rwanda's health financing policy is critical to meet the health sector's policy objectives and to make a significant impact on poverty and growth. The goals of Rwanda's health financing policy are mainly to raise funds for the sector, align sector objectives with available resources, and foster sustainable financing and equity within the health system. To achieve these objectives, MINISANTE has developed various innovative health financing strategies such as communitybased health insurance schemes (mutuelles), performancebased financing, and fiscal decentralization. The below section takes a closer look at these innovative financing strategies. Riskpooling mechanism: mutuelles: Mutuelles are communitybased health insurance schemes designed to pool risks and therefore improve financial access to health services, particularly for underserved populations. In essence, a mutuelle spreads Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 155 the financial risk across its membership base, and this risk decreases further as membership increases. To stimulate demand for health services and given the low level of resources available at the health center level, a pilot mutuelle scheme was initiated in 1999 by the government with technical support from USAID in 53 health centers in the three districts of Byumba, Kabgayi, and Kabutare. In 2005, Rwanda adopted a three year (2005­07) national policy for mutuelles aiming to develop health mutuelles nationwide. There are three levels of pooling resources to pay for health services delivered to mutuelle members: Health center level. Mutuelle sections finance all health services delivered at health centers. Their resources come from members' contributions to the scheme (RwF 1,000 per capita), the sector (according to its financial capacities), and development partners who usually pay for indigents. District level. Health mutuelles cover the cost of all services delivered at district hospitals. Financial resources come from mutuelle sections (10 percent of their revenue), districts (RwF 500), development partners, and from the National RiskPooling Basket. National level. The National RiskPooling Basket covers the costs of all services delivered in referral hospitals (that is, CHK, Butare Hospital, Ndera psychiatric center, and the center for psychosocial consultation). Funds flowing to this National RiskPooling Basket come from MINISANTE, civil servants' contributions to their Medical Insurance Scheme, contributions to the Military Medical Insurance Scheme, and contributions to private health insurance schemes. The bulk of the financing for health mutuelles comes from household dues, with support from the government and external partners. As a result of the expansion of health mutuelles, resources mobilized by the health mutuelle section became very substantial (RwF 8.3 billion in 2006). Health centers captured the largest share of the increase in health facilities revenues, showing that these additional resources enabled a boost in the quality of services offered at local health facilities. The coverage of health mutuelles targeting rural communities and the informal sector has experienced a dramatic increase over time, reaching 75 percent in 2007. Rwanda now possesses a basic health insurance infrastructure based on 403 mutuelle sections and 30 district health mutuelles that are contributing to the transformation of resource allocation mechanisms in the health sector. After the demonstration of the political will of the government and its capacity to extend health insurance coverage to the majority of the population, Rwanda needs sustained efforts in the coming decade to consolidate the financial and institutional sustainability of its decentralized health insurance systems by strengthening internal financial mechanisms, management capacities, and the governance of health mutuelles. Performancebased financing: Performancebased financing (PBF) is a strategy that links incentives to performance; it was developed by donors and MINISANTE to improve health service delivery. PBF started in Rwanda as early as 2001 with impetus from external actors: the NGO HealthNet International started a PBF scheme in 2001 in the former Butare province. In early 2002, the NGO Memisa/Cordaid started a PBF 156 World Bank Working Paper scheme in Cyangugu province. Finally, in 2005, the Belgian Technical Cooperation also started a PBF scheme: the "Ville de Kigali" model. Evidence generated by these pilot programs on the relative performance of PBF compared with traditional inputbased financing mechanisms provided a base for MINISANTE's decision to roll out PBF in all health facilities in Rwanda. A national PBF model for health centers was designed in 2006 through an extensive consultative process involving all key stakeholders. The national PBF model combines features of the piloted schemes and is adapted to the institutional environment resulting from implementation of political and decentralization reforms. Since 2006, PBF has been a major pillar of MINISANTE's strategy; it was implemented to provide additional resources and incentives to health workers to improve efficiency and the quality of care. Indeed, the quality of care had become a particularly salient issue as, with the expansion of health mutuelles, utilization rates increased drastically in all health centers. PBF is currently implemented at three levels: community, health center, and hospital levels. In the national model for health centers, payments for performance are based on the quantity of outputs achieved conditional on the quality of services delivered. Health center staff can increase their performance, and hence their earnings, by increasing the quantity of outputs and increasing the quality of services as measured by an elaborate instrument measuring quality across 13 services and 185 variables. The national PBF model for use in district hospitals consists of 52 indicators, 15 of which assess administration, 19 of which assess quality assurance, and 18 of which assess clinical activities. At the hospital level, performance is assessed through a peerevaluation mechanism, and available performance budgets range between US$70,000 and US$175,000 per hospital per year. Finally, the community PBF is based on experience gained during implementation of the health center and hospital PBF. It consists of decentralized control and decision making by a sector PBF steering committee and payment of a community health worker cooperative after performance was assessed. PBF grants are provided to the health sector in the form of block grants from the government and donors, most of which are used to supplement health worker salaries based on performance. Specifically, the contract includes indicators related to adequate coverage and has reportedly led to an increase in utilization. At the community level, about US$0.25 per capita is directly transferred to municipalities to engage, via a performancebased contract, communitybased institutions, NGOs, health promoters, private health care providers, and other related services to deliver essential interventions at the household and community levels. At the health center level, the transfer amounts to about US$1 per capita through a performancebased contract. Fiscal decentralization: Rwanda's governmentwide decentralization policy (2000) serves as an integral part of the government's national development strategy, as expressed in Vision 2020, the National Poverty Reduction Strategy Program, and the National Program for Strengthening Good Governance for Poverty Reduction in Rwanda. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 157 Fiscal decentralization (policy adopted 2001, law enacted 2006) serves as an essential component of Rwanda's decentralization agenda. It involves the decentralization or transfer of fiscal responsibilities and financial resources to decentralized units, across sectors. As stipulated in the 2007 budget, it specifically involved the following: (a) block grants to districts amounting to 5 percent (formerly 3 percent) of central government revenues per year; (b) earmarked transfers to districts for--among other things--health worker salaries and health centers; and (c) other transfers to districts including transfers, under MINALOC, for the Common Development Fund (CDF), a coordinated funding mechanism for development initiatives at local levels that allows for the mobilization and disbursement of community development resources without the complexities and bureaucracy from government and donors, and development budget transfers (mainly donor funds) under the relevant ministries. Fiscal and financial decentralization have resulted in relative autonomy in budgeting and financial management at local levels, which has facilitated prioritization of expenditure needs in situations of inadequate financing. In terms of sectoral decentralization, the key objective was to devolve responsibilities to areas where beneficiaries can best access them, whereby local government assumes responsibility for delivering services. In health, the district is viewed as the basic operational unit and cornerstone of the health system, responsible for: the organization of health services, administrative functions/logistics (that is, resource management), and supervision of community health workers. The health sector budgeting was fully decentralized, and local activities are to be aligned with the newly initiated health SWAP. Health Sector Public Expenditure Allocation There was an incremental change since 2002 in the health budget, both as a percentage of GDP, reaching 3 percent in 2007, and as a percentage of the national budget, reaching 9 percent in 2007 (see table 5.64). However, the share of domestic spending in health (recurrent) remained almost constant over the period, around 6 percent of total government expenditures. In fact, domestic spending in health increased only mechanically because total government expenditures increased. Consequently, despite the government's commitment to increase the allocation to priority sectors (including health), Rwanda is very far from the Abuja target, according to which 15 percent of public expenditure should be allocated to health. Since the 2005 fiscal decentralization process, funds being transferred to districts have increased, and in 2006, the transfers category was reaching 447 percent over the previous year. On average, transfers accounted for the largest share of public expenditure in 2007 at 39 percent, followed by goods and services at 35 percent, wages and salaries at 23 percent, and capital expenditure at 4 percent (see table 5.65). 158 World Bank Working Paper Table 5.64. Health Expenditure to National Expenditures (2002­07) 2002 2003 2004 2005 2006 2007 GDP (In RwF millions) GDP Nominal 781,500 955,200 1,137,900 1,327,100 1,583,000 1,826,200 GDP Real 1,490,577 1,494,784 1,573,404 1,684,945 1,776,836 1,826,200 National Budget (In RwF millions) National Recurrent Budget 263,289 251,225 288,972 305,058 316,014 413,516 National Development Budget 183,355 101,898 112,591 153,131 170,265 210,155 National Budget Total 446,645 353,123 401,564 458,189 486,279 623,671 Health Budget (In RwF millions) Recurrent Health Budget (including Districts) 11,978 15,958 16,224 18,179 20,256 23,627 Development Health Budget 18,619 4,872 9,052 19,238 40,783 31,539 Health Budget Total 30,597 20,829 25,276 37,417 61,040 55,166 Ratios (%) Health Budget as % of GDP 2.10% 1.40% 1.60% 2.20% 3.40% 3.00% Health Budget as % of National Budget (Recurrent) 4.50% 6.40% 5.60% 6.00% 6.40% 6.10% Health Budget as % of National Budget (Development) 10.20% 4.80% 8.00% 12.60% 24.00% 15.00% Health Budget as % of National Budget (Total) 6.90% 5.90% 6.30% 8.20% 12.60% 9.10% Annual Growth Rate of the Health Budget 0.30% 5.30% 7.10% 5.50% 2.80% Source: Health PER Report, 2008. Table 5.65. Public Health Expenditure by Economic Classification (2002­07) In Million RwF 2002 2003 2004 2005 2006 2007 Wages and Salaries 2,155 3,463 5,159 4,446 2,951 5,764 As a % of total expenditure 21% 30% 39% 30% 15% 23% Goods and Services 5,472 3,419 4,596 6,398 8,962 8,670 As a % of total expenditure 54% 30% 35% 43% 45% 35% Transfers 1,013 3,293 1,420 1,305 7,142 9,690 As a % of total expenditure 10% 29% 11% 9% 36% 39% Domestic Capital 164 17 604 1,573 840 968 As a % of total expenditure 2% 0% 5% 11% 4% 4% Others 1,255 1,174 1,341 1,157 -- -- As a % of total expenditure 12% 10% 10% 8% -- -- TOTAL 10,059 11,366 13,120 14,879 19,895 25,092 Source: Health PER Report, 2008. Between 2003 and 2006, total health expenditure per capita doubled from US$17 to US$34 (see figure 5.3). Total health expenditures come from four major financing sources: (a) government revenues, which include taxes, loans, grants, and donor contributions through GBS; (b) donor funds, partially onbudget as seen in the development budget, and partially earmarked offbudget donor funds; (c) internally generated funds from health facilities; and (d) private expenditures, mainly through household outofpocket expenditures. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 159 Figure 5.3. Per Capita Health Expenditure (1998­2006) in 2006 constant US$ Prices Source: NHA 2006. Government revenues. Public finance for health increased over time, from US$3.2 per capita in 2003 to US$12.8 per capita in 2007 (see table 5.66). The share of domestic resources to public health expenditure decreased while external health resources increased. In fact domestic resources allocated to health increased--mainly because of district transfers--but at a less rapid pace than external onbudget financing. In 2007, 55.6 percent of public resources to health came from external sources and 44.4 percent from domestic sources. Table 5.66. Sources of Public Finance for Health (2003­07) 2003 2004 2005 2006 2007 In RwF Millions Domestic ordinary MINISANTE budget 11,504 10,712 12,282 11,988 14,925 Domestic district transfers for health 4,454 5,511 5,895 6,057 8,700 Domestic (CP and FI) 303 1,767 1,055 884 External (grants and loans) 21,528 8,748 17,470 35,278 30,655 Total public health resources (on-budget) 37,486 25,276 37,416 54,380 55,165 Share of Public Health Expenditure Domestic MINISANTE resources to public health expenditure (%) 43% 65% 53% 35% 44% External health resources to total public health expenditure (%) 57% 35% 47% 65% 55% Resources per capita In RwF 1,713 2,258 3,507 6,909 7,006 In US$ 3.2 3.9 6.29 12.52 12.8 Source: Health PER Report, 2008. 160 World Bank Working Paper External financing. Donors' share of total health expenditures (THE) increased to more than 53 percent in 2006, compared with 41 percent in 2003 (see figure 5.4). This increase is impressive but distorted. A big share of external resources (42 percent) was earmarked for HIV/AIDS activities in 2006, amounting to US$7.4 per capita (out of US$17.7 in total external health resources). If HIV/AIDS financing were excluded, THE in 2006 would reach only US$25.6 per capita, which would not be a dramatic increase in per capita spending from the 2003 level. Internally generated funds from health facilities. Health facilities generate revenue internally through the sale of drugs and consumables, and other services such as consultation, hospitalization, surgery, and laboratory services. This is supplemental income on top of government subsidies, subsidies from local district authorities, and NGOs. The income generated by health centers increased by 660 percent over 2002­07. At the level of district hospitals, it increased by 305 percent, while that of reference hospitals grew by 194 percent (2008 PER). The year 2006 saw a dramatic change in fortunes when revenue generated by health centers shot up by 234 percent. This significant shift in capacity to generate revenue is a result of the mutuelles' national scaleup. Private expenditure. The share of private spending in THE increased to 28 percent in 2006 compared with 25 percent in 2003. In 2006, the share of private spending in THE was even greater than that of public spending on health (see figure 5.4). Private expenditures amounted to US$9.4 in 2006, out of which US$7.5 constituted outofpocket expenditures. Figure 5.4. Contributions to the Health Sector as a Percentage of THE (2006) Source: NHA 2006. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 161 Outofpocket spending amounted to almost US$3 per capita. The biggest share of outofpocket spending--US$1.8--occurred at private pharmacies and shops, as well as for private health care providers (traditional healers, private referral hospitals, and private clinics) (see figure 5.5). When compared to other outofpocket spending, mutuellerelated expenditures were relatively low: overall mutuelle financing amounted to US$2 per capita, out of which US$1.4 came from outofpocket spending. Figure 5.5. Out-of-Pocket Spending by Provider Type Public referral Dispensing hospital chemist Other Private referral 1% 15% 39% hospital 4% District government hospital 6% District agree hospital 3% Public health Traditional Private center healer clinic 12% 11% 9% Source: NHA 2006. The following paragraphs look at resource allocation and management within the health sector based on the NHA 2006 findings. They show that the allocation of health resources varies a lot across activities. HIV/AIDS is the major activity financed within the health sector. HIV/AIDS resources represent almost one quarter of THE, while malaria gets 14 percent of THE and reproductive health only 6 percent (see figure 5.6). The shares of malaria and reproductive health in THE decreased after 2003, while HIV/AIDS financing increased. Figure 5.6. Health Expenditure by Activity Malaria 14% Other remaining HIV/AIDS health 56% 24% Reproductive Health 6% Source: NHA 2006. 162 World Bank Working Paper This is mainly because of the striking amount of resources earmarked by some donors (Global Funds, United States President's Emergency Plan for AIDS Relief [PEPFAR]) for HIV/AIDS activities. Priorities in health spending vary across financing agents. Donors were almost the only financing source for HIV/AIDS activities, while public and private resources focused more on health services that are prioritized less by donors. Donor financing to HIV/AIDS activities increased over the years, while their reproductive health financing decreased from 37 percent to 8 percent. Private spending priority was different from that of donors, where almost onequarter of total private spending (US$2.1) was allocated to malaria (figure 5.7). Figure 5.7. Contribution to HIV, Malaria, Reproductive Health, and Other Health Activities by Financing Source 100 90 80 36% 70 Other remaining health 60 75% 8% 88% Reproductive Percent 50 Health 40 HIV/AIDS 42% 30 2% Malaria 20 10 5% 23% 3% 13% 0 4% Public (includes Private Donor parastatals) Source: NHA 2006. The functional breakdown of health funds shows that curative care had the biggest share of THE, with 26 percent of THE going toward outpatient curative care and 13.2 percent toward impatient curative care. MINISANTE managed slightly more than onequarter of THE. Development partners and households are the two main other managers of health resources. Although efforts toward the integration of donor funds into the MINISANTE budget were made, 27 percent of THE coming from external resources was managed off budget (see figure 5.8). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 163 Figure 5.8. Management of Health Sector Resources Private company direct transfer to Other, 8.1% health facilities, 1% MINISANTE, 28% Implementing agencies (incl direct donor transfers), 27% Districts, 0.04% Other min, 4% Private insurance (mutuelles), 5% Parastatals, 0.5% RAMA, 3% OOP, 23% CSR, 0.01% FARG, 1% Source: NHA 2006. Health Sector Results Most health indicators have shown clear improvement since the genocide and war of 1994, with the exception of fertility and malnutrition rate among children under five. The major progress achieved since the end of the 1990s has set the country on a positive trajectory, which provides hope that Rwanda could achieve the MDGs related to child health by 2015. In the early 1990s, Rwanda's infant and underfive mortality rates were 103 per 1,000 and 173 per 1,000, respectively. The underfive mortality rate had climbed dramatically by the end of the 1990s as a result of the aftereffects of the war. Beginning in 1998, however, there was a substantial decline in underfive mortality in the country, and from 2000 to 2005, the infant and underfive mortality rates fell to the levels achieved between 1988 and 1992. Demographic and Health Survey (DHS) estimates for 2007 suggested that the infant mortality and underfive mortality rates continued to drop, falling to 62 per 1,000 and 103 per 1,000, respectively, showing that the country is on track to reach MDG4 targets on time (see table 5.67). Table 5.67. Rwanda's Child Mortality Rates (in 1000s) Proposal to address gap MDG 2015 1990 1995 2000 2005 2007 2007­15 goal Under-Five Mortality Rate 173 209 203 152 103 50 53 Infant Mortality Rate 103 124 118 86 62 34 28 Sources: UN's Statistic Division, MDG Indicators; DHS 2005 and 2007; EDPRS, page 32. 164 World Bank Working Paper Although decreasing, the maternal mortality ratio is still very high in Rwanda, with 750 maternal deaths per 100,000 live births in 2005 (DHS 2005). The maternal mortality ratio in Rwanda is a real source of concern, and significant efforts are still needed if the country is to achieve the MDG5 target by 2015 (see table 5.68). Table 5.68. Rwanda's Maternal Mortality Ratio per 100,000 Live Births Proposal to address MDG 2015 1990 1995 2000 2005 2007 gap 2005­15 goal Maternal Mortality Ratio 1,300 2,300 1,071 750 n.a. 482 268 Sources: UN's Statistic Division, MDG Indicators; DHS 2005; EDPRS, page 32. As far as MDG 6 is concerned, there were marked improvements. Significant progress was made in the area of HIV/AIDS, and its prevalence was reduced from about 9 percent in 2000 to 3 percent in 2006. Although malaria is still the first cause of morbidity and mortality in Rwanda, the country made impressive progress in the prevention and treatment of malaria, resulting in a decrease in malaria mortality in health centers while use of health services increases. The situation regarding tuberculosis is more worrying as the incidence of TB steadily increased from 135 per 100,000 in 1990 to 562 per 100,000 in 2006. Rwanda's incidence of TB remains high in comparison with the SubSaharan Africa average. The coverage of essential health interventions was extended as a result of the combination of the increased technical capacities of health centers through the decentralization of health services, the extension of population and communitybased services, and changes in the interactions of the population with health centers resulting from the extension of mutuelle coverage. Reproductive health services experienced a significant increase during the decade. The coverage of family planning services, antenatal care services, and professional assistance at delivery significantly increased, reaching 27.4 percent, 95.8 percent, and 52.1 percent, respectively, in 2007. Sustained progress was also made in child immunization (90.4 percent coverage of children ages 12­23 months for vaccination against measles in 2007). Despite numerous nutrition interventions over the decade, however, problems persist for the reduction of malnutrition at the population, health facility, and strategic levels. Malnutrition among children under five remains relatively high in Rwanda: in 2005, more than one in every five children (22.5 percent) suffered from moderate or severe malnutrition, and 4.4 percent of children suffered from severe malnutrition, based on the weightforage indicator. The paragraphs below summarize the impact of the main health strategies implemented over the past years on households, health outcomes, use of health services, and equity. Mutuelles: Enrollment in mutuelles increased significantly between 2003 (7 percent) and 2007 (75 percent); thus doubling the use of health care services and the use of public modern services. In 2005 among individuals indicating that they had been ill during the two weeks preceding the EICV II survey, 52 percent of those who were members of the Rwanda Health Insurance Scheme (RAMA)33, 34 percent of those who were members of a health mutuelle, and 21 percent of those with no health insurance Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 165 coverage used modern health services (see figure 5.9). In contrast to individuals with health insurance coverage, those with no such coverage showed a greater tendency to use the services of traditional practitioners. Figure 5.9. Use of Health Services in Case of Illness in the Overall Population 17 9 13 20 9 11 11 35 25 16 10 No Health Insurance RAMA Health mutuelle All Public services Private services Traditional practitioners Source: EICV II. Health insurance coverage greatly reduces the amount of direct illnessrelated spending. After adjusting for inflation, patients with no insurance coverage in 2005 had illnessrelated expenses comparable to the 2000 level, and direct illnessrelated spending followed the same gradient with household income levels. On the contrary, the amount of direct illnessrelated spending in 2005 decreased among beneficiaries of health insurance schemes (see figure 5.10). Figure 5.10. Direct Disease-Related Spending by Income Quintile (2000­05) 1200 Median (RwF) 1000 800 600 400 200 0 Poorest Quintile 2 Quintile 3 Quintile 4 Richest All EICV 2000 EICV 2005 Uninsured EICV 2005 Insured Source: EICV I and II. 166 World Bank Working Paper Health mutuelle members had direct expenses nearly three times lower than uninsured individuals: members of health mutuelles spent an average of RwF 973, while for individuals with no coverage the average spending came to RwF 2,758. Direct spending from individuals with no coverage, compared with health mutuelle members, was three times higher for drugs, 3.5 times higher for hospitalization, and 2.7 times higher for consultations. Finally, further analyses on the impact of mutuelles were conducted based on the EICV II data and using bivariate methods and regression methods. These analyses aimed at addressing the following questions: First, is enrollment in health mutuelles affordable for Rwandese households in general and the poorest households in particular? What is the incidence of prepayments into health mutuelles on household expenditures? Are health mutuelle schemes socially inclusive? Second, does enrollment in health mutuelle schemes improve access to health care? How are different socioeconomic segments of the population affected by changes in access and use of health care services? Third, does enrollment in health mutuelle schemes provide coverage against financial risks associated with illness in Rwanda? Does enrollment in health mutuelles result in significant reduction in outofpocket expenditures related to illness? The main results of these analyses are summarized in box 5.1. Box 5.1. Results of Analyses on the Impact of Mutuelles on Households Affordability of enrollment in health mutuelles The premium for health mutuelles (RwF 1,000) is about 1 percent of the average annual household expenditures per capita and 2 percent of average annual household nonfood expenditures in the general population. However, the ratio of expected premium contributions to household nonfood expenditures declines steadily from 17 percent among the poorest decile to less than 1 percent among the richest decile. These results suggest that current levels of premium contributions are affordable for the majority of Rwandese, except the poorest 10 percent of households. Social inclusion in health mutuelle schemes Results show that membership in health mutuelles rises with family size. Male-headed households have a higher proportion of enrollment than female-headed households, partly because of the income difference between the two groups. It also appears that heads who completed primary school or have gone through some vocational training tend to have the highest rate of enrollment among the least educated and best educated ones. This trend is also reflected in the enrollment rate by income category. It is the middle income and middle rich who tend to have the highest participation rate in health mutuelles. Household and community characteristics driving demand for health care services Regression results suggest that demand for health care services is driven by factors such as enrollment in mutuelles, income, and distance to the nearest health care facility, among other things. One interesting result is that the effect of a mutuelle on use of modern health care providers is stronger among the poorest 40 percent of the population than among higher-income categories. However, the poorest quintile who are beneficiaries are less likely to seek care than beneficiaries of mutuelles in higher-income groups, which suggests that the poorest quintile continues to face significant barriers to seeking care in the modern health sector, even with insurance coverage. Household income protection in the health sector Households not covered by a health mutuelle spent nearly twice as much for illness-related services than insured households during the EICV II survey period. Therefore, from the point of view of protecting against unexpected illness-related expenditure, mutuelles seem to have worked quite well in Rwanda. Source: Laurence Lannes, Abebe Shimeles, François Diop. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 167 Performancebased financing: In the national health center PBF model, payments for performance were based on the quantity of outputs achieved (through casebased remuneration) conditional on the quality of services delivered. The outputs were measured monthly, while the quality is measured quarterly through the use of an elaborate supervisory checklist. Overall, the implementation of PBF is expected to translate into increases in the volume of services, better quality services, an increase in staff productivity, and greater provider enthusiasm and motivation. To assess the actual effect of PBF, an impact evaluation financed by the World Bank was conducted in 2006­08. It aimed at isolating the impact of the PBF incentives on the increase in quantity and quality of health services delivered in public facilities in Rwanda, on the improvement of the health status of the population, on the motivation and behavior of health care personnel, and on the allocation of resources received by health facilities. The main results of this evaluation are summarized in box 5.2. Box 5.2. Results of the PBF Impact Evaluation Objective and methodology The objective of the impact evaluation was to measure the health situation in Rwanda before the start of PBF and after two years of implementation in both treatment (districts with PBF) and control areas (districts without PBF). Data for the baseline survey was collected before the start of the PBF in 2006, and data for the follow-up survey was collected in 2008 to measure the health situation after a 24-month exposure period, before the roll-out of PBF in the control areas. To isolate the incentive effect from the resource effect, the comparison facilities' traditional budgets were increased an amount equal to the average PBF payments to the treatment facilities. Results The impact evaluation provides evidence that PBF is significantly associated with increased use and quality of a number of high-impact maternal and child health care services, including quality of prenatal care, institutional delivery, and child preventive care utilization. In particular, the evaluation shows that PBF has a large and significant impact on increasing institutional deliveries and preventive care visits by young children, as well as improving the quality of prenatal care, as measured by process indicators of the clinical content of care and tetanus toxoid vaccination. However, PBF seems to have no effect on the number of women completing four prenatal care visits or on the number of children fully immunized. Discussion By conditioning the PBF payment on a quality index score, the evidence suggests that the PBF gives providers the incentive to translate their knowledge into better practice. While the PBF incentives have strong effects on the use of certain services, they have little effect on others. This is probably related to the structure of the incentives. Source: Laurence Lannes, Paulin Basinga. Community health workers: Since 1995, the Community Health Unit of MINISANTE has trained about 12,000 community health workers who are living and working among their communities. As a result of decentralization, it is estimated that 60,000 community health workers in all are needed to serve all districts. The positive impact of community health workers is already being felt. Based on mobilization efforts by community health workers, Rwanda's Expanded Program on Immunization has achieved some of the best immunization levels in Africa. Family 168 World Bank Working Paper planning and reproductive health services are affordable and accessible to the majority of Rwandese. A nationwide community nutrition surveillance program was put in place and is reaping good results. Other examples include the engagement of community health workers in malaria prevention, with significant results--proving that community volunteers, when effectively used, can bring about positive results such as the increase in the use of insecticidetreated bed nets from 4 percent to more than 70 percent of the population from 2004 to 2007. Until 2008, however, the community health workers had not been formalized within the system, although the government acknowledged the need for a more formal organization and mobilization at the community level to extend the current success in other areas, many of which require more technical skill (and therefore more training). In 2008, MINISANTE developed a Community Health Policy to guide and strengthen the provision of community health services; it was then recognized that community health workers have the advantage of knowing their community well in terms of the community culture, norms, beliefs, traditions, formal and informal networks, support systems, community strengths, and, most likely, their communities' health problems. The implementation of the above presented health strategies seems to encourage equity and to improve financial access to health care services. According to the EICV II survey, in 2005, only 10 percent of respondents identified the cost of health services as the main obstacle to receiving health care (see table 5.69). Table 5.69. Reasons for Not Visiting a Government Service Facility since Last Year Reasons Valid Percent Nobody has fallen sick during this period 59.6 % Lack of drugs at facility 8.1% Impolite staff 0.3% Long distance to facility 9.4% Cannot afford payment 10.1% Poor quality services 1.7% Do not provide specialized treatment (i.e., hypertension) 3.7% Lack of cleanliness 0.7% Long waiting time 0.3% Do not know 0.7% Other 5.4% TOTAL 100% Source: EICV II. However, although health mutuelles and other strategies implemented in Rwanda contributed to remove financial barriers to access health services, inequity remains a major source of concern. The results of a benefit incidence analysis (BIA) conducted with 2005 data (DHS, EICV, and PER) to measure the benefits of public expenditure across characteristics of individuals such as income, provinces, or gender reveals, in most instances, the regressive nature of public subsidy in the health sector in Rwanda. The results of this BIA are summarized in box 5.3. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 169 Box 5.3. Main Findings of the Benefit Incidence Analysis in the Health Sector in Rwanda (2005) In 2005, most public subsidies (78.8 percent) were flowing to the four referral hospitals (mainly CHUK and King Faycal Hospital in the capital city). The remainder was almost equally distributed among district hospitals (10.1 percent) and health centers (11.2 percent). Given the low utilization of health services provided in referral hospitals compared with the high utilization of services in district hospitals and health centers, one can wonder whether this is an efficient way of spending scarce public resources. Furthermore, as most of the burden of disease in Rwanda can be treated at a lower cost at the health center level, providing more public subsidy at this level and less at the referral hospital level would have a significant impact on the health status of the population. In Rwanda, the public subsidy system is regressive at all levels, both for outpatient and inpatient care. The richest are the main beneficiaries of the system, and their benefit increases with the level of care. Overall, the richest quintile benefits from 49 percent of the total public subsidy, while the poorest receives less than 7 percent of the total subsidy. The public subsidy received by the poorest quintile is mainly flowing to the health center level, which may be related to both the needs of the poor and to their financial capacities. On the contrary, richer quintiles are more likely to seek care at the hospital level (where the cost of services is higher) than in a health center. The public subsidy unit cost for outpatient care varies from about RwF 115 at the health center level to RwF 6,502 for the referral hospital level, whereas the number of annual outpatient visits per capita is 0.55 at the health center level and only 0.02 at the referral hospital level. As far as inpatient care is concerned, the unit cost of the subsidy varies from about RwF 319 to RwF 138,693. In sum, it is clear that the public subsidy is highly distorted to the benefit of referral hospitals, which are frequented mostly by the rich urban population. None of the public subsidies for specific high-impact interventions shows a tendency to target the poor. Family planning subsidies appear to be highly regressive as the richest quintile benefits from 39 percent of the total subsidy, whereas the poorest quintile receives only 12 percent of the subsidy. Public subsidy for other health interventions (diarrhea- related interventions, antenatal care, immunization services) is relatively well balanced among the quintiles, although not progressive. In sum, the share of public subsidies targeted toward primary health care services needs to be revised to cater for the major morbidity causes. Public subsidies should be targeting the poor and women instead of benefiting those who use the most health care services. Different ways of targeting the poor and women and of setting up cross-subsidies exist and should be examined carefully to see which ones would be the more appropriate in Rwanda. Source: Laurence Lannes, Sabine Musange. 170 World Bank Working Paper Social Protection Social Protection Sector Overview In 2006, an international consultant34 conducted a detailed analysis of the social protection sector covering 2004­07 (Social Protection PER 2006) with actual data available through the year 2005. The report brought into light the reality of the social protection sector in Rwanda. In short, key conclusions of the social protection (SP) report can be summarized as follows: ONE OF THE LARGEST SECTORS: SP spent RwF 59 billion per year (US$110 million), excluding NGOs, making it one of the largest sectors of the EDPRS (larger than total spending on health and education). SCATTERED INTERVENTIONS: Donor funds were scattered over more than 80 different projects that each received less than US$300,000. (Only three projects got more than US$5 million.) Also, government funds were scattered over a large number of tiny interventions with limited strategic orientation. POOR AID POLICY ALIGNMENT: Seventyfive percent of funds came from offbudget donor projects and were not aligned with national policies and priorities. Therefore, SP ought to be reformed under the aid policy implementation. POVERTY IMPACT LIMITED: Poverty impact was unknown, at best, as no systematic monitoring of poverty impact or coordination between interventions existed. Only a limited number of people were reached by SP programs, ranging from a few hundred households to about 100,000 individuals through Victims of Genocide Fund (FARG).35 INEFFICIECY: Overheads and duplication between donor projects were large, and a majority of funds went to food aid for recurrent food crises. The main recommendation of the PER report was to balance donor and government funds and priorities to find more efficient, targeted, and sustained ways to provide SP to a greater portion of the population. In this analysis, SP was defined according to three elements: 1. Improvement of social status and protection of the rights of the vulnerable (including legal and regulatory protection) 2. Protection against welfare risks through pooling and social insurance functions 3. Transfers of income or consumption to the poor, vulnerable, and marginalized When classifying SP spending, the PER report considered the elements shown in table 5.70. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 171 Table 5.70. Summary of Main Judgments Used in Defining Social Protection36 EXPENDITURES CLASSIFIED AS SOCIAL PROTECTION Recurrent Budget Government's Social contributions to Employee Pensions, Social & Health Insurance Programs Public Subsidies to Mutuelles and Health Referrals and Emergencies Disaster Planning, Management, Relief. Support to Refugees & Returnees District Support to Women's Self-Development Support to Youth Projects MINEDUC Expenditure on Special Education Budget Expenditure on Genocide Survivors (incl. non-budgetary contributions from firms and indiv.) Protections of Rights of Poor & Vulnerable Central and District Expenditures Classified as Benefiting Vulnerable Groups, OVCs and destitute* Spending Classified as Child Protection General Expenditures Aimed at Promoting Gender Equality and Rights of Women Development Budget Labor Intensive Public Works (HIMO) Ubudehe Program Re-settlement Programs Poverty Reduction and Actions in Favor of Women Micro-Credits for Self-Support of HIV/AIDS Affected Protection de la Prevoyance Sociale Refugee Commission Donor Projects Mutuelles Component of GFATM System Strengthening Projects Food Aid for Emergencies or Vulnerable Groups Feeding School Feeding Re-settlement, Reintegration of Refugees, Returnees, and Excombatants Justice and Rights Protection for Vulnerable Groups Support to Disabled Economic, Social and Nutrition Support to Households Affected by HIV/AIDS Labor Intensive Cash or Food-for-Work Projects Ubudehe Capacity Building and Policy Support Related to the Above Projects Targeted to Specific Vulnerable Groups Source: Social Protection, PER Report, 2006. Table 5.71 shows the expenditures that are not classified as SP spending under the above definition. Table 5.71. Summary of Main Judgments Not Used in Defining Social Protection37 EXPENDITURES NOT CLASSIFIED AS SOCIAL PROTECTION Recurrent Budget Other Health and Education Expenditures Directly Met from Government's Budget Community-based Health and Other Programs Functional Literacy Programs for Youth and Adults Student Scholarships and Finance Agency for Higher Education Micro Credit unless included in other programs Province-level Spending (may incl. management and/or supervision function but not clear by data provided) Development Budget Common Development Fund (CDF), other than Labor Intensive Public Works (HIMO) and Ubudehe Donor Projects Projects Not Found in Development Assistance Database (DAD) database Food-Aid Used for General Economic Support Health Expenditures, incl. Costs of Antiretroviral Treatment NGO Projects Except Where Identified as Recipients of External Development Partners Assistance Budget Support (as reflected in government's figures) Source: Social Protection, PER Report, 2006. 172 World Bank Working Paper Based on the SP definition and the classification of SPrelated expenditures as outlined in table 5.70, the PER report identified the following overall SP trends between 2004 and 2007 (all expenditures are shown in real 2007 prices unless otherwise specified): SP spending increased from 5.7 percent of government spending in 2004 to 6.5 percent in 2005. Total government plus donor spending increased to more than US$10/person, representing about 4 percent of GDP. Donors finance twothirds of total spending on SP in 2004, as table 5.72 illustrates: Table 5.72. Total Social Protection Spending Overview 2004 2004 2005 2005 2006 2007 In million RwF Budget Actual Budget Actual Budget Budget Total Budgeted 19,019 15,653 19,109 21,772 20,479 24,621 Extra-Budget Donor Exp. -- 30,573 -- 27,027 38,803 -- TOTAL SP EXPEND: 19,019 46,226 19,109 48,799 59,282 24,621 ANALYSIS: SP as % of Total Gov. Spend. 6.9% 5.7% 5.7% 6.5% 5.3% 5.0% SP as % of GDP -- 4.4% -- 4.0% 4.4% -- Real Total Budget Spending* 21,781 17,874 19,873 22,643 20,479 22,818 Real Total SP Spending* -- 52,786 -- 50,751 59,282 -- Total Spend per capita US$ -- $8.60 -- $9.10 $10.50 -- Memo: Total Donor Spend. -- 31,448 -- 32,565 40,043 -- Source: Social Protection, PER Report, 2006, Executive Summary page 5. * 2006 prices as original table. Contributing ministries and their contributions to SP spending detailed in table 5.73. Table 5.73. Distribution of Recurrent Social Protection Expenditure 2004 2004 2005 2005 2006 2007 in % Shares Budget Actual Budget Actual Budget Budget DISTRICT TOTAL 12% 12% 12% 11% 13% 11% Primature (incl. MIGEPROF) 4% 2% 6% 4% 3% 5% MINEDUC 2% 1% 2% 1% 3% 2% MINISANTE 3% 2% 4% 4% 5% 3% MINALOC 80% 83% 77% 80% 78% 79% Central Government Total 88% 88% 89% 89% 88% 89% TOTAL ALL 100% 100% 100% 100% 100% 100% Source: Social Protection, PER Report, 2006, Executive Summary page 6. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 173 Of these ministries, MINALOC programs represented about 80 percent of recurrent SP expenditure, including: FARG support to genocide survivors Ubudehe Approach (funded by EU) HIMO laborintensive works program The districts' share of government recurrent expenditure, down from 12.3 percent in 2004 to only 11 percent budgeted in 2007, includes provisions for support of the following: Orphans and vulnerable children (OVC) Community health insurance Support for the indigent Mobilizing donor and local funding Table 5.74 shows how SP spending was allocated to various programs. Table 5.74. Overview of Allocation of Total Social Protection Funding in % Shares 2004 Actual 2005 Actual 2006 Estimate Contributions for Public Employees 9.0% 9.5% 10.1% Other Direct Support to Households 86.5% 84.2% 86.7% - Genocide Survivors 17.9% 17.3% 16.2% - OVCs 18.3% 15.7% 16.3% - Labor intense pub. works 10.0% 2.1% 1.8% - Women 2.0% 3.1% 1.9% - Health Costs (mutuelles ) 0.3% 0.7% 9.1% -Refugees & Returnees 5.6% 4.0% 4.4% -HIV/AIDS Affected 9.2% 8.7% 7.9% -Ex Combatants 1.4% 11.8% 0.8% - Disabled 0.1% 0.1% 0.3% -Ubudehe 0.0% 8.4% 11.0% Policy, Rights, Advocacy & Management 4.5% 6.3% 3.2% TOTAL 100.0% 100.0% 100.0% Source: Social Protection, PER Report, 2006, page 7. These key conclusions and findings are based on a detailed analysis of the ministries, donors, programs, and projects involved in the SP sector. Below follows a detailed summary of the expenditures in the sector as outlined in the PER report. Social Protection Sector Public Expenditure Allocation Review The following overall spending trends are illustrated in table 5.75: 174 World Bank Working Paper Table 5.75. Actual and Budgeted Total Social Protection Expenditures 2004­07 Expenditures 2004 2004 2005 2005 2006 2007 (in million RwF) Budget Actual Budget Actual Budget Budget Recurrent Budget (RT) 14,893 14,262 16,426 15,362 18,412 21,201 Development Budget (DB) 4,126 1,391 2,683 6,411 2,067 3,420 Total RT & DB Budgeted 19,019 15,653 19,109 21,772 20,479 24,621 Extra Budget Donor Expenditure 0 30,573 0 27,027 38,803 0 Total SP Expenditure 0 46,226 0 48,799 59,282 0 ANALYSIS Percentages: SP budget as % total Government 6.9% 5.7% 5.7% 6.5% 5.3% 5.0% budget SP recurrent as % Government 8.8% 8.4% 7.3% 6.8% 8.5% 6.5% recurrent SP total as % of GDP 0.0% 4.4% 0.0% 4.0% 4.4% 0.0% Million RwF (REAL): Real RECURRENT (RT) SP 17,007 16,286 17,083 15,976 18,412 19,649 spending 2006 prices Real DEVELOPMENT (DB) SP 4,712 1,588 2,790 6,667 2,067 3,170 spending 2006 prices Real TOTAL RT & DB SP 21,718 17,874 19,873 22,643 20,479 22,818 spending 2006 prices Real extra-budget donor SP 0 34,912 0 28,108 38,803 0 spending 2006 prices Real TOTAL (incl. extra budget) 0 52,786 0 50,751 59,282 0 SP spending 2006 prices Total spend per capita RwF 0 4,918 0 5,083 5,988 5,004 Total spend per capita US$ 0 9 0 9 11 0 Memo: Total donor spending 0 31,448 0 32,565 40,043 0 Source: Social Protection, PER Report, 2006, page 4. *In 2006 real prices (as in original table). Actual government and donor expenditure on SP increased from RwF 46 billion in 2004 to RwF 49 billion in 2005. This was equivalent to US$79 million and US$87 million, respectively, and represented 4 percent of GDP. 2006 estimated SP expenditure reached RwF 59 billion (more than US$100 million). The data presented in table 5.75 show significantly higher SP expenditure than previously reported in the interim report. FARG38 spending in 2004 increased by incorporating · RwF 983 million carried over from the 2003 budget · RwF 506 million of compulsory contributions from individuals and firms in 2004 · RwF 560 million of compulsory contributions from individuals and firms in 2005 Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 175 SP spending that is included in the government's actual budget increased from 5.7 percent of actual total government spending in 2004 to 6.5 percent in 2005. While actual SP spending as a percentage of the total government budget increased from 5.7 percent to 6.5 percent between 2004 and 2005, budgeted SP spending as a percentage of the total government budget went down from 6.9 percent to 5.7 percent in the same period. However, this may be misleading (especially in the development budget) as the budget does not always include projects until donor financing intentions are clear.39 Table 5.75 outlines total actual and budgeted expenditures in both nominal and real terms for 2004­07 except for 2006 and 2007, when only budgeted data are listed. Rights and advocacy work related to social equity and rights of women, children, and vulnerable groups are included in the data. The increased spending in these programs compared with previous SP estimates are not only the result of the inclusion of these programs, but also the fact that coverage of development spending by donors increased and that existing interventions previously had been missed or excluded. The following SP recurrent spending trends were identified: Recurrent spending on SP, more stable than development spending, increased in constant prices40 by 14.5 percent between 2004 and 2007, or 4.6 percent per year. While recurrent spending on SP increased in the 2004­07 period, the overall recurrent budget increased at a higher rate, resulting in a decline of SP's share of the overall budget from an 8.4 percent in 2004 (actual) to 6.5 percent in 2007 (budgeted). Actual recurrent spending on SP fell in real terms from 2004 to 2005. Including FARG funding carried over from 2003, recurrent spending on SP equaled 96 percent of budged funds in 2004 and 94 percent in 2005. Table 5.76 shows the SP recurrent spending by the central government and the districts in 2004­07. Following are some highlights from the table: MINALOC dominated spending on SP, accounting for about 79 percent of budgeted total recurrent SP spending. Spending on SP by districts declined between 2004 and 2007. The districts' share of budgeted SP spending declined from 12.3 percent in 2004 to 11.5 percent in 2007, and of actual SP spending from 12.4 percent in 2004 to 10.7 percent in 2007. The share of total recurrent SP spending by the other three ministries contributing to expenditures in this sector--PRIMATURE, MINISANTE, and MINEDUC--fluctuated between 2 percent and 6 percent, 2 percent and 5 percent, and 1 percent and 3 percent, respectively, over the 2004­07 period. 176 World Bank Working Paper Table 5.76. Recurrent Spending on Social Protection by Ministries and Local Governments In Real Terms (2007) 2004 2004 2005 2005 2006 2007 (in million RwF) Budget Actual Budget Actual Budget Budget District Total 1,775 1,640 1,543 1,379 1,656 1,551 PRIMATURE +MIGEPROF 535 297 792 491 332 726 MINEDUC 217 69 213 142 342 208 MINISANTE 380 235 484 566 601 432 MINALOC 11,598 11,034 10,393 10,344 10,329 11,147 Total 14,509 13,274 13,424 12,922 13,258 14,064 Social contributions by Gov. 5,386 5,748 6,718 5,870 6,735 6,487 FARG additional sources 697 697 711 711 673 650 TOTAL 20,591 19,719 20,854 19,504 20,666 21,201 % Shares* District Total 12% 12% 12% 11% 13% 11% PRIMATURE +MIGEPROF 4% 2% 6% 4% 3% 5% MINEDUC 2% 1% 2% 1% 3% 2% MINISANTE 3% 2% 4% 4% 5% 3% MINALOC 80% 83% 77% 80% 78% 79% Total 100% 100% 100% 100% 100% 100% Source: Social Protection, PER Report, 2006, page 6. Note: *Percentage shares of the total excluding governments social contributions and FARG additional sources. Donor assistance made significant contributions in the sector: Donor assistance represented twothirds of total identified SP expenditure. In 2006, 97 percent of expected donor assistance was not included in the budget. Aid was RwF 31 billion in 2004, RwF 32 billion in 2005, with early 2006 projections of RwF 40 billion. The expected increase in 2006 was mainly from EU spending on Ubudehe and Global Fund for AIDS, Tuberculosis and Malaria (GFATM) support for mutuelle health payments. Further expenditure analysis follows below, in the context of reviewing groups benefitting from SP spending. Social Protection Sector Results The exact number of recipients of SP spending is unclear as systematic poverty impact coordination and monitoring is lacking. An estimated 95 percent of the population is not covered by SP, and most depend on household and communitylevel mechanisms to cope with individual risk. However, some of the known beneficiaries and the mechanisms that support them are listed in table 5.77, which also shows the shares of budget spending by categories of beneficiaries. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 177 Table 5.77. Allocation of Government Social Protection Spending: Budget vs. Actual 2004­07 In Real Terms 2004 2004 2005 2005 2006 2007 (in million RwF and %) Budget Actual Budget Actual Budget Budget Contributions for Public 5,386 5,748 6,719 5,870 6,734 6,487 Employees Social Security, public 2,647 2,612 2,939 2,549 2,996 2,242 Health insurance public 2,739 3,136 3,780 3,320 3,738 4,245 Other Direct Support to 19,370 15,112 16,226 20,363 15,220 17,203 Households - Genocide survivors 11,302 11,302 10,575 10,575 10,774 11,608 - OVCs 827 809 667 641 1,018 782 - Women 2,681 1,265 999 1,904 1,280 1,019 - Disabled 232 80 196 84 174 1,220 - Health costs 295 214 293 424 787 830 - Ubudehe 0 0 1,800 5,204 112 0 - Labor intensive public works 1,968 670 889 889 561 350 Policy, Rights, Advocacy & 1,543 784 1,317 1,411 1,032 932 Management - Women and gender 762 228 456 391 162 434 - OVCs 62 664 121 791 810 219 TOTAL 26,298 21,643 24,262 27,643 22,986 24,621 2004 2004 2005 2005 2006 2007 Percentage of Total Budget Actual Budget Actual Budget Budget Social Contributions, Public Employees 20% 27% 28% 21% 29% 26% Other Direct Support to Households 74% 70% 67% 74% 66% 70% - Genocide survivors 43.0% 52.2% 43.6% 38.3% 46.9% 47.1% - OVCs 3.1% 3.7% 2.7% 2.3% 4.4% 3.2% -Women 10.2% 5.8% 4.1% 6.9% 5.6% 4.1% - Disabled 0.9% 0.4% 0.8% 0.3% 0.8% 5.0% - Health costs 1.1% 1.0% 1.2% 1.5% 3.4% 3.4% - Ubudehe 0.0% 0.0% 7.4% 18.8% 0.5% 0.0% - Labor intensive public works 7.5% 3.1% 3.7% 3.2% 2.4% 1.4% Policy, Rights, Advocacy & Management 6% 4% 5% 5% 4% 4% TOTAL 100% 100% 100% 100% 100% 100% Source: Social Protection, PER Report, 2006, page 6. 178 World Bank Working Paper From table 5.77 above, the following observations can be made: Social security and health insurance for public sector workers accounted for 25 percent to 30 percent of total SP expenditure, a much higher commitment than what was spent on mutuelles for the general population.41 Direct support to households and individuals, where FARG was the largest program, received about 70 percent of expenditures in SP. FARG accounted for around 40 percent of SP spending. Poor and vulnerable groups spending were both low and unstable: · Ubudehe was dependent on EU funding. · Laborintensive work program funding via HIMO was unclear. · District funding (table 5.76) was low and decreasing. Disability spending, based on what can be identified in the budget, was extremely low. The increase seen in 2007 related to special education. Policy, rights advocacy, and management contained some important expenditures (including rightsofchildren and gender equality promotion programs), but this is a broad category and the significant decline in 2004­07 is hard to evaluate. One of the key conclusions made in the Social Protection PER report in 2006 is that offbudget spending in the sector is significant, and yet accurate estimates of its contributions are extremely difficult because of some of the following issues: MINECOFIN maintains the Development Assistance Database (DAD) database,42 but it is not comprehensive. Some donor spending may start off as offbudget but then later be incorporated into the budget.43 Forecasting of future SP expenditure is difficult as development data forecasts and estimates are incomplete. Donor support generally targets direct support to households and individuals, with specific focus on the poor and disadvantaged, but the definition for "target groups" is not always specified. The offbudget data in the PER report is not allinclusive. Table 5.78 presents the offbudget data from 2004­06 that were available. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 179 Table 5.78. Allocation of Off-Budget Donors' Social Protection Expenditure In Real Terms 2004 2005 2006 (In Million RwF) Actual Actual Budget Direct Support to Households 40,487 32,216 42,455 Health costs (mutuelles ) 0 0 5,236 Refugees & returnees 3,519 2,473 2,936 HIV/AIDS Affected 5,861 5,376 5,261 OVCs 10,888 9,086 9,801 Ex combatants 915 7,309 560 Labor intensive public works 5,708 415 655 Policy, Rights, Advocacy & Management 1,787 2,098 1,100 TOTAL 42,274 34,314 43,555 2004 2005 2006 Percentage of Total Actual Actual Budget Direct Support to Households 96% 94% 97% Health costs (mutuelles) 0% 0% 12% Refugees & Returnees 8% 7% 7% HIV/AIDS affected 14% 16% 12% OVCs 26% 26% 23% Ex combatants 2% 21% 1% Labor intensive public works 14% 1% 2% Policy, Rights, Advocacy and Management 4% 6% 3% TOTAL 100% 100% 100% Memo Item: Food Aid RwF 26,374 19,484 22,180 Food aid as % of total 62% 57% 51% Source: Social Protection, PER Report, 2006, page 9. From table 5.78, the following conclusions of offbudget spending contributions to the sector can be drawn: Food aid accounted for 62 percent of extrabudget donor SP expenditure in 2004, but then decreased as the drought ended and nonfood SP programs increased. Direct support received between 94 percent and 97 percent of total SP spending. OVC programs were consistently the largest direct support program in the period, been the largest program, followed each year by HIV/AIDS programs, and refugee and returnee programs. Combined these programs accounted for 48 percent, 49 percent, and 39 percent of overall SP spending in 2004, 2005, 2006.44 Policy, rights advocacy, and management--although a difficult category to define--contributed between 2.5 percent and 6.1 percent of total SP expenditures, with a decrease in the 2006 budget.45 In 2006, health insurance for the general population was boosted by GFATM support and accounted for 12 percent of the total SP budget. 180 World Bank Working Paper Spending on refugees and returnees continued to be financed by World Food Programme (WFP) and UNHCR. Funding as a percentage of total SP expenditure was 8.3 percent in 2004, 7.2 percent in 2005, and 6.7 percent in 2006. Support to HIV/AIDS households, which covers food and nutrition support and helps with generating income but does not include the cost of antiretroviral treatment, contributed 14 percent, 16 percent, and 12 percent in 2004, 2005, 2006, respectively.46 OVC accounted for almost onefourth of total SP spending (which includes WFP school feeding that provided meals to 300 primary schools in food insecure areas; see footnote 15 in PER report, page 8). Demobilization (support for excombatants funded by donors, including the World Bank) represented more than 20 percent of SP in 2005, yet only 2.2 percent in 2004 and 1.3 percent in 2006. Laborintensive public works programs experienced a significant decrease in expenditure as a percentage of total SP spending. In 2004 these programs received 13.5 percent of total funds verses a mere 1.2% and 1.5% in 2005 and 2006, respectively. Finally, table 5.79 illustrates the overall allocation of government and donor spending on SP, including both on and offbudget support, to provide a full overview of all identified SP spending between 2004 and 2007: Table 5.79. Allocation of Total Social Protection Funding 2004­07 in millions RwF % of Total Social Protection Expenditures In Real 2004 2004 2005 2005 2006 2007 2004 2004 2005 2005 2006 2007 Terms Budget Actual Budget Actual Budget Budget Budget Actual Budget Actual Budget Budget Contributions 5,386 5,748 6,719 5,870 6,734 6,487 20.5% 9.0% 27.7% 9.5% 10.1% 12.7% for Public Employees -Social security, public 2,647 2,612 2,939 2,549 2,996 2,242 10% 4% 12% 4% 5% 4% -Health insurance public 2,739 3,136 3,780 3,320 3,738 4,245 10% 5% 16% 5% 6% 8% Other Direct Support to 19,370 55,599 16,226 52,579 57,675 43,642 73.7% 87.0% 66.9% 84.9% 86.7% 85.4% Households, of which: - Genocide Survivors 11,302 11,430 10,575 10,709 10,774 11,608 43% 18% 44% 17% 16% 23% - OVCs 827 11,697 667 9,726 10,819 10,749 3% 18% 3% 16% 16% 21% - Labor intensive public 1,968 6,378 889 1,303 1,216 817 7% 10% 4% 2% 2% 2% works - Women 2,681 1,265 999 1,904 1,280 1,019 10% 2% 4% 3% 2% 2% - Health costs 295 214 293 424 6,024 4,011 1% 0% 1% 1% 9% 8% - Refugees & returnees 1,238 3,563 0 2,473 2,936 0 5% 6% 0% 4% 4% 0% -HIV/AIDS affected 0 5,861 0 5,376 5,261 3,757 0% 9% 0% 9% 8% 7% -Ex-combatants 0 915 0 7,309 560 3,552 0% 1% 0% 12% 1% 7% - Disabled 217 69 190 78 169 1,220 1% 0% 1% 0% 0% 2% -Ubudehe 0 0 1,800 5,204 7,298 3,500 0% 0% 7% 8% 11% 7% Policy, Rights, 1,543 2,571 1,317 3,509 2,132 957 5.9% 4.0% 5.4% 5.7% 3.2% 1.9% Advocacy & Management TOTAL 26,298 63,917 24,262 61,957 66,541 51,086 100% 100% 100% 100% 100% 100% Source: Social Protection, PER Report, 2006, page 10. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 181 As table 5.79 shows, 85 percent of SP spending related to transfers to poor and vulnerable groups, including food aid support, of which less than onefourth was financed from the recurrent budget.47 Government expenditure for pensions and health care for public sector employees increased and accounted for more than 10 percent of total SP spending. Expenditure to support health care access for poor and vulnerable households by financing mutuelle membership represented about 9 percent of total expenditure. This category increased in large part because of support from district budgets and GFATM. Table 5.79 highlights the importance of programs targeting children (which is larger than FARG when school feeding is included) and the importance of the EUsponsored Ubudehe program. Funding to disabled groups is significantly and strikingly low. Social Protection Sector Challenges The SP sector is one of the largest sectors in Rwanda, and the spending has increased in the last few years in both nominal and real terms. The public expenditure review of SP programs showed a number of programs in various ministries and sectors that summed up to 4 percent of GDP. SP spending reaches pockets of groups qualified for benefits. For example, social security and medical insurance costs for public sector workers represents about 10 percent of total SP expenditure, and the government is encouraging community health insurance, which in 2007 covered about 75 percent of the population. Yet, according to PER findings, when it comes to individual risk, 95 percent of the population is not covered by many or any SP benefits, and most depend on household and communitylevel mechanisms to cope with individual risk. About 9 percent of SP spending provides financial support to enable about 10 percent of households to receive financial support with community health insurance fees. Therefore, the SP PER report concludes that spending must fully reflect the government's objective to manage individual and societal risks by establishing policy and institutional frameworks for the development of risk pooling and savings mechanisms, providing basic safety nets and interventions when disasters strikes. As table 5.79 illustrates, the overall emerging pattern seems to broadly address these objectives, with the main exception being the lack of funding for disabled groups and the sharp decline in laborintensive public works programs48 (although public works programs are being piloted under Vision2020 Umurenge). Yet, interventions are scattered, projects are poorly coordinated and not aligned with overall policy, and real poverty impact is limited at best and mostly unknown. Consequently, the government has initiated the following actions to improve coordination: (a) establishment of a SP Sector Working Group, with emphasis on the needs of vulnerable groups, the poor, and the disadvantaged; (b) priority setting, coordination, and rationalization of existing SP policies and ongoing programs of governments, development partners, and NGOs to enhance their impact; (c) a trend from projectstyle assistance toward budget support and SWAPs with development partners; (d) ensuring that any new initiatives are both financially sustainable and technically well structured; and (e) adopting programs to support national and local 182 World Bank Working Paper government capacity building, including training people in relevant skills, and staffing commensurate with the sector's strategic agenda. Maybe the most important of its initiatives in the SP sector is the government's comprehensive framework under Vision 2020 Umurenge, a flagship program under the EDPRS and a highly decentralized and integrated rural development program that is designed to identify vulnerable groups, define the approach to targeting and tailoring specific programs to meet the different categories of vulnerable groups, and accelerate the reduction of extreme poverty in Rwanda. Infrastructure Water and Sanitation Water and Sanitation Sector Overview Rwanda possesses water in abundance through its many lakes, rivers, and swamps (26,338 km2), which cover 8 percent of national territory. Rwanda has 22,300 natural springs and adequate rainfall, but the country uses only 12.2 percent of available sources. Along Rwanda's Water Divide Line--called the CongoNile Ridge--the Congo River Basin in the west drains 10 percent of surface water, and the Nile River Basin to the east drains an estimated 90 percent of surface water. About 69 percent of the urban population receives nationally piped water. Rwanda has few sanitation facilities, and they are close to nonexistent in rural areas. Some large institutions have their own sewage treatment plants, but they often are not supervised by responsible authorities. There is insufficient data regarding sanitation, but it is estimated that 85 percent of the population has traditional household sanitation facilities, although only 8 percent are deemed hygienic. District mayors have signed performance contracts with the government, committing to expand hygienic sanitation facilities to 100 percent of the population. Electrogaz continues to be Rwanda's supplier in urban areas and since 1976 has held monopoly rights under the supervision of MINITERRE. Technical losses were estimated at about 30 percent, many illegal connections further diminish efficiency, and the company lacks a database system and effective customer management.49 In this context, the overall performance of Electrogaz has been considered poor compared with other public utility companies in SubSaharan Africa. In 2006, MINITERRE contracted with a consultant50 to conduct a thorough review of the country's water and sanitation service (WSS) sector covering 2000­05. The key conclusions of the WSS PER report are summarized below: The WSS sector was neglected in terms of allocation in comparison with key sectors such as education and health. Sixtynine percent of the urban population and 55 percent of the rural population had access to potable water. Eight percent of the urban population and 10 percent of the rural population had access to latrines and good sanitary conditions. The total amount disbursed to the WSS sector between 2003 and 2005 was RwF 32.5 billion (US$584.9 million). Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 183 Of the total 2005 disbursement in the WSS sector, 58 percent was from donor agencies (external financing), and most of the projects were implemented in rural areas. The physical impact of WSS spending between 2003 and 2005 is summarized in table 5.80. Table 5.80. Water and Sanitation Rehabilitation, Construction, and Management 2003­05 Rehabilitation of 25 water systems Building of 35 potable water supply (AEP) systems51 Management of 1,800 water resources Construction of 120 water reservoirs Construction of 59 forages Construction of 600 public latrines Source: WSS PER Report, 2006. The total coverage rate increased progressively during 2003 and 2005, and an additional 1,060,700 people were served. Sector limitations included a weak monitoring and evaluation system (especially relating to the decentralized transfers to the districts of responsibilities for rural water supply systems), but various initiatives were already in place. The 1987 Presidential Decree set out principles for ownership and management. Financing mechanisms exist under MINALOC (see table 5.81). Table 5.81. Water and Sanitation Financing Mechanisms under MINALOC Common Development Fund (CDF) Labor Intensive Public Works (HIMO) Ubudehe Program Source: WSS PER Report, 2006. Sanitation progress did not match the progress in water supply; there is a lack of information, sanitation codes, and streamlined coordination in the sector. POLICY AND LEGAL FRAMEWORK OF THE WSS SECTOR. To understand the context of the WSS sector, it is helpful to first consider the policy and legal framework under which MINITERRE, the main coordinator of the sector, operates. Table 5.82 summarizes the four main policies that provide guidance for the WSS sector. 184 World Bank Working Paper Table 5.82. Rwanda's Water and Sanitation Sector Policy Overview VISION 2020 To increase access to potable water by 2.5 percent per year from the current rate of 57 percent and to ultimately provide drinkable water to all Rwandese by 2020. MDG (Millennium Development Goals) To reduce by half the percentage of the population that has no sustainable access to drinking water and sanitation by the year 2015. PRSP (Poverty Reduction Strategy Policy) Outlines overall improvement goals. National Investment Strategy To encourage private sector investment and provide affordable service to customers. Example of private sector investments Agenda PSP (south) serves 30 percent of potential consumers at RF 14 per 20 liters. However, the average consumption of 4­5 liter per day per person is still lower than the standard of 20 liters per day per person. Source: WSS PER Report, 2006. Legally, the sector is monitored by the 1987 Presidential Decree and the Rwanda Utilities Regulation Agency (established in 2001). Key players in the sector, besides the government and Electrogaz, include private operators, user associations, communal organizations, and intermunicipal bodies called "waterpoint communities." ORGANIZATIONAL STRUCTURE AND INSTITUTIONAL MAPPING OF THE WSS SECTOR. In 2006, MINITERRE held the main responsibility for the WSS sector, although other ministries also are involved in the sector (see table 5.83). Table 5.83. Water and Sanitation Cross-Sector Involvement Cross-Sector WSS involvement: MINITERRE ­ overall policy formulation, coordination, implementation, and so forth MINAGRI ­ manage and coordinate rural engineering, marshlands, and catchment areas MINISANTE ­ manage codification of health standards and ensure compliance MINALOC ­ financing, management, and supply of water in rural areas through DCF, Ubudehe, and HIMO MINECOFIN ­ determines tariffs and finances of sectoral investments, supervises Electrogaz DISTRICTS ­ delegate management to authorized water users and set up intermunicipal bodies Source: WSS PER Report, 2006. Table 5.84 provides additional mapping of WSSrelated institutions and players. (The role of donor activities will be highlighted in a later section.) Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 185 Table 5.84. Institutional Roles and Responsibilities in the Water and Sanitation Sector INSTITUTIONS ROLES AND RESPONSIBILITIES MINITERRE Its responsibility is to define the overall Policy of the sector through policy-making process. Funding of urban and rural water supply is by the central government using donor funds but also by NGO's using bilateral aid. WSS Department of The responsibilities consist of organization of activities involving the water and MINITERRE sanitation sector; implementation of government policies in the sector; supervision of existing infrastructure and facilities; planning and financing of water and sanitation projects and training of staff. The unit is also responsible for water resource management ELECTROGAZ The only company that produces and distributes electricity and water. The public utility is entrusted to produce and distribute drinking water on a commercial basis. The utility operates and manages the water supply in urban and semi-urban areas. DISTRICTS Responsibility for scheme planning is held at the district level but the department of water and sanitation assists with complex schemes. Districts own water systems in rural areas and are formally responsible for operation of all schemes but operation of simpler schemes is usually delegated to a community level. CDF (Community The CDF was formed in 2002 for the purpose of financing the development Development Fund) projects of the districts and towns, monitoring the use of funds allocated to development projects, playing an intermediary role between the districts and towns, and donor financing, in particular the development projects of these entities. CDF creates a favorable framework for channelling funds to the beneficiaries in the sector UBUDEHE Ubudehe is a Rwandan tradition woven around the culture of working together as a community to resolve problems of the rural dwellers. Groups of households join forces to work the fields of the entire community. The program finances water and sanitation projects in the rural areas. The program also creates a platform for Information, Education and Communication (IEC) campaigns for hygiene education. HIMO A labor intensive infrastructure development program, which was adopted in 2003, with the objective to contribute to poverty reduction by providing employment opportunities. Water and sanitation has and will continue to apply this approach in the execution of water and sanitation infrastructure both at program and project levels. Rwanda Utilities The agency is charged with regulating the WSS sector, through regulating tariff Regulatory Agency levels to prevent monopolistic exploitation by utilities and tariff structures to (RURA) promote equity objectives. Monitors performance in the sector and advises the government on matters pertaining to the sector, Arbitrates and offers dispute resolution procedures for conflicts. NGOs Play role of providing water and sanitation facilities in the rural areas for example AVSI, MSF Belgique, CICR, COFORWA, Rwanda Red Cross, AVSN. COMMUNITY & Sanitation provision is performed by private organization and individuals which INDIVIDUALS also contribute as beneficiaries to the construction or set up of water, and sanitation infrastructure. Under the water and sanitation program, the community funds 10 percent of the capital cost. NGO-led schemes typically involve similar community funding. The majority of the schemes are delegated to community committees which appoint someone to look after facilities. Source: WSS PER Report, 2006, page 18. POPULATION COVERAGE OF WSS. Rwanda's 3rd General Census (2002), which provided the basis for the PER report on WSS in 2006, estimated Rwanda's population at 8.2 million people and an annual population growth rates of 3 percent. Current estimates indicate a population closer to 9 million people. Based on the PER analysis and the 3rd General Census, it was estimated that 83 percent of the population lives in rural areas. Rwanda experienced 186 World Bank Working Paper rapid urban growth rates between 1991 and 2002, increasing from 5.5 percent to 16 percent of the total population (MINECOFIN 2004c), and the population between the ages of 25 and 64 had the highest growth rate. However, looking beyond 2002, the highest growth rates occur in age groups 15­19 and 20­24. Table 5.85 summarizes estimated population trends from the 3rd General Census. Table 5.85. Population Trends 1997­2020 AGE Growth Growth BRACKET 1997­ 1997­ In thousands 1997 2002 2007 2012 2017 2020 2007 2020 00­04 1,177 1,456 1,529 1,519 1,694 1,889 30% 60% 05­09 1,121 1,110 1,384 1,462 1,630 1,818 23% 62% 10­14 611 1,102 1,093 1,365 1,522 1,697 79% 178% 15­19 474 1,052 1,082 1,077 1,201 1,338 128% 182% 20­24 445 865 1,027 1,059 1,181 1,317 131% 196% 25­64 2,044 2,340 2,951 3,693 4,117 4,590 44% 125% 65+ 248 238 234 245 272 305 ­6% 23% TOTAL 6,120 8,163 9,300 10,420 11,617 12,954 52% 112% Source: WSS PER Report, 2006, page 4. Original Source: 3rd General Census of Population and Housing 2002. Water and Sanitation Public Expenditure Allocation Review Institutions contributing to the sector use a number of channels (the way funds are mobilized and allocated) and sources of funds. This section looks at total resource flow in the sector and financing mechanisms. Table 5.86 shows the allocation and execution of expenditures by institutions. All expenditures are in 2007 real prices unless otherwise indicated. Key trends from table 5.86 include the following: Actual disbursement measured in real terms increased from RwF 8.5 billion in 2003 to RwF 25.1 billion in 2005, a 197 percent increase. Actual disbursement measured in real terms increased between 2003 and 2004 by 18 percent and between 2004 and 2005 by 151percent. Domestic funding by the government increased in real terms by 5 percent and 142 percent in 2004 and 2005, respectively, while external funding increased by 32 percent and 158 percent in the same years. Government and external financing increased actual disbursement in real terms with 153 percent and 240 percent, respectively for the entire period between 2003 and 2005. Actual resources of MINITERRE increased each year by 516 percent and 454 percent between 2003 and 2005. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 187 Table 5.86. Water and Sanitation Resource Flow 2003­05 Growth Growth 2003 2004 2005 2003-2004 2004-2005 Budget Actual Budget Actual Budget Actual Budget Actual Budget Actual GOVERNMENT 2,947 4,188 4,100 4,379 8,632 10,610 39% 5% 111% 142% MINITERRE 228 127 925 781 4,121 4,327 305% 516% 346% 454% ELECTROGAZ 2,685 2,792 3,164 3,164 4,312 4,312 18% 13% 36% 36% Recurrent 1,386 1,386 2,501 2,501 3,542 3,542 80% 80% 42% 42% Development 1,299 1,405 662 662 769 769 ­49% ­53% 16% 16% MINAGRI 0 0 0 0 157 152 n.a. n.a. n.a. n.a. MINISANTE 33 418 11 10 9 9 ­66% ­98% ­20% ­8% MINALOC 0 851 0 424 33 1,811 n.a. ­50% n.a. 327% UBUDEHE/Province 0 0 0 0 33 23 n.a. n.a. n.a. n.a. CDF 0 851 0 325 0 1,682 n.a. ­62% n.a. 418% HIMO 0 0 0 100 0 105 n.a. n.a. n.a. 6% EXTERNAL 8,485 4,277 14,585 5,643 19,029 14,534 72% 32% 30% 158% FINAN. Projects 8,485 3,876 14,585 5,228 19,029 13,632 72% 35% 30% 161% NGOs 0 401 0 415 0 901 n.a. 4% n.a. 117% TOTAL 11,432 8,464 18,685 10,022 27,662 25,144 63% 18% 48% 151% RESOURCES Gov. Allocation Growth 2003-2005 193% 153% External Finan. Allocation Growth 2003-2005 124% 240% Overall Total Resource Allocat'n Growth 2003-2005 142% 197% Source: WSS PER Report, 2006, page 21. Original Source: Government budget, ELECTROGAZ and NGOs. Actual disbursement by MINISANTE in the WSS sector decreased significantly in the period. MINAGRI, which had no resource allocation for the WSS sector in 2003 and 2004, received RwF 152 million in 2005 (in 2007 real prices). Financing mechanisms under MINALOC received fluctuating resource allocations. For example, the CDF allocation for the WSS sector went down in 2004 by 62 percent but was significantly increased in 2005 by 418 percent, while the new mechanism, HIMO (established in 2003), received a 6 percent allocation increase in its first year as measured in real terms. Electrogaz continued to receive increased overall actual allocations from MINECOFIN, with 13 percent and 36 percent actual yeartoyear disbursement increases in 2004 and 2005. Electrogaz's largest actual disbursement increases occurred in its recurrent budget, showing 80 percent and 42 percent yeartoyear increases in real term allocations for 2004 and 2005, respectively. This contrasts with the lower allocations for its development budget, which decreased by 53 percent between 2003 and 2004, and then increased by 16 percent in 2005--but still at a lower allocation in actual terms in 2005 compared with two years earlier 188 World Bank Working Paper (RwF 1,405 million in 2003 versus RwF 769 million in 2005 measured in real terms). The donor community provided 58 percent of actual disbursements in the sector, compared with 42 percent by the Government. MINITERRE and Electrogaz received the greatest portion of government funding in the sector with 17 percent each of total sector funding, followed by MINALOC receiving 7 percent through its CDF. EXPENDITURES BY SUBPROGRAM. Water and sanitation under MINITERRE operates through four subprograms (see table 5.87). Table 5.87. Water and Sanitation Expenditure by Subprogram 2003­05 2003 2004 2005 2003 2004 2005 % of % of % of Sub-program Budget Actual Budget Actual Budget Actual Total Total Total Management Support 149 85 105 68 39 15 1% 1% 0% Sanitation 759 839 2,460 362 3,513 1,713 10% 4% 7% Potable Water 10,311 7,504 15,323 9,544 23,433 23,066 89% 95% 92% Water Resource Mgmt 216 38 796 50 677 350 0% 0% 1% TOTAL 11,435 8,464 18,685 10,023 27,662 25,144 100% 100% 100% Source: WSS PER Report, 2006, page 21. *Percentage calculated on actual expenditures. The potable water program was by far the largest of the four programs in the sector, with a consistent allocation in the period of 89 percent to 95 percent of the total. The sanitation program received significantly less than the potable water program, with a 7 percent resource allocation in 2005, down from 10 percent in 2003. This subprogram was further plagued with lack of data for sanitation, making an overall assessment of allocation and effectiveness of resources allocated difficult. The 2006 WSS PER report pointed out that while on a national level, subprograms were the same in recurrent and development budgets, the lack of integration of donor activities made it difficult to see trends in recurrent, development, and overall expenditures. Furthermore, there is an overlap of project data between the different ministries and MINECOFIN. ROLE OF DONOR AGENCIES. The donor community plays a significant role in the WSS sector. As noted earlier, the donor community provided 58 percent of actual disbursements in the sector compared with 42 percent for the government. Electrogaz received the greatest portion of government funding to the sector with 17 percent of total sector funding, followed by MINALOC receiving 7 percent of government funding through its CDF. Donor funding follows a convention signed by MINECOFIN and the donor. Table 5.88 shows the principal donors. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 189 Table 5.88. Multilateral and Bilateral Donors in the Water and Sanitation Sector Principal multilateral donors Principal bilateral donors World Bank Germany Foundation for International Development Assistance (FIDA) Austria African Development Bank Belgium Arabic Bank for Economic Development for Africa (BADEA) Japan RFA United Nations Children's Fund (UNICEF) European Union(EU) Source: WSS PER Report, 2006. As table 5.89 illustrates, actual disbursements in the sector increased progressively in real terms between 2003 and 2005, from RwF 3.6 billion in 2003 to RwF 13.6 billion in 2005, a 276 percent increase in actual disbursements in the threeyear period. Table 5.89. Water and Sanitation Project Disbursement and Realization 2003­05 In Real Terms (in million RwF) 2003 2004 2005 Exec. Exec. Exec. PROJECTS DONOR Budget Actual % Budget Actual % Budget Actual % 8 ex communities Kigali RFA 1,710 1,698 99% 2,066 1,894 92% 847 928 110% AEPA Gikongoro Austria 745 0 0% 660 14 2% 425 95 22% PGNRE IDA 180 8 4% 754 36 5% 517 197 38% Bugesera Karenge EU 1,527 25 2% 1,604 14 1% 2,368 3,414 144% 4 AEP in rural areas EU 1,527 0 0% 0 0 n.a. 0 0 n.a. AEPE of Kigali BAD 0 0 n.a. 0 0 n.a. 1,918 265 14% PNEAR BAD 0 0 n.a. 0 0 n.a. 1,441 98 7% PEAMR IDA 1,302 1,120 86% 7,279 1,027 14% 8,976 4,689 52% AEP Butare BADEA 1,059 266 25% 1,921 1,564 81% 771 2,086 271% Technical Support BADEA 74 74 100% 65 65 100% 15 15 100% PDRCIU FIDA 149 149 100% 0 0 n.a. 0 0 n.a. Ubudehe EU 0 0 n.a. 0 0 n.a. 0 93 n.a. New Water PNUD 0 0 n.a. 199 199 100% 0 0 n.a. Sources&Extens. Competitiveness IDA 0 0 n.a. 0 0 n.a. 1,751 1,751 100% Projects/W WES UNICEF 360 285 79% 40 0 0% 0 0 n.a. TOTAL 8,633 3,624 42% 14,586 4,813 33% 19,029 13,631 72% TOT Proj Disburs. Growth 2003-05 120% 276% Source: WSS PER Report, 2006, page 25. Original Source: CEPEX/MINECOFIN. Note: APEA=Potable Water Supply and Sanitation; PGNRE=National Program for Water Resource Management; AEP=Potable Water Supply; PNEAR=National; Program for Water Supply and Sanitation in Rural Areas; PEAMR=Water Supply and Sanitation Project in Rural Areas; PDRCIU=Umutara Community Resource and Infrastructure Development Program. 190 World Bank Working Paper However, overall execution in the sector remains significantly low, at 42 percent, 33 percent, and 72 percent each of the three years, respectively. Many projects underexecuted. Only one consistently met its target at 100 percent in the three years, while only a handful of other projects met their targets in one of the years in the period. The remaining projects significantly overexecuted at rates of 144 percent to 271 percent. SECTOR DEVELOPMENT BUDGET COMPARED WITH OTHER PRIORITY SECTORS. This section compares WSS development spending to other priority sectors, including energy, health, and education. According to the estimates of the WSS PER of 2006, the following differences between priority sectors existed at the time of the report: WSS received 17 percent and 22 percent of overall development priority spending in 2003 and 2004, respectively (see table 5.90). Health received an increasing portion of the overall development priority spending in 2003 and 2004, and education got a decreasing portion over the two years. Yet health and education together represented 80 percent and 76 percent of overall development priority spending in 2003 and 2004, respectively. Although WSS received less than the two largest sectors, health and education, these two larger sectors also have WSSrelated expenditures benefiting the WSS sector. In 2003 and 2004, energy was, by far, the lowestreceiving priority sector in terms of both budgeted and allocated resources. However, WSS, with rather low execution rates of 51 percent in 2003 and 61 percent in 2004, had significantly higher execution rates than the energy sector. Table 5.90. Water and Sanitation Sector Development Budget 2003­0452 2003 2004 2003 2004 SECTOR Budget Actual Exec. % Budget Actual Exec. % % of Tot % of Tot Health 4,871 4,779 98% 10,333 9,051 88% 22% 46% Water & Sanitation 7,089 3,623 51% 7,136 4,339 61% 17% 22% Education 12,200 12,435 102% 12,240 5,854 48% 58% 30% Energy 6,486 696 11% 5,112 321 6% 3% 2% TOTAL 30,647 21,533 70% 34,821 19,566 56% 100% 100% Source: WSS PER Report, 2006, page 21. Original Source: CEPEX/MINECOFIN. *Percentages calculated on actual. INVESTMENT REQUIREMENTS IN WSS. The government has estimated that US$925 million will be required over the next 16 years for the country to attain rural water and sanitation targets, including US$70 million every year over the next 3 years (as stated in 2006). Identified investment requirements in urban areas include the following: Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 191 An estimated US$750 million will be required to address water and sanitation needs in the city of Kigali. The key objective in Kigali is to cut Electrogaz's resource deficit of approximately 32,000 m3/day (a deficit estimated to grow to 100,000 m3/day in 2020 if water resources are not mobilized) In the short term (2005­07) an estimated investment of US$75.3 million is required.53 In the long term (2008­20) an estimated investment of US$626 million is required. Identified investment requirements in rural areas include the following: An estimated 30 percent of existing infrastructure requires rehabilitation. To address existing and new investment needs to reach rural area objectives by 2020, an estimated US$925 million were determined in 2006 to be required (of which US$85.4 million was secured) (WSS PER 2006). Tables 5.91, 5.92, and 5.93 show the WSS investment requirements by types of programs, geographic regions, and coverage targets. Table 5.91. Breakdown by Program Component Showing New Investment Requirements in Rural Areas to Meet 2020 Objectives COMPONENTS US$ millions A. STUDIES & PLANNING TOOLS $10.95 Putting in Place a Planning System $1.38 Engineering and Background Studies $9.57 B. INSTITUTIIONAL SUPPORT & CAPACITY BUILDING $24.11 Capacity Building at the Central Level $4.77 Capacity Building at the Decentralized Level $1.90 Support to the Private Sector and NGOs $13.52 Training of Managers and Technicians $3.92 C. DEVELOPMENT of WSS INFRASTRUCTURE $626.80 Development of Water Infrastructure $361.05 Development of Sanitation Infrastructure $247.91 Support Measures $17.84 D. PROGRAM MANAGEMENT $14.51 Program Management $4.46 Technical Assistance $10.05 PROGRAM TOTAL $676.37 Physical Contingencies $98.34 Price Escalation $150.12 TOTAL OVERALL COST $924.83 Source: WSS PER Report, 2006, page 46. Original Source: Natural Rural Water Supply and Sanitation Program (Metha 2004). 192 World Bank Working Paper Table 5.92. Water and Sanitation Investment Requirements by Region Popul. Popul. Water Sanitation Supplied Supplied Investment Access Access w/Water w/ Sanitat'n US$ REGIONS Rate Rate 2020 2020 million SOUTH 46% 7% 1,554,751 2,489,861 $197.22 EAST 43% 6% 2,212,257 3,062,705 $265.12 NORTH 37% 5% 1,971,566 3,034,924 $246.12 WEST 40% 5% 1,772,812 2,574,179 $216.35 TOTAL 7,511,386 11,161,669 $924.81 Source: WSS PER Report, 2006, page 47. Original Source: Based on Table 12 in WSS PER Report, 2006. Table 5.93. Water and Sanitation Coverage Targets and Estimated Investment Requirements Total Invest. Requirements Added Popul't Public To Be Rehab/ Investmt. 2002 2005 2015 Covered New Replace Total Requiremt Access Access Access Mill/Year Rural 41% 55% 85% 222 13 31 44 40 Water Urban 66% 69% 85% 303 24 4 28 0 TOTAL 53% 57% 85% 525 37 35 72 40 Rural 8% 10% 66% 342 3 2 5 1 Sanitation Urban 88% 10% 66% 264 4 1 5 0 TOTAL 8% 10% 66% 606 7 3 10 1 Source: WSS PER Report, 2006, page 48. Original Source: United Nations Population Fund (UNFPA)/PNEAR/PAEMR/Government Investment Plan/Policy. Water and Sanitation Sector Results This section looks at the WSS sector results, with a special emphasis on 2001­05 as covered in the sector's most recent WSS PER (2006). Overall, significant success was reached in the water subsector, with the percentage of the population with access to safe drinking water growing steadily toward the goal of 80 percent by 2010. The percentage of the population with access to proper sanitation facilities is much lower, although it is also rising. By 2020, it is the government's goal to ensure that 100 percent of the population has access to both. In line with the advancement of the Aid Coordination, Harmonization, and Alignment agenda, the ministry in charge of water and sanitation has put in place a policy of encouraging donors to concentrate their efforts in specific regions of the country. Furthermore, the ministry, in collaboration with donors, has undertaken reforms since 2004, including establishing sectoral working groups, and it has also developed a resultsdriven MTEF. As part of the decentralization process, districts will play an increasing role in implementation in the WSS sector. Table 5.94 highlights the reforms undertaken in WSS since 2004. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 193 Table 5.94. Water and Sanitation Reform Measures 2004­05 Theme/Sector Reform Measures Undertaken Year Water Sectoral working group established 2005 Water policy drafted 2005 Water supply activities at district and community level implemented 2004/05 Source: WSS PER Report, 2006. With respect to the action points identified during the Joint Budget Support Review (JBSR) in September 2007, the ministry reported that systems to manage basic services were being installed in regions that did not previously have such systems, and the ministry envisioned that these systems would become functional in the near future. The ministry is also working on a program to encourage investment in the water and sanitation subsector, which will continue until the end of the year. A number of donors are in the process of mobilizing funds to scale up their efforts in this area. Furthermore, 2008 was identified as the International Year of Water and Sanitation, and the government is aggressively pursuing several initiatives to show Rwanda's commitment to water and sanitation in the country. KEY FINDINGS BASED ON PERFORMANCE INDICATORS. As table 5.95 shows, slight progress in the sector was achieved, based on sector performance indicators measured in the years 2001 and 2005. The greatest percentage improvement in the indicators occurred in the percentage of the rural population with access to potable water--a 14 percentage point improvement between 2001 and 2005. The percentage of the urban population with access to potable water grew less rapidly in the period, with a coverage rate of 66 percent in 2001 improving to 69 percent in 2005. Sanitation significantly lagged behind drinking water in the percentage of both urban and rural populations having access to hygienic sanitation. In 2005, the population with access to such services was a mere 10 percent and 8 percent in urban and rural areas, respectively. The functioning of existing water systems saw only a 5 percentage point improvement in the fouryear period, but it still is the indicator closest to reaching its 2015 MDG target of 85 percent coverage. Most systems relating to the WSS sector were not managed by professional operators, having clear implications on how systems were operated, managed, maintained, and kept in compliance with existing standards. Overall, all performance indicators for the sector were far off their 2010 and 2015 targets, with sanitation lagging behind the most. 194 World Bank Working Paper Table 5.95. Water and Sanitation Performance Indicators 2001­15 WATER & SANITATION PERFORMANCE INDICATORS 2001 2005 2010 2015 % of population with access to potable water in rural areas 41% 55% 66% 85% % of population with access to potable water in urban areas 66% 69% 75% 85% % of functioning of existing water systems 70% 75% 90% 85% % of systems managed by professional operators 4% 8% 25% 35% % of population with access to hygienic sanitation in urban areas 6% 10% 34% 66% % of population with access to hygiene sanitation in rural areas 6% 8% 34% 66% Source: WSS PER Report, 2006, page 30. Original Source: MINITERRE. URBAN WATER SUPPLY AND SANITATION SERVICES. Overview of urban water supply: Access to potable water in Kigali and other urban areas was around 69 percent in 2005, and Electrogaz provided water services to Kigali and many other urban centers. At time of the 2006 PER report, Electrogaz operated 22,134 private connections and 156 standpipes (123 working and 33 not working). Electrogaz's central problem was considered to be the discrepancy between production capacity and demand. In 2006, daily production capacity was 25,000 m3, compared with an estimated demand of 60,000 m3. Therefore, plans are in place to expand production through a treatment plant in Nyabarongo River. Furthermore, in August 2003, the government entered into a fiveyear performancebased management contract with a consortium led by Lahmayer International (Germany) to provide technical assistance. No further note in the PER report revealed the progress of this performancebased contract. Overview of urban sanitation: No real piped wastewater system existed in urban areas, and the storm water system was inadequate. Instead, a drainage system called "Ruhurura" was used in specific areas. Some big premises had autonomous sewage systems on a small scale, but industrial waterwaste and lack of supervision and control were significant problems. Table 5.96 shows the sanitation coverage rates according to the National Census of 2001. Table 5.96. Urban Sanitation Coverage Rates 2001 85% of urban people crude health installations 40% of urban families had toilets with septic tanks 53% of urban families had pit latrines without water 7% of urban families had no latrines at all Source: WSS PER Report, 2006. Some public toilets existed and were usually managed by private individuals near markets and taxi parks. The cost for an individual to use such facilities was about RwF 50. Also, a solid waste collection and transportation site existed at Nyanza, managed by the city of Kigali, and was complemented by services of some associations that collected and recycled solid waste to produce charcoal and plastics. Nonetheless, the overall sanitation services provided by these groups were limited. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 195 Performance indicators in urban WSS: Table 5.97 shows performance indicators for water production, water sales, losses, connections, and revenue turnover covering all systems within the urban and semiurban areas from 2001 to 2005. This service was the responsibility of Electrogaz. Water production increased slightly in the fiveyear period, with a slight 3 percent decrease in 2005 resulting from a drop in water levels from a drought and environmental degradation. Twenty percent of Electrogaz water production occurred outside Kigali. Water sales continued to increase, although at various rates each year. Overall in the fiveyear period, sales increased by 3 million m3, as a result of reduced technical and nontechnical losses over the period. Technical and nontechnical losses in the period were reduced from 55 percent in 2002 to 30 percent in 2005, yet Electrogaz still struggled to meet increasing urban demand and was not operating at maximum capacity. Revenue turnover increased from RwF 2.4 billion in 2001 to RwF 3.3 billion in 2005, representing cumulative growth of 38 percent from 2001 to 2005 and equaling an average turnover growth rate of 7 percent. This increase in turnover is attributed to tighter supervision of illegal connections and debt recovery from customers. The number of connections increased in the urban areas in the fiveyear period, showing a cumulative growth rate in connections of 30 percent. The increases in connections occurred mainly in the urban areas of KigaliNigali, Gisnenyi, and Ruhengeri, but Electrogaz was still far behind in meeting its urban demand. Accessibility to water systems expressed by pipe concentration increased in the period (not shown in table 5.97). Table 5.97. Water and Sanitation Performance Indicators 2001­05 Growth % Growth % INDICATOR 2001 2002 2003 2004 2005 2004-05 2001-05 Water Production (m3) in mill. 14.4 15.3 15.5 16.3 15.8 ­3% 10% Annual change 6% 1% 5% ­3% Water Sales (m3) in mill. 7.4 8.5 8.9 9.1 10.0 10% 35% Annual change 15% 5% 2% 10% Tech. & Nontech. Loss 49.0 55.0 47.0 32.0 30.0 ­6% ­39% Annual change 12% ­15% ­32% ­6% Connections (total) 30,644 32,230 34,871 38,479 39,708 3% 30% Annual change 5% 8% 10% 3% Revenues (RwF) in bill. 2.4 2.7 2.9 3.0 3.3 10% 38% Annual change 13% 7% 3% 10% Source: WSS PER Report, 2006, page 32. Original Source: Water Statistics Division/Electrogaz. 196 World Bank Working Paper Notable is that most urban customers pay substantially more for water than Electrogaz's average rate. Electrogaz sells water to standpipe operators at a flat rate of RwF 200 per m3, but there is no control over prices that operators charge end users (WSS PER 2006). Hence, the normal price charged by standpipe operators for 20 liters within Kigali is RwF 20 (WSS PER 2006). Poor household have little alternative for safe resources, but this price represents about three times higher than the price that wealthier households pay, because they have more options for paying for water. Yet, tariffs are below the cost of treatment and do not allow for proper operation and maintenance (WSS PER 2006). RURAL WATER SUPPLY AND SANITATION SERVICES Overview of rural WSS: The proportion of the rural population benefiting from main household supply and tap water is negligible. Rural communities depend on communal sources such as wells, springs, and standpipes. Although some communal sources provide safe drinking water, the rural population still largely depends on water from unprotected and untested sources. The price the rural population paid at the source in 2005 was RwF 5­15 for 20liter jar, of which RwF 3 was kept by the person employed to look after of the source (WSS PER, 2006). Figure 5.11 illustrates the distances it takes for householders to reach a communal water source, by the former provinces. Figure 5.11. Distance to Reach Water-Access Point in Ex-Provinces Kigali Ville Kibuye Cyangugu Gisenyi Gikongoro Byumba Ruhengeri Butare Kigali Ngali Kiburgo 0 200 400 600 800 1,000 1,200 1,400 Distance in meters Source: WSS PER Report, 2006, page 16. Table 5.98 highlights the water and sanitation infrastructure in rural and semi urban areas, where the majority of the population still relies heavily on communal water resources. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 197 Table 5.98. Water Infrastructure in Rural and Semiurban Areas 200554 Rural Piped Sys. Taps per Connections Popul. 2002 Piped Protected per '000 '000 per '000 EX-PROVINCE Census System Taps Connections Sources Inhabitants Inhabitants Inhabitants Butare 722,616 63 585 3,206 2,512 0.09 0.81 4.44 Byumba 712,372 103 930 1,149 1,575 0.14 1.31 1.61 Cyangugu 609,504 59 634 1,482 1,388 0.10 1.04 2.43 Gikongoro 492,607 43 418 897 1,481 0.09 0.85 1.82 Gisenyi 867,225 121 985 3,029 1,918 0.14 1.14 3.49 Gitarama 864,594 104 1,048 2,492 3,262 0.12 1.21 2.88 Kibungo 707,548 56 520 1,969 387 0.08 0.73 2.78 Kibuye 467,745 85 772 784 2,696 0.18 1.65 1.68 Kigali Nagali 792,542 65 912 375 906 0.08 1.15 0.47 Ruhengeri 894,179 103 806 1,757 2,430 0.12 0.90 1.96 Umutara 423,642 19 434 106 113 0.04 1.02 0.25 TOTAL 7,554,574 821 8,044 17,246 18,668 0.11 1.06 2.28 Source: WSS PER Report, 2006, page 35. Original Source: Water and Sanitation Department/MINITERRE. Butare had the highest rate of connections per inhabitant, or 18 percent of all connections, followed by Gisenyi, Gitarama, Kibungo, and Cyangugu. Kibuye had the highest rate of taps per inhabitant, followed by Byumba, Gitarama, KigaliNagali, and Gisenyi. Kibuye had the highest rate of piped water systems, followed by Byumba, Gisenyi, Gitarama, and Ruhengeri. The rural population in Gisenyi, Gitarama, Kibuye, and Byumba had overall better access to all three systems, while Butare's population had a higher concentration of connections than the other exprovinces. Additional data suggest55 that Butare and Gitarama had the highest coverage rates relating to population having access to physical infrastructure, while Cyangugu and Ruhengeri made the biggest improvements in coverage rates thanks to projects such as Rural Water Supply and Sanitation Project (PEAMR), Water Environment and Sanitation (WES), and World Health Organization (WHO). The overall physical improvements in infrastructure in rural areas in 2005 are listed in table 5.99. Table 5.99. Rural Infrastructure Improvements in 2005 Rehabilitation of 25 water and sanitation systems Construction of 59 forages Construction of 120 water reservoirs Construction of 600 public latrines Management of 1,800 water resources Source: WSS PER Report, 2006. These physical infrastructure improvements in 2005 covered an additional 1 million inhabitants, and hence, access to potable water increased to 55 percent in 2005 198 World Bank Working Paper from 41 percent in 2002 for the rural population, and to 69 percent in 2005 from 66 percent in 2002 for the urban population.56 Table 5.100 below shows how performance and outcomes are aligned with actual resources spent in the sector. Table 5.100. Water and Sanitation Sector Actual Expenditure Compared with Outcomes in 2003­05 In Real Terms 2003 2004 2005 TOTAL Total Expenditure of Sector (in Millions RwF) 8,464 10,023 25,144 43,632 Additional Population Served in Rural Areas 200,000 167,000 598,000 965,000 Additional Population Served in Urban Areas n.a. n.a. 95,700 95,700 TOTAL ADD'L POPULATION SERVED 200,000 167,000 693,700 1,060,700 Source: WSS PER Report, 2006, page 37. Original Source: MINITERRE. Between 2003 and 2005, an additional 965,000 persons in rural areas were served and an additional 95,700 persons in urban areas--an additional 1.1 million people overall. With a total resource allocation of RwF 43.6 billion, this expenditure translated into a cost of RwF 41,135 for each additional person served. More recently, the government of Rwanda reported further progress in the WSS sector (not yet reflected in the WSS 2006 PER). These results included a 6.8 percent improvement in access to clean water between 2006 and 2007, bringing to 71 percent the total population with access to safe drinking water. Also, an additional 636,766 people had access to safe water through the construction of more than 1,000 new water points by the end of 2007. Rwanda was able to increase access to rural water for an average of 600,000 people since 2005. This is a steady and outstanding performance. The private sector participation in the operation of the piped, rural water supply systems continues to be a success, of which the PRSGsupported Rural Water Supply and Sanitation Project (RWSSP), led by the World Bank as a publicprivate partnership program, serves as an example of success. Under the RWSSP an additional 20,000 citizens gained access to clean water in the first part of 2008 following the completion of infrastructure construction by the end of 2007. As of April 2008, 28.9 percent of systems were already privately managed and the portion of privately managed systems is expected to continue to increase in the near future. Also, in terms of water resource management, the target of at least 40 percent of the springs protected in accordance with national standards is on track. Spring protection is implemented by the districts. Unfortunately, the 2009 target of 80 percent was overestimated and required revisions. Additionally, the monograph of the Kagera Basin was completed in October 2007. This monograph was the first step in achieving a water resources management target of an investment plan for the Kagera (large and small scale) in the Nyabarango and Muvumba Basins. In the area of sanitation, further progress can be noted as well. In 2006, 85 percent of the population had traditional household sanitation facilities, but only 38 percent of them were considered by the 2004 UNICEF/WHO Joint Monitoring Program report as hygienic onsite sanitation. All district performance contracts signed in 2006 committed to expand their respective household sanitation coverage to 100 percent. In 2007, newly released data suggested Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 199 that 42 percent of the household sanitation had been achieved against an EDPRS target of 45 percent. Water and Sanitation Challenges The government has shown increased commitment to investment in infrastructure and is planning on expanding such investments to ensure economic and private sector growth. In the future, the share of infrastructure in public expenditure needs to be increased to foster growth. However, during the JBSR in September 2007, the Water and Sanitation Ministry reported a decline in the WSS sector allocation, despite the government's commitment to increase funding for this sector. This apparent decrease was because some externally financed projects were not included. However, overall the allocations increased significantly after 2003. Furthermore, some development partners expressed concern that spending in the urban centers may be receiving preference over the rural regions. While this disparity was expected to be corrected by 2009, measures were taken to address the issues immediately. The government indicated that the relative increase in urban spending was the result of the Kigali Water Project, which was then anticipated to be finished in 2008, at which point the funding would become more even. Also, concern was raised regarding the consistency of sanitation indicators used by the government. In part this was because the National Institute of Statistics of Rwanda (NISR) was using different indicators than the ministry, whose indicators were more nuanced, and refinements of indicators are ongoing. Also, it was suggested that affordability, rather than physical access, was becoming the primary barrier to access to clean drinking water in regions where infrastructure was being erected. For example, the increase in the number of people accessing clean water between 2000 and 2005 was lower than expected because of cost, despite the increase in facilities. There had previously been some discussion that poorer members of communities were going to receive vouchers to ensure that they receive access to water, and the status of this needs to be followed up. The ministry has noted that a publicprivatepartnership system is being attempted as an alternative to the civil society­based system that was used following the erection of new facilities and that proved to be unsatisfactory. Assuming this management system is successful, donors emphasized that a propoor subsidy system must be in place to ensure universal access where facilities exist. Another key challenge for the ministry is to ensure proper investments in the sector and to continue to work toward attracting investors. Lastly, the Water and Sanitation Ministry is moving from MINITERRE to MININFRA, and this transition may also pose some unforeseen challenges. Energy For nearly 10 years since the tragic events of 1994, Rwanda's development partners had paid scant attention to the energy sector, except for limited, emergency rehabilitation of the electricity distribution network damaged by war; negotiations over a management contract for Electrogaz; and establishment of a regulatory authority. The sector was starved of investments and even proper maintenance. Ministerial (energy, commerce, mining) and Electrogaz operating capabilities suffered because of the loss of manpower throughout the skill hierarchies and insufficient external assistance. The processes of planning, budgeting, auditing, and decision making were severely eroded 200 World Bank Working Paper as a result. In retrospect, it is remarkable--and credit to those who were engaged in sector operations during the period--that the lights stayed on and petroleum supplies were not disrupted. Against that backdrop, Rwanda was the first SubSaharan country to engage its development partners in the energy sector via budget support operations (rather than only via projects, as was the rule and still remains the rule in most lowincome countries). Budget support aid--rather than conventional project aid--enabled Rwanda to (a) weather the combined crises of power shortages and high world oil prices, via a variety of tax, subsidy, and tariff instruments; and (b) begin the process of strengthening sector institutions, building human capacity, and engaging a wider network of aid partners. Even so, the macroeconomic damage of the combined energy crises is not trivial, mounting perhaps to more than 10 percent of GDP loss cumulative over five years, significantly eroding the benefits of debt reduction and adding to budgetary stresses and forced shifts in priorities. Sector scope:57 From the viewpoint of public budgeting, the energy sector consists almost entirely of (a) production, acquisition, and distribution of grid electricity; (b) acquisition, storage, and distribution of petroleum products; and (c) promotion of corporate investments in fuel resource exploration, development, and use. Government budgetary involvement in the supply chains for traditional biomass cooking fuels is marginal, both as a share of the total expenditures in that subsector and as the portion of the supply chain (for example, limited to reforestation activities on public lands). Scope of government involvement: With some exceptions, energy expenditures are generally consumer financed, with public expenditures meeting the investment financing gaps and associated risks of international borrowing. Ideally, government budgetary involvement in the electricity and oil products supply systems should be limited at best--and only when necessary--to investment or debt guarantees. The bulk of government attention, usually at small budgetary cost, should go toward formulation and enforcement of regulatory structures for corporate entities, not unlike other regulations on corporate finance and general commercial activities. This is so because distribution of energy is a commercial activity, and consumer expenditures are generally expected to meet the cost of production and delivery, as is nearly always the case with the traditional woodfuels supplies, mostly to the poor and the middle class. As in most developing countries, however, the situation is less than ideal in Rwanda. Premium energy sources--grid electricity and oil products--that serve only a tiny fraction of the population--about 7 percent, mostly in urban areas--have needed government financial support because of (a) critical dependence of modern economic growth on reliable supplies of highquality modern energy products, in particular electricity; and (b) the transformative role of electrical lighting and appliances in the delivery of human development services (health, education) provided or financed by the government. In each of these cases, public investments in expanding access and maintaining reliability are legitimized by the enabling role of modern energy, on the one hand, and market failures in attracting private investments, on the other. In Rwanda, failures in aid design and administration also added to the urgency of public expenditures in the sector. Types of government budgetary involvement: Unlike other sectors, government budgetary involvement in the energy sector involves tax revenues from the sector, Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 201 portfolio investments in corporate entities, operating subsidies, own expenditures for energy use or appliance acquisition, and general expenditures on administration and regulation (see table 5.101). The types of budgetary items, with reporting ministry name in parentheses, include the following: Tax revenues from energy imports and sales: import duties, VAT, and other charges on oil products; VAT on electricity sales; duty and tax exemption for Electrogaz oil products purchases (MINECOFIN). Operating subsidies to Electrogaz: direct cash subventions or payments for rental and fuel expense for thirdparty generators (MINECOFIN until 2006; MININFRA beginning 2007 for a portion). Capital subsidies to Electrogaz and joint ventures: in turn, equity injections (cash, debt capitalization, or asset purchase on behalf of the companies), debt guarantees, debtforequity swaps, and debt service in case of failure by those companies (MININFRA and MINECOFIN). Own expenditures on energy (electricity, oil products, biofuels) and energy using appliances or vehicles (individual ministries). Administration and regulatory activities (MININFRA, except that the Rwanda Utilities Regulatory Agency (RURA) budget in 2005­06 may have been under MINECOFIN). While energy taxes are a significant share of the total revenues and can be said to offset the impact of higher electricity and oil products prices on the government's own consumption--a sizeable portion of the total electricity and oil products market--the following discussion neglects energy tax revenues as well as expenditures for the government's own use, and reports data on public expenditures on electricity provision, administration and regulation, research and development, and special programs. Role of general budget support: Budget support aid could not have come at a more pressing time. Beginning just around the time the government of Rwanda began negotiations with the World Bank on the first PRSC/G series (December 2003), the country was caught in a major power shortage, a combined result of neglected investments in generation and network, steady demand growth, and low hydroelectricity reservoir levels (resulting, in turn, from poor rainfall and overwithdrawals). This crisis was met by purchase of emergency generators and lease contract for additional generators, all relying on diesel fuel. Following the energy crisis in 2003, world oil prices also began their upward climb. As a landlocked country far from the coast and without railway or pipeline access, Rwanda was perhaps the first SubSaharan African nation to be doubly hit by increasing use of diesel for power and a higher price for it. The government boldly increased the electricity tariffs in two stages--from RwF 42 to RwF 112 per kWh--in rapid succession, after which it chose to stabilize the tariff and absorb the additional operating costs of diesel generation. Thus, while consumers paid for most of the sharply higher cost of supply, the government covered about 25 percent of the total. 202 World Bank Working Paper Table 5.101. Public Budget and the Energy Sector, Rwanda 2001­07 In RwF millions 2001 2002 2003 2004 2005 2006 2007 2005­07 A. Contributions to Fiscal Revenues Liquid Fuels: Import Duty 3,163 3,625 4,563 4,973 4,851 4,960 4,438 14,249 Consumption Tax (some fuels) 5,667 5,920 6,916 7,550 7,195 7,090 6,204 20,489 VAT 3,661 4,317 5,262 5,707 5,590 5,832 5,147 16,569 Electricity: VAT 1,013 1,393 1,361 1,533 2,427 4,002 4,974 11,403 Electrogaz Corporate Tax -- -- 1,772 -- -- -- 0 Subtotal 13,504 15,255 19,874 19,763 20,063 21,884 20,763 62,710 B. Main Energy-Related Expenditures Road Maintenance Fund ­2,000 ­2,000 ­2,000 2,677 5,866 7,770 9,891 23,527 MININFRA Energy Recurrent budget 126 150 380 248 778 MININFRA: Energy Development 237 455 232 5,900 21,748 27,648 Budget MINITERRE Forest 937 1,515 2,965 5,417 Protection/Replanting C. Assistance in Overcoming 2003 Crisis Waive tax Electrogaz thermal fuel -- -- -- -- 1,678 4,475 5,658 11,811 imports Buy 7.5 MW Jabana diesel units -- -- -- -- 2,016 -- -- 2,016 Rent 15MW diesel units from -- -- -- -- -- 2,295 2,887 5,182 Agrekko Fuel for 5MW this capacity -- -- -- -- -- 2,141 4,545 6,686 (Mukangwa) Waive import tax high-efficiency -- -- -- -- -- -- 500 500 bulbs Subtotal 2,000 2,237 2,455 3,035 10,647 24,476 48,442 83,565 Source: Gaspard Ayibushije and Christopher Willoughby, Rwanda Energy Sector Public Expenditure Review Draft 6, March 2008. Note: Development expenditures under externally financed projects are not shown. Also not shown are equity and quasiequity to Kibuye Power Limited. Excluding the value of import duties and the tax exemption granted for fuel for supplies to the Electrogaz grid, the government subsidized Electrogaz tariffs by about US$25 million between 2005 and 2007 (about RwF 14 billion for purchase, lease, and partial fuel); and in 2008 the government was projecting subsidize about US$10­15 million more, depending on the evolution of oil prices. The government also moderated tariffs by a renegotiation of the sales price of Sinelac hydroelectricity to the utilities in Rwanda, Burundi, and the DRC. While this has no budgetary impact as such, financial restructuring of Sinelac, and settlement of crossdebts among the three countries' governments and utilities, still remains to be completed. In retrospect, GBS enabled the government to take emergency actions that would have been extremely difficult, if not impossible, under the usual project approach, considering the limited capacity to handle donor project procedures and the tightness in world power equipment markets. Indeed, the IDA Urgent Electricity Rehabilitation Project, negotiated in October 2004 and made effective in mid2005, faced delays in its planned capacity addition of a heavy fuel oil powered generator. The government and Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 203 Electrogaz has also held negotiations for the development of additional generating capacity for the Electrogaz grid--principally, development of Lake Kivu methane for power production (jointly with private investors), and construction of Rukarara and Nyabarongo hydro plants--as well as expansion of the grid. GBS also permitted the government to undertake some investments in the areas of nonconventional energy--for example, installation of microhydrobased rural independent grids, biogas plants, and initiation of biogas and improved cookstoves programs for households. These initiatives demonstrate the government's appreciation of the problems of the rural poor otherwise excluded from most of the activities afforded by modern energy (other than, say, motorized public transport). Table 5.102. Ameliorating Shortages, Expanding Geographic Access Generating capacity for the Electrogaz grid: Placed in service 2004­07: o Electrogaz Gatsata and Jabana diesel generators ­ grant-financed by government of Rwanda o Rental generators ­ lease payments and partial fuel payments grant-financed by government of Rwanda Anticipated 2008­10: Electrogaz-owned: o Heavy fuel oil generator 20 MW ­ IDA credit o Rukarara hydro 9.5 MW ­ government of Rwanda o Nyabarango hydro 27.5 MW ­ government of Rwanda Independent generators: o Kibuye Power Kivu gas pilot plant 4.5 MW ­ joint venture o REC Kivu gas 7.5 MW ­ joint venture o Additional under planning (~70 MW) Generating capacity for independent grids (microhydro) Placed in service 2004­07: (precise information not available) Anticipated 2008­10: o More than 20 stations and grids, cumulative > 10 MW Table 5.103. Initiatives in Nonconventional Energy Via Central Government o Biogas program (15,000 digesters; 14 m EUR) Via Electrogaz o Grid-connected solar PV plant o Distribution of energy-saving lamps Via District Governments o Household biogas o Improved cook-stoves 204 World Bank Working Paper Results and Impacts The main results from the GBS to the energy sector can be separated into two broad areas: realization of physical installation targets, and gradual movement toward institutional restructuring and reform. Table 5.104. shows supplies via the Electrogaz grid for the period 200307. Table 5.104. Electrogaz Supply/Demand Balance 2003 2004 2005 2006 2007 Supplies (million kWh) Hydro (domestic) 119 82 66 39 37 Thermal (oil) -- 12 50 140 153 Domestic gas (Kivu - pilot + KP1) Solar 1 Total domestic 119 94 116 178 190 Imports (hydro) 123 129 87 62 61 of which, SNEL 3 16 21 21 20 Sinelac 114 113 65 40 41 Total available 242 222 204 240 252 Estimate of unconstrained reqmts. 259 280 302 402 Oil thermal as % of total 5% 25% 58% 61% Total sales 177 163 157 186 192 Growth 14% ­8% ­3% 18% 3% Line losses (tech, nontech) 65 59 46 54 n.a. as % of available 27% 27% 23% 23% Value of power imports US$ millions 9 10.2 4 2 2 Fuel requirements (m liters) <1 13 37 n.a. Value of fuel imports, US$ millions < $1.0 10 27 n.a. Average cost of supply US$/kWh 0.08 0.14 0.20 Source: 2001­06 Electrogaz data from previous World Bank reviews; 2007­10 from Electrogaz financial model Dec 2007 and Quarter III 2007 report to MININFRA. As regards institutional restructuring and reforms, progress has been slow. The Electrogaz board acted swiftly in renegotiating the management contract for Electrogaz with a consulting company and appointed a replacement director general in January 2006. In retrospect, it appears that Electrogaz's situation had become too dire for a quick turnaround by three expatriate staff under the management contract. The board also approved an increase in Electrogaz water tariffs in late 2006, which not only had a positive impact on Electrogaz's overall finances, but also reduced the crosssubsidy from electricity to water operations. By 2008, the government decided to separate electricity and water distribution operations and is considering unbundling the Electrogaz grid, permitting independent distributors to purchase bulk power and sell it to the public under small area concessions. In the area of petroleum products pricing, the government generally permitted passthrough of higher import costs to the final consumers, but sought to dampen the inflationary impact via moderating duties and Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 205 taxes. Because oil products consumption rose in spite of price increases, total tax revenues from oil products were not adversely affected until 2007, as seen in table 5.101. (Increased VAT from higher electricity sales made up for the drop in oil products tax revenue.) Impacts Because largescale grid electricity investments have long lead times, exacerbated in recent years because of the worldwide boom in the electricity supply industry, some of the impacts of major government development expenditures will not be manifest until after expanded, lowercost supplies begin to be fed into the Electrogaz grid. However, the most important contribution GBS made is in avoiding what could have been a macroeconomic catastrophe: major power shortages. Consumers take grid electricity for granted; only its absence reminds them of its importance. Additionally, moderation of tax rates on oil products allowed gradual adaptation to spiraling price increases in the world market, with limited impact on inflation, which was a worry given the electricity tariff increases. Looking Forward Recall that in the first phase of GBS for the energy sector, circumstances compelled the government to utilize budget support funds to subsidize emergency thermal power, but it was also able to invest in new generating capacity for the Electrogaz grid and initiate a remarkable program of microhydro independent grids and biomass energy interventions for the poor. Such a broadbased, comprehensive approach to public interventions has also attracted greater cooperation from development partners that had shied away from energy investments in the past. Looking forward, the government seeks to build on these pioneering practices via initiating an energy SWAP that would look at the investment and capacitybuilding needs throughout the sector and the delivery chains, and mobilize additional public as well as private financing. This too would be a pioneering effort to take the Harmonization and Alignment agenda to the next level, the first time in an infrastructure sector. Table 5.105 broadly outlines the implications of the Harmonization and Alignment agenda for the energy sector. Outstanding Challenges Table 5.106 shows the planned total energy sector MTEF for 2008­10. While it does not show the organizational responsibilities and conditions for fiscal transfer, especially when investing in longterm electricity supply system assets, it shows that the objectives are well defined: expansion of access to electricity, and arresting the environmental damage due to heavy reliance on biomass fuels. The main challenge is to reduce the sector dependence on public expenditures and attract greater private sector investments on competitive, commercial terms. 206 World Bank Working Paper Table 5.105. Implications of Paris Declaration Harmonization and Alignment Agenda for the Energy Sector Commitment Target Issue/challenge OWNERSHIP Partners have operational Energy development is not entirely within government budgetary control; sector development strategies. finances usually covered by customer payments. SWAP scope and procedures addressing private ownership and cost recovery yet to be defined and tested. Intergroup conflicts on scope, financing of changes. ALIGNMENT At least 80% of aid flows on No "national" investment strategy; investment strategies are defined by budget. corporations, public and private, in pursuit of their business goals. Progress on measures of Company systems and procedures subject to confidentiality rules and agreements procurement and public with lenders. Aid flows via country systems may be limited to subsidy payments and financial management small procurements. performance; % donors and aid flows via country systems. Capacity ­ 50% of technical Some technical cooperation activities highly specialized and funding targets for cooperation flows implemented "coordinated programs" may not work. through coordinated programs Partner countries: Ministries, regulatory agencies, and companies (public and consistent with national private) work on arm's length and occasionally adversarial basis; each needs to development strategies. have specific performance indicators and capacity-strengthening plan. No monolithic "national development strategies" and "country-led capacity Reduce by one-third the stock strengthening strategies" possible. of program implementation Donors: Capacity weaknesses and fragmentation at their own end, and inconsistent units (PIUs). policy biases (about sector reforms, least-cost planning). Harmonizing support for technical cooperation and capacity strengthening a difficult challenge. More predictable aid Lumpy investments with long gestation periods may not permit predictability of disbursement. project-based disbursements. Percentage of bilateral aid Exports of energy equipment and services of considerable, if declining, value to untied. major OECD donors; globalization has introduced greater competition in equipment markets, and fuel/electricity exports to developing countries are of negligible value to the traditional donors, whose export edge is now in renewable energy (RE) and energy efficiency (EE) equipment and services. OECD donor preference for RE/EE at risk in reducing tied aid, if applied at the sectoral level. Increasing presence of nontraditional donors (that is, outside of the Development Assistance Committee of the OECD)­for example, China, India, major oil producers. HARMONIZATION 66% of aid flows in the context Not applicable at the sectoral level, and not practical for large, lumpy investments. of program-based approaches. SWAPs yet to be defined; links with private sector finance and associated regulation make the numerical target meaningless. Government expenditures a significant but small share of the total consumer expenditures. Limited experience with use of explicit (nonproject) budgetary instruments for subsidizing noncommercial aspects of the supply chains­for example, expansion of rural access to modern energy. 40% joint missions, 66% joint Not applicable at the sectoral level, and not practical for energy sector capital country analytical work expenditures. Analytical work and missions by the private investors and financiers cannot be subject to such aims. Historical donor fragmentation (large vs. small; credits vs. grants; utility and mining vs. off-grid and renewables) and several bilateral donors' exit in the past decade make numerical targets difficult. MANAGING FOR One-third reduction in number Some quantitative targets relatively easy to adopt, but subject to operational and RESULTS of countries without financial performance targets at the corporate level and arm's-length regulation. transparent and monitorable Corporate targets aimed at bankability and financial risk management, not political performance assessment preferences necessarily. frameworks. Qualitative targets such as transparency of decision making, adequate provisions for fair dispute resolution, improvement of investment climate, and equitable burden sharing for social objectives difficult to define and monitor. MUTUAL Mutual assessment reviews in Some relatively easy to do because energy sector technical, management, and ACCOUNTABILITY place. financing practices have fairly common international norms, and most bilateral and multilateral donors are engaged in different parts of the sector. The entry of nontraditional donors (China, India, or private charities) and possible conflicts between country needs and some traditional donors' domestic or foreign policy goals (environment, human rights) may affect the comprehensiveness of mutual reviews. Project aid with sovereign debt or guarantees has its own financial and contractual covenants. Source: Data compiled by Nikhil Desai. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 207 Table 5.106. Energy Sector Medium-Term Expenditure Framework 2008­10 (in RwF millions) 2010 Total 2008-10 Increased Access to Electricity Development Budget Generation, of which hydro-electricity 22,891 75,271 methane gas 5,940 11,677 thermal power (heavy fuel oil) 21,340 other (solar, geothermal, etc.) 1,650 4,630 Transmission and distribution 19,065 59,062 Subtotal, increased access to electricity 49,546 171,980 All other 2,350 7,398 Total 51,896 179,378 Electricity investments as % of total 95% 96% Recurrent Budget 15,057 43,869 Source: Data compiled by Nikhil Desai. With the restoration of lake levels in mid2008 and anticipated additions of lower cost generating capacity (2009­10), it will be possible to move from a "crisis management" mode of operations to sustainable planning. Table 5.107 shows that the government intends to make up for the lost ground in the "soft" areas of planning, legal and regulatory framework, and institutional/human capacity development. Table 5.107. Key Activities Proposed for 2008­12 1. Investments in increased access to electricity · Additional generation capacity (37 MW hydro, 82 MW Kivu gas) · Studies on alternative fuels (geothermal, peat, biofuels) and pilot plants · Off-grid access plan and investments in independent grids · Fiscal framework for performance-based subsidies · Solar electrification of health centers, schools, and sector offices 2. Legal and regulatory framework · Financial restructuring of Sinelac, Electrogaz · Tariff study · Business plans for Electrogaz, RURA · Law separating water and electricity distribution · Regulations for renewable energy and energy efficiency · Performance contract for Electrogaz · Private management contracts for rural independent grids · Distribution liberalization (unbundling of Electrogaz, third-party access; tentative) · Human resource capacity plan · Establishment of a National Energy Development Agency (NEDA) with a performance contract between MININFRA and NEDA Sources: Presentations by Secretary General, MININFRA, Joint Budget Support Reviews. 208 World Bank Working Paper Background Note58 DISTINCT FEATURES OF ENERGY SECTOR FOR BUDGETING AND PROGRAMMATIC AID Public budgeting for the energy sector has some very different features from those in the social service ministries such as health and education, or administrative ministries such as foreign affairs, commerce, or agriculture. These derive, in part, from the physical nature of energyeconomic interactions and the supply of commodities to a wider variety of customers than just people--that is, productive enterprises and service sector institutions. With the exception of the government's own use of energy and investments in energy production for its own use (generators, solar systems), the transmission of fiscal flows are generally via enterprises; governments do not engage in direct provision of energy commodities. Companies in the electricity or oil/gas supply chains are the primary channels for taxes and subsidies; direct subsidies to households or other consumers are rare. These enterprises, even if fully or partially state owned, are subject to corporate legal, management, and finance rules; investments are on the balance sheet, for instance, and are generally expected to be recovered over time from customers. Also, revenue and subsidy flows--and the government's own expenditures--are overwhelmingly in the modern fuels: fossil fuels and electricity. From a public finance perspective, traditional biomass fuels (wood, charcoal, dungcakes) are usually of small concern; they neither yield much revenue nor take much subsidy or public investments, if any, nor are they subject to much economic regulation, although government expenditures for its own use of biomass can sometimes be significant.59 Public expenditures in the energy sector take a variety of forms: support (equity, sovereign debt, tax incentives) for longterm supply investments with public, private, or mixed ownership; recurrent expenditures in own use and administration and regulation of the sector; onetime or recurrent subsidies and tax incentives. Energy production and sales, and trade in the equipment for energy production and use, are also significant sources of government revenue. The quality and predictability of regulatory decisions matter; in other words, results are measured not only against not how much a government was able to spend but also what it enabled others to spend-- investors in profitable opportunities, customers in preferred appliances. Local and international private sector involvement is significant, especially in nonelectric fuels (oil, gas, coal, biomass) and equipment (vehicles, electrical pumps and motors, lighting, electronics) supply chains, and it is generally capable of expansion. Since energy supply investments are capital intensive and long term and can easily crowd out other public investments, public expenditures on improving the private investment climate--regulatory capacity and associated "knowledge systems" that can facilitate mobilization of private finance--are key to commercial functioning of the sector. Therefore, the effectiveness of public expenditures is to be measured more in the quality and predictability of regulatory decision making, and their influences on public or private corporate entities, than in terms of budget execution and incidence of household benefits. The distinctive features of the energy sector from a public finance viewpoint can be described as follows. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 209 From a growth and budgeting perspective-- Intermediate as well as final consumption--growth and welfare impacts: Most energy forms are used for final household consumption as well as intermediates in industry, transport, service sectors (and even intermediates in transformation to secondary fuels). Indeed, for modern energy forms, household (including personal vehicle transport) final demand rarely accounts for more than a third of the total. The intermediate use is more important, and supply disruptions for those users can entail significant losses in growth and tax revenue since a high share of direct tax revenues--income tax or tax on goods and services--comes from entities that use oil products and electricity. Revenue potential: Energy is often a net revenue earner for governments, and not just for energy exporters. Energy tax revenues come from oil products (import duties, VAT, other taxes) and VAT on electricity sales, although governments frequently subsidize electricity prices. Cost recovery and corporate management: Recurrent energy expenditures in the economy are by and large met by consumers, and the role of public expenditures is relatively small (usually for government's own use). Government support to investments often takes the form of borrowing from aid agencies under a sovereign guarantee and onlending to parastatals that are supposed to repay with return on capital (that is, financed by the consumers). Limited geographic/demographic reach: Energy investment expenditures in the economy have a large public sector role in a very narrow market, namely, grid electricity provision by the public utility. Most other energy investment expenditures--such as in supply infrastructure for oil products or wood/charcoal, and generally in all energy enduse equipment ranging from vehicles to pumps, motors, and lighting systems to stoves and boilers--are made by the private sector (including households). Government expenditures for own use significant: Governments' own purchases of electricity and oil products (including fuel for selfgeneration in some cases) are a significant share of their total budget as well as of the overall markets for these fuels. Governments also occasionally procure electricity production equipment (generators, solar PV systems) and of course are significant buyers of energy enduse equipment as well (vehicles, lighting systems, airconditioning equipment, water heaters, cooking appliances). Facilitative and regulatory role: In addition to the budgetary instruments, the government also has at its disposal pricing policy instruments, nonprice regulation, knowledge development and dissemination, and energy investment promotion assistance. Portions of the energy sector that are in private hands--for example, oil marketing companies, independent power producers, producers and sellers of biomass fuels--can be influenced by indirect measures rather than budgetary expenditures. Because shortrun price elasticities of demand are generally low, it is difficult to respond to unexpected demand surges by use of pricing instruments alone. Because shortages and outages carry high opportunity costs, reliability requires that 210 World Bank Working Paper supply investments are made in a timely fashion. In turn, this implies that supply agents need to be financially viable and "bankable." In other words, price regulation, where applied, needs to ensure sufficient revenues to provide for reliable supplies and spare the consumers the high cost of shortages; years of financial gain from subsidies can be wiped out by a few weeks of unreliable or lowquality supplies. Finally, because electricity production costs can vary considerably depending on primary fuel potentials and because electricity trade is limited and can be costly, even at relatively short distances, electricity prices and reliability can become significant considerations in many investment decisions; hence, governments' general investment promotion and regulatory strategies (that is, apart from those in the energy sector) may also need to take into account their energy regulatory practices. Conversely, energy regulatory practices need to be tailored according to the varieties of supply models available and adaptable. From an industrial policy and planning perspective-- Industrial, tradable commodities for mass use, versatile and substitutable: Outputs in the energy sector are commodities whose consumption can be volumetrically measured at the customer level. Like water, individual fuels are generally uniform, with tight, internationally common quality standards and minor variations at most (important because unacceptable quality can damage the enduse equipment). Not only are some of these products extremely versatile in application, there is a considerable degree of substitutability among fuels, including for electricity generation, so long as the necessary equipment to use the alternate fuel is available and among supplier sources for primary fuels (sometimes even electricity). This leads to a considerable potential for competition among fuels and supply sources (adjusted for transport costs), which is largely the driver of technological change and management gains from private sector participation. Versatility in application and substitutability in transformation or final use, in turn, depend on a great diversity in types of equipment for energy production and use. Tradability of fuel commodities and equipment: In turn, the markets for fuels and production/use equipment are generally international in scope, with varying degrees of competition at different points in the supply chains. Logistical requirements for fuels can be quite different and involve high capital costs, for example, electrical grids, pipelines, and storage facilities. In as much as oil products are overwhelmingly the fuel of choice for transport, and in other markets the "fuels of last resort," world oil market developments reverberate through the supply chains. Service on demand and network/logistical chains: Products such as grid electricity, piped or bottled gases, and transport fuels are expected to be served on demand. Grid electricity is expected to be available instantaneously, since it cannot be stored (except for pumped hydro), and storage in DC power or other energy forms is costly and extremely limited. For most other fuels, transport and inventory costs are high depending on distance and energy density, and investments in the transport networks are capital intensive. Yet, a Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 211 certain degree of redundancy and excess capacity is built into the supply networks to be able to provide service on demand. Above all, many power and oil/gas sector assets have to meet exacting operational performance standards simply because of the premium on ondemand, uninterrupted delivery; a school or a hospital can shut down for a day, a bridge can be washed out for a month, but an entire power system just cannot be allowed to shut down. Utilities and petrol pumps cannot have vacations. Supply investments lumpy, with long lead times: Most investments in electricity generation (including hydro) and mineral fuels (fossil or nuclear fuels and geothermal) require a relatively long time for planning--from resource assessments to procurement documents--and implementation. These investments are "lumpy," that is, increments are large and noncontinuous, and economic returns on them depend on optimal capacity utilization. Dependence on external knowledge and finance: Energy sector assets-- power plants, networks, oil/gas wells, pipelines, refineries--and their operations tend to be technically complex and have to meet challenging, changing international standards, especially when operations are internationally linked (for example, grid interconnections, pipelines). Investments in these assets take complex financing arrangements and risk management options. These financing arrangements, whether public or private, and corporate management practices also tend to have international norms, and the markets for investment finance are also fairly competitive. ENERGY AND PUBLIC BUDGETING: TAXES, TARIFFS, SUBSIDIES For a variety of reasons and to varying extents, governments in SubSaharan Africa, as governments elsewhere, have been significant players in energy markets.60 They have invested heavily in the supply chains for modern energy sources--primary fuels such as oil, gas, coal, hydro or geothermal energy, as well as secondary fuels such as oil products and electricity--in the last 40 or so years. In some cases, foreign private sector has also contributed to these investments, either under a joint venture with governments or via other contracting procedures for revenue. Grid electricity distribution and in some cases generation and transmission as well typically have been in the hands of public sector corporations, occasionally under multicountry agreements. While popular access to oil products and electricity has been very limited, mostly in urban areas and growing only slowly (mostly from new customers in urban areas), reliable supplies of these fuels is critical for economic growth in general and functioning of vital infrastructure networks in particular. From the public finance point of view, supply shocks--shortages, and even sharp increases in prices--carry the risk of limiting economic growth and hence budgetary revenues. Therefore, investments in modern energy supplies have traditionally commanded prime attention from governments and their aid partners. Public investments in the energy sector are often a high share of the longterm investments in a country, and the energy companies are among the biggest corporate entities. These investments usually have been financed by concessional credits and loans from multilateral and bilateral aid agencies for repayment over a very long 212 World Bank Working Paper period (30 to 50 years) carried by sovereign guarantees. Such terms have been deemed necessary because most of the productive assets have a long economic life and slow amortization and debtservice schedules. Such debt was extended to public sector companies that often have a weak equity base. As owners and lenders, governments have tended to exercise substantial direct or indirect control on corporate decision making in the energy sector--licensing and intermediate/enduser tariffs and retail prices--and this influence has routinely curtailed efficient investments in the energy supply chains. Supply entities themselves have not generated sufficient finance and have typically suffered losses and provided little return on investments, and in some cases failed to service the debt. Public investments in the energy sector are geared at provision of commodities that serve both the household customers' demands for final consumption as well as business and institutional customers' demands for productive or incomeearning activities. Hence, energy supply is generally and properly regarded to be a commercially viable activity, because the product provided is mostly a private benefit and, importantly, the demand for it depends considerably on the customers' investments in enduse equipment and appliances. One implication is that energy suppliers have to meet customers' demand for energy commodities on a reliable basis; otherwise the customers' investments are underutilized (and they may have to make additional investments in alternatives), with adverse impacts on economic growth and stability, including public revenues from taxation of productive enterprises. Another implication is that the suppliers, in turn, can fulfill these expectations only if they are able to recover their costs and finance longerterm growth--meaning from customers' payments rather than the public treasury--on a reliable basis, since the planning and commissioning of energy supply assets is a technologically complex task and may take several months to years, depending on the asset type and size. Therefore, the broad general premise for a public finance review is that because nearly all of energy use is for the benefit of the final user61--that is, energy is not a public good--and, in turn, because such use can be measured and charged volumetrically at the individual level, customers should be expected to pay for their benefits, and these revenues ought to be adequate for the provider to cover the cost of service. The primary objective of public policy should be to facilitate the operation of supply agents--mostly corporate entities, with varying degrees of public or private investments--for reliable, leastcost supplies, and public expenditures should ideally be limited to addressing market failures. This can take the following five principal forms: Support to supply agents for commercially unviable investments in access expansion to under or unserved populations. Partial risk guarantees and transaction support for other private sector or joint venture investments (that is, avoiding additional public equity and fully guaranteed balance sheet debt) for optimal risk allocation (private and public). Research, development, and demonstration for the adaptation or development of promising technologies. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 213 Planning, policy making, reviews, and public accounting of energy budgets (including tax revenues and own expenditures on energy use as well as regulatory agencies). Information systems and capacity building for all public budgeting and regulatory, legal, environmental, and health and safety affairs. This ideal may not be met for a long time, but it is important to recognize that expenditures on subsidies for operating costs or commercial parts of capital investments have an inherent urban and prorich bias, and the opportunity costs of underinvestment are very high (leading to supply shortages or continued deprivation from access). To the extent that recurrent or capital subsidies are required because of inadequate tariffs, highcost technical standards, and inefficient operations, regulatory direction should be toward ameliorating those weaknesses and reforming the tax/subsidy regimes. These requirements for subsidies, marketenabling activities, and broad parameters of regulation policy need to be articulated and budgeted for, with justifications for incremental expenditures and a process for reporting and accountability. Government Revenues from the Energy Sector Relative to total tax revenues, SubSaharan African governments also derive significant revenues from taxation of energy commodities and that of equipment for energy supply and use. Relatively little attention was paid to the intended purposes and impacts of such taxation on household expenditures or costs of production, but a proper analysis of the role of the state in sector finances should include revenues from the energy sector as well. Parastatal Finances Public finance reviews in the energy sector must also address finances of corporate entities with public participation. Unlike other sectors in government budgeting, capital expenditures on energy (usually electricity, and oil/gas when there are state owned or joint venture enterprises in oil/gas production, processing, or distribution) are usually done on a balancesheet basis and occasionally on a nonrecourse or limited recourse project finance basis. Treatment of government equity, debt, or guarantees, and rules for tariff setting, in turn, determine how these expenditures are recovered from customer revenues--that is, returns on equity, repayment of sovereign guaranteed or other debt, and premiums for guarantees. Because some of the impacts of longlife investments are apparent only in the long term, it is difficult to establish a shortterm correspondence between government expenditures and results in terms of improvements in service quality. Quite often governments have suffered losses on their equity and guarantees in electric utilities because of inadequate tariffs, inadequate collection, or poor capacity utilization. Further, because electricity supply costs vary according to generation source, distance, and accounting practices, whereas tariffs are often averaged over regions with implicit crosssubsidies across different consumer groups, electric utilities are quite possibly the most opaque means of hiding intergenerational and intergroup equities. That is, it is next to impossible to assign benefits of public expenditures to particular classes of 214 World Bank Working Paper households (except to the extent that those who lack access to electricity are obviously denied any such benefits). Instruments of Energy Fiscal Policy Governments use a wide variety of fiscal instruments in the energy sector as shown below. Public financial involvement becomes more complex if in addition to government support to energy sector investments, other stateowned enterprises are major consumers or nonpayers of energy, or providers of equity or debt to energy sector institutions. a. Sources of revenue Import duties User fees VAT Taxes on resource extraction and exports Returns on public investments (stateowned enterprises, joint ventures) b. Longterm, usually largescale investments Electricity supply, including primary resources, or generation only Mineral fuels extraction (Frequently financed by external credits and a significant share of national debt) c. Recurrent expenditures Own use of electricity, gas, transport fuels Public water sector use of electricity Expenditures on energy sector management (energy ministry, agencies) Expenditures on related management (environmental and safety protection, water resource management, mining regulation) Financing of "national contributions" to donor projects d. Subsidies--recurrent or onetime Budgetary allocations to electric utilities for meeting operating costs Capital subsidies for access expansion Preferential tax treatment (Implicit) subsidy due to concessional loans, forgone returns on public equity, and assuming contingent liabilities (sovereign guarantees) e. Economic regulation (revenue/expenditure implications whereof) Investment approvals (and associated tax incentives, if any) Water use allocation (including international arrangements) Royalties, production sharing from resource extraction Price regulation at wholesale/retail levels f. Onetime emergency expenditures On accidents in the energy supply chains Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 215 While private investment in energy supply systems has grown in recent years-- primarily for nonutility power generation and oil/gas exploration and production--not only does the government's involvement remain extensive, energy shares relative to other public budget or trade items are also generally quite high. This is all the more relevant when considering that revenues from other sectors of the economy are also dependent on energy use and that expenditures on energy appliances-- road/air/water/rail vehicles, motors, pumps--are also significant and contribute to tax revenues (import duties, VAT). Typically, the following ratios are of interest from public finance view: Energy revenues to total budget revenues Government energy expenditures to the total budget Sales to government to total fuel or electricity sales Government capital subsidies--implicit in concessional finance and lowcost, or even negative cost, equity, and explicit in grant financing of investment--to total subsidies for longterm assets62 (particularly since denationalization waves since the 1990s for other stateowned enterprises) Government budgetary (recurrent) subsidies to total subsidies and to total budget Public equity and debt to energy companies to total public investments over time Public borrowing for the energy sector--at least up until the mid1990s--to the total external public debt Values of energy63 imports and exports64 to the respective totals Finally, where the government or public corporate entities--for example, water suppliers and public transport--do not pay for their purchases, and an energy supply company fails to service its publicly guaranteed debt, a complex web of crossdebt entanglements is created. Table 5.108 on page 216 illustrates a typology of public finance instruments and associated indicators of performance. 216 World Bank Working Paper Table 5.108. A Typology of Public Finance Instruments for Energy Public finance Related energy Result Overall objective Issue instrument policy instrument accountability Governmental Recurrent expenditure on Legislative and regulatory Annual performance planning, policy, and institutional and human development and audits, benchmarks; regulatory capacity resource development application collection and publication of statistics Reliability, quality specifications State equity Tariff setting, regulatory Returns on investment reviews State debt Debt service Public resources for Sovereign risk guarantees new investments Financing facilities (special Annual performance Reliable and secure funds) audits, benchmarks supplies of modern energy, prevent or moderate Tax incentives New private investment shocks Cost recovery, Conditionality for Cost, quality, performance Bankability of supply commercial equity/debt/subsidy benchmarks companies sustainability Tax rates, subsidies Reliability criteria, open Temporary relief, gradual Sudden supply crises, access storage for oil adjustment to market temporary price products prices increases in imported fuels Financial instruments for risk management Lack of competition, Liberalization, Investment Technical Assistance for New private Investment Overreliance on the assistance business development, fiscus esp. small-scale independent providers Commercially Capital subsidy Establishing subsidy Funds for rural grid unattractive capital criteria and administration electrification, retail Expand access to investments procedures systems for self-provision electricity and modern (solar, wind), supply fuels to promote broad- chain development based growth Commercially Reduction in recurrent Exception for small lifeline No recurrent subsidies unattractive sales subsidies tariffs if any, marginal cost growth pricing Improving the efficiency of Standards and technical Annual performance own expenditures assistance audits, benchmarks Reduction and/or redirection of Electricity access for public Standards and technical Annual performance Fiscal space for other operating social service institutions assistance audits, benchmarks priorities expenditures and Elimination of cross-debts Annual performance subsidies (own bills, public water audits, benchmarks supplies) Tax enforcement and Resource development Revenue management application revenues (production, framework for resource Public revenue exports of petroleum, taxation, incl. resource enhancement and electricity) sharing management General budget revenues Predictability, moderating (import duties, VAT) price shocks Resource use conflicts (land, water) Coordination with International Development of regional environmental, natural cooperation grid interconnections, resource management, (contracts, treaties) pipelines, storage facilities and foreign affairs objectives and operations Transport, municipal Budgeting for road water sector maintenance, energy needs implications of public water systems Source: Data compiled by Nikhil Desai. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 217 Transport Transportation Overview Transport costs in Rwanda are high compared with other regions because of its landlocked geographic location. The rolling terrain and heavy rainfall raises the cost of road construction and maintenance, and as a result Rwanda's transportation costs are some of the highest in Africa. It is estimated that more than 40 percent of the total value of imported or exported goods is transportationrelated costs, compared with 6.5 percent to 11 percent in developed countries. This high transportation cost has a strong negative impact on economic growth and development. Rwanda has a total road transportation network of about 14,000 km, of which 2,860 km are classified as national roads and 1,835 km as district roads. As road transportation is very expensive, the government continues to consider alternatives. Air transport consists of two international airports (Kigali and Kamembe) and five aerodromes spread across the country (Gisenyi, Butare, Ruhengeri, Nemba, and Gabiro). The only lake transportation available is on Lake Kivu. Currently, no railway system exists. Studies are underway or planned for largescale regional transportation projects to lower costs: IsakaKigali railway Lake and river transport on Lake Victoria and Akagera River New international airport at Bugesera Also to address the problem of high transportation costs, the government has initiated several transport facilitation initiatives, including the OneStop Border Post program, acquisition of more modern equipment, and reduction of nontariff barriers to help reduce import and export costs. As of 2007, 11 percent of national roads and 15 percent of district roads were considered in good condition. A key feature of the government's road program includes the Road Maintenance Fund. The sector is guided by a Transport Policy and a Road Maintenance Strategy. Routine maintenance guidelines in Kinyarwanda guide rural roads routine maintenance initiatives in the districts. Traditional arrangement structures with development partners (including the EU, AfDB the World Bank) and MININFRA were strengthened, and a partnership with NUR was initiated for road assessments, inspection, maintenance quality assurance, and capacity building. MININFRA has also benefited from additional capacity support from development partners in the form of technical assistance. Figure 5.12 shows the transport sector's transitional structure under its institutional framework. MININFRA is guided by several national investment guidelines, including the following: Vision 2020 National Investment Strategy 2002 ­ MINECOFIN/CEPEX LongTerm Investment Framework 2006­20 Concept Note for Public Investment Policy EDPRS document MTEF 218 World Bank Working Paper Figure 5.12. Transport Transitional Structure Secretary General MININFRA Planning, Policy and Capacity Building Unit (UPPR) Transport Coordinator Road Works Service Road Network Planning & Management Service Prospectives Service Road Investments Programming Road Maintenance and Road Network Transport Planning, and Management Management Other Transport Modes, Transport Facilitation Source: MININFRA, Transportation, JBSR VII, April 2008. Guiding principles and priorities for the transportation ministry includes selectivity, promoting publicprivate partnerships, decentralization, program approach/SWAP, macroeconomic equilibrium, and clarity (distinguish and identify recurrent costs and investment components). Furthermore, the ministry's decentralization framework is based on two key principles: Focus on financial decentralization and not on administrative decentralization. Take into account lack of capacity at the district level by building on progressive capacity improvement before transferring full responsibilities to minimize technical and financial risks. As part of the decentralization efforts, MININFRA is working to strengthen the reporting mechanism between itself and the local governments at the district level. Mechanisms in place include (a) quarterly reports to MINECOFIN with a copy to the Road Maintenance Fund (RMF) and to MININFRA, (b) RMF approval to district actions plans for new disbursement, (c) regular consultation between MINECOFIN and MININFRA to assess progress in EDPRS­Multiannual Plan (PPA20082012), (d) organized trainings and distribution of reporting format (including EDPRS indicators and disbursements), and (v) EDPRS quarterly reports. Transportation Expenditures The transportation sector has yet not completed a PER. Hence, the analysis of the sector is limited to a general overview as presented by the sector on an annual basis. Presentation of data in April 2008 at the JBSR in Kigali provides an update on expenditure allocations in the sector for the period ending December 2007. The following section is a summary of these expenditure allocations as presented at this venue in 2008. As can be seen in figure 5.13, the largest portion of the transportation budget in 2007 went to the construction of 40 km KicukiroNyamataMayange Road, representing 17 percent of the total budget allocations but 49 percent of total actual Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 219 spending. This represents an overexecution in 2007 of 295 percent. Budget allocations for this project in 2006 were 22 percent with actual spending reaching 37 percent of total sector spending, again showing an overexecution (of 170 percent). Figure 5.13. Key Budget Lines by Total Development Execution 2007 RRSSP: Rwanda Rehabilitation of Road Sector Support Gitarama-Ngororero Program: Ruhengeri Road (46.6km) Gisenyi Road 12% 15% Construction of Mayange-Nemba Road (20km) Construction of 14% Kicukiro-Nyamata- Mayange Road All others (40km) 10% 49% Source: MININFRA, Transportation JBSR VII, April 2008. Also figure 5.13 shows, the Rwanda Road Sector Support Program for the RuhengeriGisenyi Road represented 15 percent of total spending, up significantly from the previous year's budget allocations and limited execution rates. The third largest project in 2007 was the construction of the 20 km Mayange Nemba Road, a project not available in 2006. This program executed at only 71 percent. The fourth largest project in 2007 was the rehabilitation of the GitaramaNgororero Road of 46.6 km. This project executed at 98 percent. Combined, these four largest projects accounted for 90 percent of total transport sector spending. Table 5.109 shows transportation projects as of 2004. Overall, the sector has improved its execution rate from 130 percent in 2005 to 99 percent and 105 percent in 2006 and 2007, respectively. However, large variations in execution rates exist among projects, with some spending limited portions of their budget allocations while others, such as the KicukiroNyamataMayange Road project, significantly overspent. The IsakaKigali railroad project underexecuted in 2007 at only 37 percent and spent no funds in 2006, although only a small portion of the budget was allocated that year for the project still in its early stages. The construction of the Bugesera airport however, has seen execution rates of 112 percent and 100 percent in 2006 and 2007, respectively. In terms of recurrent budget spending in 2007, the budget was RwF 12.1 billion, but with an overexecution of 114 percent, the total recurrent spending for the sector reached RwF 13.8 billion. Overall, the various programs within the recurrent budget executed well, ranging from 100 percent to 119 percent. The programs included support to transport sector management, road network rehabilitation and asphalting, road network management, road network maintenance, and the road maintenance fund (RMF). Out of the total recurrent budget of RwF 13.8 billion, 75 percent was allocated for the RMF. Table 5.110 highlights the recurrent spending for the year 2007. 220 World Bank Working Paper Table 5.109. Transport Sector Development Expenditures by Program 2004­07 In million RwF, nominal terms 2004 Executed 2004 Original Revised* Actual* % PROJECT: Assistance Technique Direction Des Routes 39.2 n.a. n.a. n.a. PROJECT: Reparation Route Gitarama-Butare-Akanyaru 111KM 1,508.3 n.a. n.a. n.a. PROJECT: Rehabilitation Route Kigali-Kayonza 66 KM 1,891.8 n.a. n.a. n.a. PROJECT: Amen. Asphaltage Route Kicukiro-Nyamata-Nemba 58 KM 1,398.3 n.a. n.a. n.a. PROJECT: Reconstruction de la Route Cyangugu-Bugarama 1,081.4 n.a. n.a. n.a. PROJECT: Voiries Urbaines 12.6 KM Ville de Kigali 1,584.1 n.a. n.a. n.a. PROJECT: Expropriation Pour les Routes in Ville Kigali 300.0 n.a. n.a. n.a. PROJECT: Construction des Routes tra Estates Gikondo 350.0 n.a. n.a. n.a. PROJECT: Etude de Rehabilitation de la Route Kigali-Ruhengeri-Gisenyi 495.0 n.a. n.a. n.a. PROJECT: Reconstruction du Pont Gashora II 75.4 n.a. n.a. n.a. TOTAL - 2004: 8,723.5 0.0 0.0 0 2005 Executed 2005 Original Revised Actual % PROJECT: Etude du Terminal Rail a Isaka 40.0 40.0 4.1 10% PROJECT: Rehabilitation de la Ceinture des Voiries au Magerwa 100.0 100.0 0.0 0% PROJECT: PDST: Projet Developpement du Secteur de Transports 1,202.0 1,202.0 210.1 17% PROJECT: Etude de Rehabilitation de la Route Kigali-Ruhengeri-Gisenyi 357.5 357.5 127.7 36% PROJECT: Rehabilitation de L'Aeroport de Kanombe (Phase II) 5,000.0 5,000.0 7,228.9 145% PROJECT: Etude Construction de l'Aeroport de Bugesera 300.0 300.0 300.0 100% PROJECT: Reparation Route Gitarama-Butare-Akanyaru (111KM) 3,000.3 3,000.3 5,160.6 172% PROJECT: Rehabilitation Route Kigali-Kayonza 66 KM 1,390.0 1,390.0 7,111.3 512% PROJECT: Amen. Asphaltage Route Kicukiro-Nyamata-Nemba (58KM) 1,609.0 1,609.0 190.9 12% PROJECT: Infras. Routieres: Reconstr. Cyangugu-Bugarama 39KM; Pont Gashora 70KM; Route Cyangug 2,510.0 2,510.0 1,958.7 78% PROJECT: Voiries Urbaines 12.6 KM Ville de Kigali 1,506.0 1,506.0 0.0 0% PROJECT: Construction des Routes Tra Estates Gikondo 100.0 100.0 0.0 0% TOTAL - 2005: 17,114.8 17,114.8 22,292.4 130% 2006 Executed 2006 Original Revised Actual % PROJECT: Etude Chemin de Fer ISAKA-KIGALI 640.0 640.0 0.0 0% PROJECT: Etude Construction Aeroport de Bugesera 300.0 300.0 335.3 112% PROJECT: PDST: Projet de Developpement du Secteur des Transports 333.0 333.0 329.9 99% PROJECT: Construction Route Gitarama-Ngororero-Mukamira 2,371.0 2,371.0 166.7 7% PROJECT: Infrastructures Routieres: Reconstr., Cyangugu-Bugarama 39km, Pont Gashora 70km 2,431.0 2,431.0 6,590.1 271% PROJECT: Rehabilitation Route Ruhengeri-Gisenyi 3,073.1 3,073.1 0.0 0% PROJECT: Amenagement et Asphaltage Route Kicukiro-Nyamata-Nemba 58km 2,952.0 2,952.0 5,007.4 170% PROJECT: Constrction Route Kibuye-Ruganda 1,000.0 1,000.0 611.2 61% PROJECT: Appui Technique a la Direction des Routes 52.7 52.7 26.1 50% TOTAL - 2006: 13,152.8 13,152.8 13,066.7 99% 2007 Executed 2007 Original Revised Actual % 31 - PROJECT: PPSSP: Rwanda Road Sector Support Programme: Ruhengeri - Gisenyi Road 4,151.0 4,151.0 3,789.1 91% 06 - PROJECT: Bugesera Airport Project 450.0 450.0 450.0 100% 20 - PROJECT: Amenagement et Asphaltage de la Route Mayange-Nemba (20km) 4,946.0 4,946.0 3,492.9 71% 21 - PROJECT: Rehabilitation de la Route Gitarama - Ngororero (48.6KM) 3,074.0 3,074.0 3,000.1 98% 22 - PROJECT: Rehabilitation de la Route Ngororero - Mukamira (56KM) 2,055.0 2,055.0 0.0 0% 23 - PROJECT: Amenagement et Asphaltage de la Route Kicukiro - Nyamata - Mayange (40KM) 4,178.0 4,178.0 12,309.1 295% 24 - PROJECT: Appui Technique a la Direction des Routes 30.0 30.0 44.6 149% 25 - PROJECT: Etude Amenagement de la Route Kabarondo - Akagera Lodge 83.0 83.0 80.1 96% 27 - PROJECT: Infrastructures Routieres: Reconstr. Cyangugu - Bugarama 39KM, Pont Gashora... 1,650.0 1,650.0 716.1 43% 28 - PROJECT: ISAKA - KIGALI RAILWAY PROJECT 1,200.0 1,200.0 448.8 37% 30 - PROJECT: Preparation du Project de Developpement du Secteur de Transport 110.0 110.0 0.0 0% 32 - PROJECT: Project de Facilitation du Transport et du Commerce en Afrique de l'Est 1,210.0 1,210.0 320.1 26% 08 - PROJECT: Etude de Construction du Quai D'Accostage sur le Lac Kivu 80.0 80.0 39.0 49% 09 - PROJECT: Etude de Faisabilite de la Route Base - Byumba - Nyagatare (130KM) 297.0 297.0 297.0 100% 10 - PROJECT: Etude de Faisabilite de la Route Ntendezi - Nyamashake - Karongi - Rubavu (173KM) 340.0 340.0 317.6 93% 11 - PROJECT: Etude de Faisabilite de la Route Faisant Ceinture du Lac Muhazi 250.0 250.0 0.0 0% TOTAL 2007: 24,104.0 24,104.0 25,304.4 105% Source: Data compiled by Annika Kjellgren from MINECOFIN Budget Law. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 221 Table 5.110. Transport Sector Recurrent Expenditures by Program 2007 Portion of Budget Actual Execution Total Actual In million RwF Recurrent Recurrent Rate (%) Budget (%) Support to Transport Sector Management 576.7 574.1 100% 4% Road Network Rehabilitation and Asphalting 2130.0 2126.8 100% 15% Road Network Management 736.0 734.0 100% 5% Road Network Maintenance 12.6 12.6 100% 0% Road Maintenance Fund 8674.0 10333.5 119% 75% TOTAL 12129.2 13781.0 114% 100% Source: MININFRA, Transportation, JBSR VII, April 2008. Transportation Results During the JBSR in September 2007, a wide range of issues were raised, including the transport sector 2007 budget execution (both the development budget and the recurrent budget), the 2007 New Development in Transport Programs and Projects, the RMF, the MTEF 2008­12 and budget for 2008, and the EDPRS. Overall, progress on the main road projects was tangible, according to the government's own assessment, including the RuhengeriGisenyi Road rehabilitation, the contract for which was signed in May 2007. Most recent progress in transportation infrastructure includes road networks, rural gravel road networks, routine maintenance, and bridge networks. Table 5.111 provides the progress as of 2007. As can be seen, road rehabilitation currently involves four major road networks, all of which are in different stages of progress. Table 5.111. Road Network Progress as of 2007 · Kigali-Ruhengeri Road Rehabilitation: Financial agreements signed in Sept 08 (US$49 million from ACFG and IDA). Next steps: procurement process for works and supervision. · Ngorolero-Mukamira Road rehabilitation: Financial agreements signed in Nov 07(BADEA, OPEC, Kuwait Fund). Next steps: procurement process for works and supervision. · Butare-Cyangugu Road rehabilitation: AfDB appraisal mission. · Bugarama-Ntendezi-Kibuye-Gisenyi Road construction: Feasibility studies completed in June 2006. Source: MININFRA, Transportation, JBSR VII, April 2008. Furthermore, rural gravel network progress as of 2007 included the initiation of seven new projects, covering more than 200 km of rural gravel roads. Table 5.112 also includes a project planned for May 2008. Furthermore, technical studies were initiated to assess these rural gravel networks. These studies cover the following rural gravel networks, an additional approximately 400 km of rural gravel roads: 222 World Bank Working Paper Table 5.112. Rural Gravel Network Progress as of 2007 (ongoing maintenance contracts) Budget Project (in RwF billions) Start End Km Nyanza-Kibuye 1.95 12/2007 02/2009 69 Mudasomwa-Gisovu 1.46 11/2007 11/2008 52.5 Butare-Kibeho-Muse 2.23 11/2007 01/2009 53 Nyankora-Nasho 1.17 05/2008 09/2009 42 Kabarondo-Akagera-Ihema 1.66 05/2007 05/2008 36 Source: MININFRA, Transportation, JBSR VII, April 2008. KazabeGashubi (54 km) ByimanaBuhandaKitabi (98 km) CyakabiriNyabikenkeNdusu (75 km) GashirabwobaNyamirundi (30 km) NgoroleroVungaNyakinama (45 km) Muhazi Lake RingRoad (98 km) Other activities of the road subsector include the routine maintenance of the national 1,135 km paved network, including multiyear maintenance contracts (renewal on performance basis) to be signed with local contractors. Furthermore, progress in construction of bridges includes the reconstruction of Gashora II and III Bridges (ongoing), Rusizi Bridges (two in total) supported by technical studies within Economic Community of Great Lake Countries (CEPGL) context, and maintenance work on the Rusumo Bridge. The government has also assessed alternatives to its current transportation options. Rwanda currently does not have a railroad system but is working on the IsakaKigali Railway project. In the spring 2008, nine missions to assess the project had been completed, and seven of the following studies had been undertaken with drafts presented in April 2008: (a) preliminary tasks, (b) topographic and alignment studies, (c) engineering studies, (d) railway track studies, (e) electric and mechanical equipment studies, (f) environmental and social impact studies, (g) economics studies, (h) financial studies, and (i) legal and institutional framework analysis. Figure 5.14 shows the planned network for the assessed railway project. During 2008, progress was made on the oil pipeline between Kigali and Kampala. The following list highlights the next steps that were identified to move this project forward beyond 2008: March 19, 2008: Signing of memorandums of understandings between Government of Rwanda, Government of Uganda and TAMOIL Africa April­May 2008: Preparation of the technofeasibility study June­July 2008: Discussions of the first draft of the project agreement End of August 2008: Signing of the project agreement Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 223 Figure 5.14. Isaka-Kigali Railway Project, "Kigali ­ Dar es Salaam Rail Corridor" Source: MININFRA, Transportation, JBSR VII, April 2008. New key developments within the sector for the period include at least four key initiatives. First, MININFRA held training programs involving MININFRA and district technical staff to address issues of transport expenditures, planning, and regulation. Specialized training courses included road design, construction, maintenance, and management. Second, transport facilitation to simplify border crossing procedures through the OneStop Border Post Program was still ongoing, with the intent to encourage further import and export and, more important, lower existing export/import costs. Third, the ministry was working to upgrade its equipment. Utilization of modern and efficient equipment to facilitate transport included new cargo scanners and an electronic cargo tracking system along transport corridors. And finally, a regional integration program (including Transit Transport Coordination Authority [TTCA], Transit Transport Facilitation Agency [TTFA], Transit Transport Facilitation Agency [EAC], and CEPGL) was underway. Going forward, the progress made over the last several years will be an important baseline for the new poverty reduction strategy, the EDPRS. As can be seen in table 5.113, two key indicators relating to the transportation sector will help guide the efforts of the EDPRS and allow the government and other stakeholders to hold the sector accountable for key developments in its crucial road network. 224 World Bank Working Paper Table 5.113. 2006 Road Conditions Status Serving as a Baseline for EDPRS 2008­12 EDPRS EXPECTED PERFORMANCE EDPRS STRATEGIC STRATEGIC OUTCOMES Baseline Targets Targets Targets Target Target OBJECTIVES (Results) INDICATORS 2006 2008 2009 2010 2011 2012 Key Strategic Results 1.3 Economic % of Classified 11% 16% 19% 23% 27% 31% infrastructure National built Road network in good condition % of Classified 15% 22% 28% 36% 43% 50% District Road network in good condition Source: EDPRS Performance Matrix 20082012. Results and Challenges The government should envisage a borrowing strategy to ensure that the large projects are fully funded. Prior to that, the government needs to take into consideration the debt sustainability issue and define which of the projects are regarded as priorities. As of September 2007, the sector intended to complete a BIA as part of its SWAP in the sector. Concerns over financing the Bugesera Airport were noted, and as of September 2007 the government was in the final stages of the feasibility study. An ongoing challenge is the effort needed to seek longterm partnerships to close the financing gaps. The RMF budget was approved in 2007 by the board, and the agency is implementing the program agreed upon with MININFRA. Other issues that need to be addressed include institutional arrangements within MININFRA and its affiliated agencies (FER, RURA). MININFRA is attempting to strengthen its communication with districts to enhance dialogue and strengthen collaboration. Nonpriority Sectors Snapshots This section provides a snapshot of all the nonpriority sectors' data between 2004 and 2007 (in nominal terms). All data are in million RwF. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 225 Table 5.114. Snapshot--PARLEMENT--Office of Parliament (Million RwF) PARLIAMENT 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0101 Activites Legislatives - Chambre des Desputes 1,362.9 1,362.9 1,625.8 119% 0102 Controle de l'Action Gouvernementale - Chambre des Desputes 311.7 311.7 103.5 33% 0103 Service d'Appui - Chambre des Desputes 716.0 716.0 668.7 93% 0104 Activities Legislatives - Senat 555.9 555.9 606.7 109% 0105 Controle de l'Action Gouvernmentale - Senate 75.9 75.9 42.9 57% 0106 SERVICES D'APPUI-SENAT 446.7 446.7 279.7 63% 0107 Inspection, Controle, Audit et Organes de Transparence 860.8 860.8 770.5 90% TOTAL 2004 4,330.0 4,330.0 4,097.7 95% PARLIAMENT 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0101 Legislative Activities - Chambers of Deputies 1,489.4 1,489.4 1,389.8 93% 0102 Control of Government Actions - Chamber of Disputes 156.5 156.5 26.8 17% 0103 General Services and Planning 676.5 676.5 648.0 96% 0104 Management Activities - Senate 684.4 684.4 602.2 88% 0105 Control of Government - Senate 77.2 77.2 74.8 97% 0106 Support Services - Senate 472.1 472.1 348.1 74% 0107 INSPECTION , CONTROLE,AUDIT ET ORGANE DE TRANSPARENCE 846.9 846.9 836.4 99% 0108 RESPECT PRINCIPES FONDEMENTAUX ENONCES AUX ART 9&54 LA 34.6 34.6 35.6 103% CONSTITUTION TOTAL 2005 4,437.7 4,437.7 3,961.7 89% PARLIAMENT 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0201 PLANIFICATION ET COORDINATION DES ACTIVITES - CHAMBRE DES 2,235.1 2,232.4 1,875.6 84% DEPUTES 0202 DIPLOMATIE PARLEMENTAIRE - CHAMBRE DES DEPUTES 123.4 123.4 44.0 36% 0203 CONTROLE DE L'ACTION GOUVERNEMENTALE- CHAMBRE DES DEPUTES 196.2 196.2 6.4 3% 0204 ELABORATION ET VOTE DES LOIS - CHAMBRE DES DEPUTES 80.5 80.5 65.8 82% 0205 SERVICES D'APPUI - CHAMBRE DES DEPUTES 439.2 439.2 606.0 138% 0206 ACTIVITES LEGISLATIVES - SENAT 749.5 749.5 723.5 97% 0207 CONTOLE DE L'ACTION GOUVERNEMENTALE - SENAT 95.8 95.8 66.8 70% 0208 RESPECT PRINCIPES FONDEMENTAUX ENONCES AUX ART 9&54 DE LA 83.1 83.1 73.8 89% CONSTITUTION 0209 SERVICES D'APPUI-SENAT 465.2 481.0 417.5 87% 0210 INSPECTION , CONTROLE,AUDIT ET ORGANE DE TRANSPARENCE 947.1 875.9 826.8 94% TOTAL 2006 5,415.0 5,357.0 4,706.1 88% PARLIAMENT 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0201 PLANIFICATION ET COORDINATION DES ACTIVITES - CHAMBRE DES 2,059.0 2,093.1 0.0 0% DEPUTES 0202 DIPLOMATIE PARLEMENTAIRE - CHAMBRE DES DEPUTES 158.4 126.3 0.0 0% 0203 CONTROLE DE L'ACTION GOUVERNEMENTALE- CHAMBRE DES DEPUTES 177.8 134.3 0.0 0% 0204 ELABORATION ET VOTE DES LOIS - CHAMBRE DES DEPUTES 71.6 87.7 0.0 0% 0205 SERVICES D'APPUI - CHAMBRE DES DEPUTES 415.5 441.8 0.0 0% 0206 ACTIVITES LEGISLATIVES - SENAT 695.0 709.2 0.0 0% 0207 CONTOLE DE L'ACTION GOUVERNEMENTALE - SENAT 78.3 76.1 0.0 0% 0208 RESPECT PRINCIPES FONDEMENTAUX ENONCES AUX ART 9&54 DE LA 67.6 60.6 0.0 0% CONSTITUTION 0209 Diplomatie Parlementaire et Reseaux Parlementaires- Senat 118.2 112.2 0.0 0% 0210 Communication Senateurs - Population - Senat 49.5 49.5 0.0 0% 0211 Services d'Appue - SENAT 413.7 413.7 0.0 0% 0212 Inspection, Controle, Audit et Organes de Tranparence 1,019.4 1,019.4 0.0 0% TOTAL 2007 5,323.8 5,323.8 0.0 0% Source: MINECOFIN data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. 226 World Bank Working Paper Table 5.114 (continued) Source: MINECOFIN--data shown in nominal terms. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 227 Table 5.115. Snapshot--PRESIREP--Office of the President Prog S/Prog BUDGET BUDGET EXECUTION % 0201 Coordination et Services d'Appui 2,288.0 2,288.0 2,362.4 103% 0202 Services de Conseiller Technique 111.9 111.9 51.0 46% 0203 Coordination de Politique du Gouvernement 162.2 162.2 177.4 109% 0204 Palais Presidentiel 404.3 404.3 383.8 95% 0205 BONNE GOUVERNANCE 1,215.8 1,215.8 1,099.6 90% 0206 Securite des Bien et Des Services 1,995.1 1,995.1 2,009.1 101% 0207 Protocole 58.6 58.6 48.2 82% TOTAL 2004 6,235.8 6,235.8 6,131.5 98% PRESIREP 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0201 Coordination and Support Services 2,193.7 2,193.7 2,202.8 100% 0202 Technical Advice Services 114.8 114.8 35.6 31% 0203 Coordination of Government Policies 125.8 125.8 120.6 96% 0204 Presidential Palase 446.4 446.4 500.5 112% 0205 Good Governance 1,447.1 1,447.1 1,680.1 116% 0206 Security of Goods and Services 2,620.8 2,620.8 2,638.5 101% 0207 Protocol 144.3 144.3 79.3 55% TOTAL 2005 7,093.0 7,093.0 7,257.4 102% PRESIREP 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0101 COORDINATION ET SUIVI DES FONCTIONS DE LA 2,671.2 2,673.8 2,417.1 90% PRESIDENCE 0102 SERVICE DE CONSEILLES TECHNIQUES 55.3 55.3 36.1 65% 0103 COORDINATION DE LA POLITIQUE DU GOUVERNEMENT 80.7 80.7 58.6 73% 0104 PALAIS PRESIDENTIEL 531.2 531.2 693.3 131% 0105 BONNE GOUVERNANCE 1,544.6 1,439.2 1,518.1 105% 0106 SECURITE DES BIENS ET DES SERVICES 2,773.2 2,973.2 3,258.6 110% 0108 SCIENCE,TECHNOLOGIE ET RECHERCHE SCIENTIFIQUE 0.0 112.6 3.3 3% TOTAL 2006 7,656.3 7,866.0 7,985.1 102% PRESIREP 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0101 COORDINATION ET SUIVI DES FONCTIONS DE LA 3,083.1 3,083.1 0.0 0% PRESIDENCE 0102 SERVICE DE CONSEILLES TECHNIQUES 74.8 74.8 0.0 0% 0103 COORDINATION DE LA POLITIQUE DU GOUVERNEMENT 58.0 58.0 0.0 0% 0104 PALAIS PRESIDENTIEL 307.7 307.7 0.0 0% 0105 BONNE GOUVERNANCE 1,451.4 1,451.4 0.0 0% 0106 SECURITE DES BIENS ET DES SERVICES 3,627.2 3,627.2 0.0 0% TOTAL 2007 8,602.3 8,602.3 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. 228 World Bank Working Paper Table 5.116. Snapshot--PRIMATURE--Office of Prime Minister PRIMATURE 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0401 POLITIQUE GENERALE DU GOUVERNEMENT 57.8 57.8 4.9 8% 0402 PROGRAMME DU GOUVERNEMENT 63.8 63.8 0.0 0% 0403 LOIS, REGLEMENT ET DECISIONS DU CONSEIL DES MINISTRES 63.2 63.2 0.0 0% 0404 RENFORCEMENT DES CAPACITES DE COORDINATION 789.7 789.7 861.6 109% 0405 TRADUCTION DES DOCUMENTS OFFICIELS ET INTERPRETARIAT 73.3 73.3 49.7 68% 0406 Systeme National d'Information 1,151.0 1,151.0 888.7 77% TOTAL 2004 2,198.9 2,198.9 1,804.8 82% PRIMATURE 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0401 General Government Policy 48.4 48.4 62.9 130% 0402 Government Programmes 104.9 104.9 19.9 19% 0403 Roles, Regulations and Decision Making by the Cabinet 53.4 53.4 92.2 172% 0404 Strengthening co-ordination Capacity 695.9 695.9 993.0 143% 0405 Translation and Interpretation of Official Documents 63.1 63.1 38.2 60% 0406 National System of Information 1,309.6 1,309.6 863.5 66% TOTAL 2005 2,275.3 2,275.3 2,069.6 91% PRIMATURE 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0401 POLITIQUE GENERALE DU GOUVERNEMENT 42.4 42.4 75.2 177% 0402 PROGRAMME DU GOUVERNEMENT 89.1 89.1 19.3 22% 0403 LOIS, REGLEMENT ET DECISIONS DU CONSEIL DES MINISTRES 44.6 44.6 126.2 283% 0404 RENFORCEMENT DES CAPACITES DE COORDINATION 714.5 716.1 981.2 137% 0405 TRADUCTION DES DOCUMENTS OFFICIELS ET INTERPRETARIAT 9.3 9.3 0.0 0% 0406 GESTION DES RISQUES ET DES CATASTROPHES 137.4 137.4 98.9 72% 0407 SERVICE POLITIQUE DE L'INFORMATION ET DEVELOPPEMENT DES MEDIAS 1,276.9 1,318.4 1,167.8 89% 0408 APPUI INSTITUTIONNEL 106.3 106.3 78.8 74% 0409 APPUI A L'AUTOPROMOTION DE LA FEMME 112.3 112.3 12.6 11% 0410 PROMOTION DE L'EGALITE/EQUITE DU GENRE DANS LE DEVELOPPEMENT 39.1 39.1 3.9 10% 0411 PROMOTION ET PROTECTION DE LA FAMILLE 38.6 38.6 1.2 3% 0412 PROMOTION ET PROTECTION DES DROITS DE L' ENFANT 103.0 115.8 39.1 34% 0413 APPUI AU CONSEIL NATIONAL DES FEMMES 94.9 136.7 83.0 61% 0414 APPUI AU SECRETARIAT DE SUIVI DE PLATE FORME D'ACTION DE BEIJING 70.9 80.4 59.6 74% 0415 OBSERVATOIRE DU GENRE 29.6 29.6 0.0 0% TOTAL 2006 2,908.8 3,016.2 2,746.9 91% PRIMATURE 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0401 POLITIQUE GENERALE DU GOUVERNEMENT 4.6 4.6 0.0 0% 0402 PROGRAMME DU GOUVERNEMENT 65.3 65.3 0.0 0% 0403 LOIS, REGLEMENT ET DECISIONS DU CONSEIL DES MINISTRES 46.6 46.6 0.0 0% 0404 RENFORCEMENT DES CAPACITES DE COORDINATION 758.1 758.1 0.0 0% 0405 GESTION DES RISQUES ET DES CATASTROPHES 135.8 135.8 0.0 0% 0407 PROMOTION ET PROTECTION DE LA FAMILLE 35.0 35.0 0.0 0% 0408 Coordination et Suivi-Evaluation 249.4 249.4 0.0 0% 0409 GENRE ET PROMOTION DE LA FEMME 325.1 325.1 0.0 0% 0410 SERVICE POLITIQUE DE L'INFORM ET DEVELOPP. DES MEDIAS 1,410.8 1,410.8 0.0 0% TOTAL 2007 3,030.6 3,030.6 0.0 0% Note: Actual 2007 program data was not provided by the time of analysis/writing. Source: MINECOFIN - data shown in nominal terms Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform Table 5.116 (continued) Source: MINECOFIN--data shown in nominal terms. 229 230 World Bank Working Paper Table 5.117. Snapshot--COURS.SUP--Office of the Supreme Court COURS.SUP 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0501 Coordination des Activites Administratives et Juridictionnelles 1,436.9 1,436.9 1,557.4 108% 0502 Rendement de la Justice a la Population 1,055.1 1,055.1 368.0 35% 0503 Inspection Juridictionnelle 11.4 11.4 0.0 0% 0504 Gestion de la Carriere des Magistrats 13.2 13.2 7.0 53% 0505 Juridictions GACACA 1,615.3 1,615.3 543.6 34% TOTAL 2004 4,131.8 4,131.8 2,476.1 60% COURS.SUP 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0501 Access to Justice by the Population 2,379.2 2,379.2 2,260.8 95% 0502 Efficient Judicial Organization 80.7 80.7 9.5 12% 0503 Materials, Financials and Human Resource Management 19.2 19.2 2.4 12% TOTAL 2005 2,479.0 2,479.0 2,272.7 92% COURS.SUP 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0501 COURT SUPREME 2,855.7 2,984.1 2,673.5 90% 0502 HAUTE COURT DE LA REPUBLIQUE ET SES CHAMBRES 65.8 65.8 64.4 98% 0503 12 TRIBUNAUX DE GRANDE INSTANCE 123.8 123.8 138.9 112% 0504 60 TRIBUNAUX DE BASE 412.1 412.1 282.9 69% 0505 GESTION ET COORDINATION DES SERVICES 51.0 51.0 92.7 182% TOTAL 2006 3,508.5 3,636.9 3,252.3 89% COURS.SUP 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0501 COURT SUPREME 796.8 869.6 0.0 0% 0502 Justice Rendue par la Cour 559.3 486.5 0.0 0% 0503 Justice Rendue par les cours de Grande Instance 1,203.5 1,203.5 0.0 0% 0504 Justice Rendue par les Tribunaux de Petite Instance 972.1 972.1 0.0 0% 0505 GESTION ET COORDINATION DES SERVICES 49.0 49.0 0.0 0% TOTAL 2007 3,580.6 3,580.6 0.0 0% Source: MINECOFIN data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 231 Source: MINECOFIN--data shown in nominal terms. Table 5.117 (continued) 232 World Bank Working Paper Table 5.118. Snapshot--MINADEF--Ministry of Defense MINADEF 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0601 Armee Rwandaise 4,798.3 4,798.3 4,843.1 101% 0602 Armee de l'Air 1,516.8 1,516.8 1,533.3 101% 0603 Armee Marine 274.5 274.5 96.0 35% 0604 Mechanised 517.2 517.2 220.0 43% 0605 Garde Republicaine 23.7 23.7 0.0 0% 0606 Coordination des Services Militaires 18,460.4 18,460.4 18,805.8 102% TOTAL 2004 25,591.0 25,591.0 25,498.1 100% MINADEF 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0601 Rwandan Army 6,138.4 6,138.4 5,642.6 92% 0602 AirForce 866.5 866.5 362.6 42% 0603 Navy Army 134.8 134.8 111.0 82% 0604 Mechanised 208.0 208.0 51.1 25% 0605 Republican Guard 30.4 30.4 0.6 2% 0606 Coordination of Military Service 19,761.6 19,761.6 19,961.3 101% 0607 Peace Keeping Force 3,385.4 5,230.5 9,186.0 176% TOTAL 2005 30,525.1 32,370.2 35,315.1 109% MINADEF 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0601 COORDINATION DES SERVICES MILITAIRES 31,675.5 31,706.2 35,986.8 114% 0602 FORCES RWANDAISES DE LA DEFENSE 6,935.2 6,935.2 6,003.0 87% TOTAL 2006 38,610.6 38,641.3 41,989.8 109% MINADEF 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0601 COORDINATION DES SERVICES MILITAIRES 5,566.2 5,566.2 0.0 0% 0602 FORCES RWANDAISES DE LA DEFENSE 27,357.6 27,357.6 0.0 0% TOTAL 2007 32,923.9 32,923.9 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 233 Source: MINECOFIN--data shown in nominal terms. Table 5.118 (continued) 234 World Bank Working Paper Table 5.119. Snapshot--MININTER--Ministry of Internal Security MININTER 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0701 Service de la Police 5,737.5 5,737.5 5,249.7 91% 0702 Prisons 754.2 754.2 393.8 52% 0703 Services de Gestion et d'Appui 191.3 191.3 38.3 20% 0704 Suivi et Evaluation des Politiques du MININTER 62.2 62.2 201.1 323% TOTAL 2004 6,745.3 6,745.3 5,882.9 87% MININTER 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0701 POLICE NATIONALE 6,958.5 6,958.5 6,230.6 90% 0702 SERVICES DES PRISONS 600.0 600.0 482.1 80% 0703 SERVICE DE GESTION ET D'APPUI 150.0 150.0 170.9 114% 0704 PLANIFICAT'N, EVALUAT'N ET MONITORING DES 41.7 41.7 7.3 18% POLITIQUES DU MININTER TOTAL 2005 7,750.3 7,750.3 6,890.9 89% MININTER 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0701 POLICE NATIONALE 7,611.0 7,761.0 7,461.7 96% 0702 SERVICES DES PRISONS 746.2 746.2 675.4 91% 0703 SERVICE DE GESTION ET D'APPUI 241.1 239.8 217.4 91% 0704 PLANIFICAT'N, EVALUAT'N ET MONITORING DES 17.6 17.6 8.3 48% POLITIQUES DU MININTER 0705 TECHNOLOGIE DE L'INFOR ET DE LA COMMUNICAT'N DU 42.6 42.6 31.0 73% MININTER TOTAL 2006 8,658.4 8,807.1 8,393.8 95% MININTER 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0701 Administration du MINITER 273.7 280.8 0.0 0% 0702 POLICE NATIONALE (IN 2007 - this is divided under prog 0702- 0.0 0.0 0.0 n/a 0706) 0702 Service Corporatif de la Police 7,438.7 7,438.7 0.0 0% 0703 Investigation du Crime et Dectective (incl Police National) 361.0 361.0 0.0 0% 0704 Functionnement General (incl Police National) 289.9 289.9 0.0 0% 0705 Services Specialises et Services d'Appui (incl Police National) 261.0 261.0 0.0 0% 0706 Ecole des Formations des Policiers (incl Police National) 115.9 115.9 0.0 0% 0707 Administration du Prisons 675.9 675.9 0.0 0% 0708 Bien Etre des Prisonniers 140.2 140.2 0.0 0% 0709 Infrastructures 56.4 49.8 0.0 0% TOTAL 2007 9,612.7 9,613.2 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 235 Source: MINECOFIN--data shown in nominal terms. Table 5.119 (continued) 236 World Bank Working Paper Table 5.120. Snapshot--MINAFFET--Ministry of Foreign Affairs and Cooperation MINAFFET 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0801 service de gestion et d'appui 507.4 507.4 491.6 97% 0802 PROTOCOLE D'ETAT 102.3 102.3 81.8 80% 0803 Relations Bilaterales 125.2 125.2 46.8 37% 0804 Relations Multilaterales 31.1 31.1 14.4 46% 0805 REPRESENTATION DIPLOMATIQUE et Consulaires 2,921.3 2,921.3 2,990.6 102% TOTAL 2004 3,687.3 3,687.3 3,625.2 98% MINAFFET 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0801 service de gestion et d'appui 276.1 276.1 308.3 112% 0802 PROTOCOLE D'ETAT 50.7 50.7 36.3 72% 0803 RELATIONS AVEC AFRIQUE, ASIE ET OCEANIE 249.1 249.1 219.1 88% 0804 RELATIONS AVEC EUROPE 11.5 11.5 36.8 320% 0805 REPRESENTATION DIPLOMATIQUE A L'ETRANGER 4,135.2 4,135.2 4,001.2 97% 0806 RELATIONS AVEC AMERIQUE ET NATIONS UNIES 65.5 65.5 37.0 57% 0807 COMMUNICATION DIASPORA RWANDAISE 20.1 20.1 23.3 116% TOTAL 2005 4,808.2 4,808.2 4,662.1 97% MINAFFET 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0801 service de gestion et d'appui 431.8 431.8 445.8 103% 0802 PROTOCOLE D'ETAT 53.5 53.5 62.2 116% 0803 RELATIONS AVEC AFRIQUE, ASIE ET OCEANIE 257.6 257.6 469.7 182% 0804 RELATIONS AVEC EUROPE 27.8 27.8 79.9 287% 0805 REPRESENTATION DIPLOMATIQUE A L'ETRANGER 4,384.3 4,534.9 4,393.5 97% 0806 RELATIONS AVEC AMERIQUE ET NATIONS UNIES 41.0 41.0 22.8 55% 0807 COMMUNICATION, REL. PUBLIQUES ET DIASPORA RWANDAISE 25.0 25.0 8.8 35% 0808 INFORMATION, COMMUNICATION ET TECHNOLOGY 28.5 28.5 6.4 23% TOTAL 2006 5,249.4 5,400.0 5,489.1 102% MINAFFET 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 0801 service de gestion et d'appui 446.9 446.9 0.0 0% 0802 PROTOCOLE D'ETAT 66.0 66.0 0.0 0% 0803 RELATIONS AVEC AFRIQUE, ASIE ET OCEANIE 295.0 295.0 0.0 0% 0804 RELATIONS AVEC EUROPE 92.8 92.8 0.0 0% 0805 REPRESENTATION DIPLOMATIQUE A L'ETRANGER 5,765.4 5,765.4 0.0 0% 0806 RELATIONS AVEC AMERIQUE ET NATIONS UNIES 53.4 53.4 0.0 0% 0807 COMMUNICATION, REL. PUBLIQUES 15.5 15.5 0.0 0% 0808 INFORMATION, COMMUNICATION ET TECHNOLOGY 42.0 42.0 0.0 0% 0809 Agence Rwandaise de la Diaspora 143.0 143.0 0.0 0% TOTAL 2007 6,919.8 6,919.8 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 237 Source: MINECOFIN--data shown in nominal terms. Table 5.120 (continued) 238 World Bank Working Paper Table 5.121. Snapshot--MINICOM--Ministry of Commerce, Industry, Investment Promotion, Tourism and Cooperatives MINICOM 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1001 ORIENTATION ET COORDINATION DES PROGRAMMES DU MINISTERE 119.7 119.7 194.5 163% 1002 RENFORCEMENT DES CAPACITES INSTITUTIONNELLES 26.6 26.6 24.0 90% 1003 PROMOTION DU COMMERCE 80.3 80.3 93.0 116% 1004 PROMOTION DES EXPORTATIONS 1,176.0 1,176.0 821.7 70% 1005 DEVELOPPEMENT DU TOURISME 212.4 212.4 155.1 73% 1006 DEVELOPPEMENT INDUSTRIEL ET PROMOTION DE L'ARTISANAT 811.3 811.3 658.5 81% 1007 PROMOTION DES COOPERATIVES 20.8 20.8 6.9 33% TOTAL 2004 2,447.0 2,447.0 1,953.8 80% MINICOM 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1001 Direction and Coordination of Ministerial Programmes 539.9 539.9 652.6 121% 1002 Strengthening of Institutional Capacity 32.6 32.6 14.2 43% 1003 Trade Promotion 30.7 30.7 12.9 42% 1004 Export Promotion 1,380.5 1,380.5 1,304.9 95% 1005 Development of Tourism 496.6 496.6 116.5 23% 1006 Industriel Development and Promotion of Craft Industry 1,191.1 1,791.1 1,562.9 87% 1007 Promotion of Co-Operations 58.6 58.6 32.6 56% TOTAL 2005 3,730.1 4,330.1 3,696.6 85% MINICOM 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1001 ORIENTATION ET COORDINATION DES PROGRAMMES DU MINISTERE 432.8 401.3 369.5 92% 1002 RENFORCEMENT DES CAPACITES INSTITUTIONNELLES 25.8 25.8 9.6 37% 1003 PROMOTION DU COMMERCE 23.1 23.1 11.2 49% 1004 PROMOTION DES EXPORTATIONS 789.1 789.1 1,072.6 136% 1005 DEVELOPPEMENT DU TOURISME 60.6 60.6 32.1 53% 1006 DEVELOPPEMENT INDUSTRIEL ET PROMOTION DE L'ARTISANAT 102.2 102.2 48.5 47% 1007 OFFICE RWANDAIS DE NORMALISATION 464.3 805.9 613.2 76% 1008 AGENCE RWANDAIS DE PROMOTION DES INVESTISSEMENTS 1,758.9 1,766.7 2,034.2 115% 1009 CENTRE D'APPUI AUX PETITES ET MOYENNES ENTREPRISES DU RWANDA 126.1 126.1 126.1 100% 1010 PROMOTION DES COOPERATIVES 236.4 236.4 308.7 131% TOTAL 2006 4,019.4 4,337.2 4,625.7 107% MINICOM 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1001 ORIENTATION ET COORDINATION DES PROGRAMMES DU MINISTERE 649.3 649.3 0.0 0% 1002 PROMOTION DU COMMERCE 54.4 54.4 0.0 0% 1003 PROMOTION DES EXPORTATIONS 375.4 375.4 0.0 0% 1004 DEVELOPPEMENT INDUSTRIEL ET PROMOTION DE L'ARTISANAT 93.2 93.2 0.0 0% 1005 PROMOTION DES COOPERATIVES 257.6 257.6 0.0 0% 1006 Coordination du Systeme d'Information Communication et Technologie 18.6 18.6 0.0 0% 1007 OFFICE RWANDAIS DE NORMALISATION 924.2 924.2 0.0 0% 1008 Promotion des Investissements et EXPORTATIONS 1,797.3 1,797.3 0.0 0% TOTAL 2007 4,170.0 4,170.0 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 239 Source: MINECOFIN--data shown in nominal terms. Table 5.121 (continued) 240 World Bank Working Paper Table 5.122. Snapshot--MINISTR--Ministry of Science, Technology, and Scientific Research MINISTR 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1101 Appui Institutionnel 73.0 73.0 0.0 0% 1102 Reformes Governmental et Incentives 25.7 25.7 0.0 0% 1103 Commission National de la Science, Technologie et Innovation 35.0 35.0 0.0 0% 1104 Interventions a Tous les Niveaux de Recherche 97.0 97.0 0.0 0% 1105 Intervetions a Tous les Secteurs 218.7 218.7 0.0 0% 1106 Institutions de Recherche 1,044.3 1,044.3 0.0 0% TOTAL 2007 1,493.7 1,493.7 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Table 5.123. Snapshot--MINECOFIN--Ministry of Finance and Economic Planning MINECOFIN 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1201 PLANIFICATION ECONOMIQUE ET BUDGETISATION 1,525.5 1,525.5 1,464.5 96% 1202 EXECUTION,SUIVI ET RAPPORT DU BUDGET 30,757.7 30,757.7 28,912.6 94% 1203 MOBILISATION DES RESSOURCES 963.1 963.1 427.1 44% 1204 APPUI au FONCTIONNEMT EFFICACE du MINIS. des FINANCES ET 69,367.9 63,299.4 43,230.7 68% PLANIFICAT'N ECONO. TOTAL 2004 102,614.2 96,545.8 74,035.0 77% MINECOFIN 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1201 Development and Strengthening of Economic & Financial Capacity 1,685.5 1,685.5 2,027.5 120% 1202 Management of Public Spending and common Resources 20,409.3 23,499.4 24,323.8 104% 1203 National and Sectoral Planning and Macro Economic Management 1,260.4 1,260.4 762.8 61% 1204 Mobilisation of Internal and External Resources 67,036.5 56,440.3 32,700.4 58% 1205 INSPECTION , CONTROLE,AUDIT ET ORGANE DE TRANSPARENCE 690.1 690.1 495.8 72% 1207 Compensation for People Injured by Auto-Motor Vehicles 64.1 64.1 0.0 0% 1208 Analysis and Research of Adequate Solution Public Sector SocioEco Problems 51.6 51.6 0.0 0% TOTAL 2005 91,197.6 83,691.5 60,310.3 72% MINECOFIN 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1201 PLANIFICATION ECONOMIQUE ET BUDGETISATION 1,784.4 1,324.3 1,106.9 84% 1202 EXECUTION,SUIVI ET RAPPORT DU BUDGET 77,843.5 63,877.1 65,230.5 102% 1203 MOBILISATION DES RESSOURCES 4,723.5 4,723.5 4,191.4 89% 1204 APPUI au FONCTIONNEMT EFFICACE du MINIS. des FINANCES ET 2,393.4 2,288.1 2,266.8 99% PLANIFICAT'N ECONO. TOTAL 2006 86,744.8 72,212.9 72,795.6 101% MINECOFIN 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1201 PLANIFICATION ECONOMIQUE ET BUDGETISATION 1,918.0 1,908.0 0.0 0% 1202 EXECUTION,SUIVI ET RAPPORT DU BUDGET 59,894.1 75,400.1 0.0 0% 1203 MOBILISATION DES RESSOURCES 7,542.1 7,542.1 0.0 0% 1204 APPUI au FONCTIONNEMT EFFICACE du MINIS. des FINANCES ET 2,993.9 3,003.9 0.0 0% PLANIFICAT'N ECONO. TOTAL 2007 72,348.1 87,854.1 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 241 Source: MINECOFIN--data shown in nominal terms. Table 5.123 (continued) 242 World Bank Working Paper Table 5.124. Snapshot--MINIJUST--Ministry of Justice MINIJUST 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1301 Gestion Administrative et Planification 482.2 482.2 404.2 84% 1302 LEGISLATION ET SERVICE A LA COLLECTIVITE 49.6 49.6 17.2 35% 1303 Relationes Institutionnelles 10.8 10.8 0.0 0% 1304 REGLEMENT DU CONTENTIEUX ET AFFAIRES JURIDIQUES DE L' ETAT 133.2 133.2 45.5 34% 1305 Assurer un Foctionnement Efficace du Systeme Judiciaire 13.3 13.3 13.6 102% 1306 Respect et Application de la Loi 742.2 742.2 668.5 1066% 1307 Reforme Legislative 62.7 62.7 61.7 44% 1308 Formation et Recyclage Continue du Personnel Judiciaire 139.1 139.1 120.6 130% 1309 REPRESENTATION DU RWANDA AU TPIR A ARUSHA 93.1 93.1 66.6 29% 1310 TRAVAUX D'INTERET GENERAL 230.0 230.0 69.3 8% 1311 Administration Autonome et Efficace du Parquet 875.3 875.3 883.6 507% 1312 Poursuite de l'Action Publique 174.3 174.3 73.5 0% TOTAL 2004 3,005.7 3,005.7 2,424.6 81% MINIJUST 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1301 CABINET 138.2 138.2 578.0 418% 1302 RELATION PUBLIQUE ET GESTION DES RESSOURCES INTERNES 140.1 140.1 121.5 87% 1303 Improvement of Policy and Programme of Judicial System 34.1 34.1 11.7 34% 1304 Mediation and Collective Services 42.3 42.3 18.5 44% 1305 Committee of Mediators 20.1 20.1 0.0 0% 1306 State of Right Promotion 678.1 678.1 651.0 96% 1307 Regulation of Litigation and Judicial Affairs 52.8 52.8 19.9 38% 1308 Training and Continous Retraining of Judicial Staff 134.8 134.8 126.1 94% 1309 Justice for All 22.2 22.2 0.0 0% 1310 Community Works for Prisoners 300.0 300.0 248.3 83% 1311 Coordination of Financial and Administrative Activities 620.6 620.6 1,016.8 164% 1312 Conception, Elaboration and Adoption of Ethic and Dontological Manuals 6.5 6.5 5.3 81% 1313 Representation and Pick-up Charge of ICTR Activity 98.7 98.7 74.1 75% 1314 Development of Monitoring System 29.6 29.6 10.6 36% 1315 Planning and Documentation 125.0 125.0 10.6 8% 1316 Development of Local Network 31.8 31.8 0.0 0% 1317 Information and Awareness 39.4 39.4 7.5 19% 1318 Development of Enlarge Productive Partnership 49.1 49.1 22.0 45% 1319 Respect of Human Right and Genocide Victim Protecion and Vulnerable Groups 55.8 55.8 0.0 0% 1320 Litigation of Genocide related GACACA Jurisdictions Activities 15.1 15.1 0.0 0% 1321 Rursue of People Victims of Genocide and Follow up of ICTR Activities 42.0 42.0 0.6 1% 1322 GACACA Jurisdictions 1,780.7 1,780.7 3,661.8 206% TOTAL 2005 4,457.2 4,457.2 6,584.3 148% MINIJUST 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1301 CABINET 131.4 131.4 169.4 129% 1302 RELATION PUBLIQUE ET GESTION DES RESSOURCES INTERNES 463.2 478.1 471.1 99% 1303 UNITE ICT 25.2 25.2 29.1 116% 1304 PLANIFICATION ET POLITIQUE JUDICIAIRE 21.0 21.0 19.2 92% 1305 LEGISLATION ET SERVICE A LA COLLECTIVITE 92.0 188.1 88.4 47% 1306 PROMOTION DE L'ETAT DE DROIT 725.3 626.9 594.8 95% 1307 COMITES DES CONCILIATEURS 7.8 7.8 0.0 0% 1308 UNITE JUSTICE ET RELATIONS INSTITUTIONNELLES 19.4 19.4 8.2 42% 1309 RENFORCEMENT DE L' ORGANE D' EXECUTION DES JUGEMENTS(HUISSIERS) 6.3 6.3 0.0 0% 1310 REPRESENTATION DU RWANDA AU TPIR A ARUSHA 118.0 118.0 86.9 74% 1311 REGLEMENT DU CONTENTIEUX ET AFFAIRES JURIDIQUES DE L' ETAT 51.2 51.2 23.7 46% 1312 TRAVAUX D'INTERET GENERAL 552.6 552.6 424.6 77% 1313 INSTITUT SUPERIEUR DE PRATIQUE ET DEVELOPPEMENT DU DROIT 141.6 158.4 123.4 78% 1322 JURIDICTIONS GACACA 1,558.5 1,808.5 2,000.1 111% TOTAL 2006 3,913.6 4,193.0 4,038.9 96% MINIJUST 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1301 Administration du MINIJUST (incl. Cabinet, Ressources Internes, ICT) 523.8 523.8 0.0 0% 1302 LEGISLATION ET SERVICE A LA COLLECTIVITE 29.7 29.7 0.0 0% 1303 REGLEMENT DU CONTENTIEUX ET AFFAIRES JURIDIQUES DE L' ETAT 551.9 551.9 0.0 0% 1304 Services Auxiliares 583.3 583.3 0.0 0% 1305 PLANIFICATION ET POLITIQUE JUDICIAIRE 518.5 518.5 0.0 0% 1306 Protection des Droits de l'Homme 37.8 37.8 0.0 0% 1307 Promotion des Droits de L'Homme 51.9 51.9 0.0 0% 1312 TRAVAUX D'INTERET GENERAL 764.6 764.6 0.0 0% 1313 INSTITUT SUPERIEUR DE PRATIQUE ET DEVELOPPEMENT DU DROIT 250.9 250.9 0.0 0% 1322 JURIDICTIONS GACACA 4,001.2 4,001.2 0.0 0% TOTAL 2007 7,313.7 7,313.7 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 243 Source: MINECOFIN--data shown in nominal terms. Table 5.124 (continued) 244 World Bank Working Paper Table 5.125. Snapshot--MIJESPOC--Ministry of Youth, Sports, and Culture (prior to March 2008) MIJESPOC 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1501 Service d'Appui au Niveau Central 102.2 102.2 152.5 149% 1502 MOBILISATION DES JEUNES 52.3 52.3 41.1 78% 1503 PROMOTION DES SPORTS ET LOISIRS 223.3 223.3 275.4 123% 1504 Promotion Culturelle 91.8 91.8 121.4 132% 1505 Conservation des Archives 28.0 28.0 12.8 46% 1506 RECHERCHE ACQUISIT'N CONSERVAT'N et VALORISAT'N DU PATRIMOINE 207.1 207.1 205.5 99% 1507 Memoire du Genocide 134.9 134.9 60.2 45% 1508 Encadrement des Structures et Organisations des Jeunes 91.3 91.3 67.5 74% 1509 Appui a la Gestion 40.0 40.0 23.3 58% TOTAL 2004 971.0 971.0 959.6 99% MIJESPOC 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1501 Central Support Services 138.6 138.6 145.1 105% 1502 MOBILISATION DES JEUNES 41.4 41.4 27.6 67% 1503 PROMOTION DES SPORTS ET LOISIRS 904.3 904.3 906.4 100% 1504 Cultural Promotion 48.3 48.3 38.7 80% 1505 Preservation of Archives 15.7 15.7 6.2 40% 1506 Research, Acquisition, Conservation and Enhancement of National Heritage 197.4 197.4 181.5 92% 1507 Memorial of Genocide 91.0 91.0 66.0 73% 1508 Training of Youth Structures and Orgainsations 101.5 101.5 76.2 75% 1509 Promoting Domumentary Activities 16.1 16.1 9.7 60% TOTAL 2005 1,554.2 1,554.2 1,457.3 94% MIJESPOC 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1501 RELATION PUBLIQUE ET APPUI A LA GESTION DES SERVICES INTERNES 224.7 204.2 230.6 113% 1502 AMELIORATION DES POLITIQUES ET PROGRAMMES DU MIJESPOC 25.3 25.3 25.8 102% 1503 DEVELOPPEMT DE LA TECHNOLOGIE, DE L'INFORMAT'N ET COMMUNICAT'N 11.9 11.9 7.8 65% 1504 MOBILISATION DES JEUNES 24.3 51.5 14.8 29% 1505 DEVELOPPEMENT DES INSTITUTIONS DE LA MEMOIRE 79.7 79.7 61.1 77% 1506 RECHERCHE ACQUISIT'N CONSERVAT'N et VALORISAT'N DU PATRIMOINE 208.8 251.2 239.2 95% 1507 PROMOTION DE L'ACTIVITE DOCUMENTAIRE ET CONSERVATION DES 149.5 149.5 47.2 32% 1508 PROMOTION DE LA CULTURE ET DES ARTS 32.2 32.2 81.7 253% 1509 PROMOTION DES SPORTS ET LOISIRS 786.1 786.1 712.5 91% 1510 ENCADREMENT DES STRUCTURES ET ORGANISATIONS DES JEUNES 106.0 118.5 106.5 90% TOTAL 2006 1,648.5 1,710.2 1,527.2 89% MIJESPOC 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1501 Appui au Niveau Central 192.9 192.9 0.0 0% 1502 Amelioration des Politiques et Programmes du MIJESPOC 80.4 80.4 0.0 0% 1503 Developpement de la Technologie, de l'Information et Communication 24.2 24.2 0.0 0% 1504 MOBILISATION DES JEUNES 31.2 31.2 0.0 0% 1505 DEVELOPPEMENT DES INSTITUTIONS DE LA MEMOIRE 109.8 109.8 0.0 0% 1506 RECHERCHE ACQUISIT'N CONSERVAT'N et VALORISAT'N DU PATRIMOINE 266.1 266.1 0.0 0% 1507 PROMOTION DE L'ACTIVITE DOCUMENTAIRE ET CONSERVATION DES 119.9 119.9 0.0 0% 1508 PROMOTION DE LA CULTURE ET DES ARTS 154.2 154.2 0.0 0% 1509 PROMOTION DES SPORTS ET LOISIRS 810.2 810.2 0.0 0% 1510 ENCADREMENT DES STRUCTURES ET ORGANISATIONS DES JEUNES 102.8 102.8 0.0 0% TOTAL 2007 1,891.7 1,891.7 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 245 Table 5.126. Snapshot--PARQUET GENERAL--Office of the Prosecutor General PARQUET GENERAL 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1701 DEVELOPPEMENT ET RENFORCEMENT DES RESSOURCES 1,318.7 1,467.7 1,327.8 90% HUMAINES 1703 RENFORCEMENT DES CAPACITES DES SERVICES D APPUI 10.7 10.7 15.8 148% ET DE COORDINATION 1704 RENFORCEMENT DES CAPACITES DU SERVICE DE 12.1 12.1 0.0 0% PLANIFICATION 1705 DEVELOPPMT ET MODERNISAT'N DES INFRAST. 70.0 70.0 27.8 40% ,COLLECTE,TRAITEMT ET ECHANGE 1706 PARTENARIAT AU NIVEAU NATIONAL ET INTERNATIONAL 35.0 35.0 14.1 40% 1707 CREATION D'UN SERVICE D'APPUI AUX VICTIMES ET 24.0 24.0 21.7 90% PROTECTION DES TEMOINS 1708 PREPARATION EN VUE DE PRENDRE LA RELEVE DU TPIR 31.1 31.1 15.5 50% 1709 POURSUIVRE des PREVENUS de CRIME de GENOCIDE et 44.2 314.2 292.0 93% CEUX AYANT FUI le PAYS TOTAL 2006 1,545.8 1,964.8 1,714.7 87% PARQUET GENERAL 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 1701 Gestion des Ressources Propres 1,336.3 1,336.3 0.0 0% 1702 Poursuite Judiciaire 190.7 190.7 0.0 0% 1704 Protection des Victimes et des Temoins 71.0 71.0 0.0 0% TOTAL 2007 1,598.0 1,598.0 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. Table 5.127. Snapshot--MIGEPROF--Office of Gender and Family Promotion (as of 2006 under Office of Prime Minister) MIGEPROF 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2001 Appui Institutionnel 365.1 365.1 228.8 63% 2002 Appui a l'Autopromotion de la Femme 140.9 140.9 95.6 68% 2003 Promotion de l'Approche du Genre dans le Developpement 77.7 77.7 38.7 50% 2004 Promotion de l'Equite socio-Juridique 68.6 68.6 12.1 18% 2005 Rehabilitation de la Familie 5.6 5.6 4.6 82% 2006 Promotion des Droits de l'Enfant 49.6 49.6 39.2 79% TOTAL 2004 707.5 707.5 418.9 59% MIGEPROF 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2001 Appui Institutionnel 212.2 212.2 221.0 104% 2002 Appui a l'Autopromotion de la Femme 137.9 137.9 50.8 37% 2003 Promotion de l'Approche du Genre dans le Developpement 80.0 80.0 29.0 36% 2004 Rehabilitation de la Familie 59.9 59.9 11.4 19% 2006 Promotion des Droits de l'Enfant 157.4 157.4 53.3 34% 2007 Support to Women National Council 108.1 108.1 90.1 83% 2008 Support to Beijing Secretariate 72.6 72.6 38.9 0% TOTAL 2005 828.3 828.3 494.5 60% 246 World Bank Working Paper Table 5.127 (continued) Source: MINECOFIN--data shown in nominal terms. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 247 Table 5.128. Snapshot--MIFOTRA--Ministry of Public Service and Labor MIFOTRA 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2001 RENFORCEMENT DES CAPACITES DES DES SERVICES DE 168.8 168.8 184.6 109% GESTION ET D APPUI 2002 REFORME DE LA FONCTION PUBLIQUE 419.2 419.2 157.6 38% 2003 PROMOTION DU TRAVAIL ET DE LA SECURITE SOCIALE 83.5 83.5 44.5 53% 2004 Planification, Developpement et Reforecement des Capacites des 158.1 158.1 164.3 104% Ressources Humaines 2005 Formation Professionnelle et Apprentissage des Metiers 325.4 325.4 117.7 36% 2006 STRATEGIE DE RENFORCEMENT DES CAPACITES DES 81.7 81.7 30.2 37% RESSOURCES HUMAINES TOTAL 2004 1,236.7 1,236.7 698.9 57% MIFOTRA 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2001 RENFORCEMENT DES CAPACITES DES DES SERVICES DE 251.9 251.9 272.8 108% GESTION ET D APPUI 2002 REFORME DE LA FONCTION PUBLIQUE 2,021.7 2,021.7 1,949.9 96% 2003 PROMOTION DU TRAVAIL ET DE LA SECURITE SOCIALE 154.8 154.8 102.2 66% 2004 Planning, Development and Strengthening of Resources Capacity 225.6 225.6 128.8 57% 2005 STRATEGIE DE RENFORCEMENT DES CAPACITES DES 68.1 68.1 65.1 96% RESSOURCES HUMAINES TOTAL 2005 2,722.0 2,722.0 2,518.7 93% MIFOTRA 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2001 RENFORCEMENT DES CAPACITES DES DES SERVICES DE 182.5 182.5 109.9 60% GESTION ET D APPUI 2002 REFORME DE LA FONCTION PUBLIQUE 1,509.1 1,509.1 1,488.5 99% 2003 COMMISSION DE LA FONCTION PUBLIQUE 134.9 124.9 0.0 0% 2004 PROMOTION DU TRAVAIL ET DE LA SECURITE SOCIALE 120.8 120.8 78.5 65% 2005 PLANIFICAT'N , DEVELOPP. ET RENFORCMT DES CAPACITES 447.7 410.3 528.5 129% DES RESSOURCES HUMAINES 2006 STRATEGIE DE RENFORCEMENT DES CAPACITES DES 56.8 56.8 26.8 47% RESSOURCES HUMAINES TOTAL 2006 2,451.8 2,404.3 2,232.1 93% MIFOTRA 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2001 RENFORCEMENT DES CAPACITES DES DES SERVICES DE 250.6 250.6 0.0 0% GESTION 2002 Modernisation de l'Administration Publique 1,285.0 1,285.0 0.0 0% 2003 Promotion de l'Emploi 282.0 214.7 0.0 0% 2004 Administration DU TRAVAIL 103.0 103.0 0.0 0% 2005 PLANIFICAT'N , DEVELOPP. ET RENFORCMT DES CAPACITES 285.6 352.9 0.0 0% DES RESSOURCES HUMAINES TOTAL 2007 2,206.2 2,206.2 0.0 0% Source: MINECOFIN--data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. 248 World Bank Working Paper Table 5.128 (continued) Source: MINECOFIN--data shown in nominal terms. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 249 Table 5.129. Snapshot--MINALOC--Ministry of Local Government, Good Governance, Community Development, and Social Affairs MINALOC 2004 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2301 Decentralisation 3,430.6 3,430.6 2,670.9 78% 2302 DEVELOPPEMENT COMMUNAUTAIRE 966.2 966.2 404.4 42% 2303 PROTECTION SOCIALE 6,838.3 6,838.3 6,831.1 81% 2308 Appui Institutionnel 416.7 416.7 337.2 90% 2309 Bonne Gouvernance 546.0 546.0 491.7 90% TOTAL 2004 12,197.9 12,197.9 10,735.4 88% MINALOC 2005 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2301 ADMINISTRATION DU TERRITOIRE 3,389.6 3,389.6 3,348.3 99% 2304 Vulnerable Groups Unit 7,913.4 7,913.4 7,870.3 99% 2305 Community Development Unit 686.6 686.6 623.1 91% 2307 Public Relation Unit and Domestic Resources Management 401.4 401.4 397.7 99% 2307 Good Governance 1,099.7 1,099.7 2,454.8 223% TOTAL 2005 13,490.6 13,490.6 14,694.1 109% MINALOC 2006 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2301 ADMINISTRATION DU TERRITOIRE 137.5 137.5 110.1 80% 2302 DEVELOPPEMENT COMMUNAUTAIRE 936.1 936.1 889.6 95% 2303 PROTECTION SOCIALE 9,231.1 9,231.1 9,218.7 100% 2305 DECENTRALISATION FISCALE ET FINANCE LOCALE 5,027.7 5,027.7 5,006.5 100% 2306 RELATIONS PUBLIQUES ET GESTION DES RESSOURCES 483.7 429.1 366.1 85% INTERNES 2307 TECHNOLOGIE DE L'INFORMATION ET DE COMMUNICATION 36.1 36.1 17.9 50% 2309 BONNE GOUVERNANCE 3,997.1 3,997.1 3,666.0 92% TOTAL 2006 19,849.3 19,794.8 19,275.0 97% MINALOC 2007 ORIGINAL* REVISED^ ACTUAL ^ Execut. Prog S/Prog BUDGET BUDGET EXECUTION % 2301 Appui aux Services 365.0 365.0 0.0 0% 2302 Boone Gouvernance et Decentralisation 3,274.4 3,274.4 0.0 0% 2303 PROTECTION SOCIALE 11,228.1 11,228.1 0.0 0% 2305 Developpement Communautaire 862.5 862.5 0.0 0% 2306 Finance Locale 67.0 67.0 0.0 0% 2307 Administration du territoire 1,172.4 1,172.4 0.0 0% TOTAL 2007 16,969.4 16,969.4 0.0 0% Source: MINECOFIN data shown in nominal terms. Note: Actual 2007 program data was not provided by the time of analysis/writing. 250 World Bank Working Paper Table 5.129 (continued) Source: MINECOFIN--data shown in nominal terms. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 251 Notes 1 Please note that the PER tables included in this section have been, to the extent feasible, converted into 2007 prices to make the analysis of this chapter 5 consistent with the overall real price assumptions in other sections of this report. However, sometimes the conversion of the original PER tables has not been feasible or the original tables have already been converted but using other base years and, in some cases, CPI instead of GDP deflator. These differences are noted when applicable. 2 In 2008, Rwanda's parliament became the first in the world where women claim the majority (56 percent). 3 Projet de relance des activités d'urgence pour le production agricole. 4 Union des coopératives rizicoles au Rwanda. 5 Little of these funds had been raised at the time the 2007 Agriculture PER Report (Rwanda Public Expenditure Review Agriculture, 2007, on behalf of MINECOFIN, Martin Fowler, Claver Garisabo, Sam Kanyarukiga, Augustin Mutijima) was written. 6 Subsidy figures are from Agriculture PER, 2007. 7 As identified in Agriculture PER, 2007. 8 Food and Agriculture Organization of the United Nations and World Health Organization estimates 9 Rwanda Public Expenditure Review Agriculture, April 2007. 10 Rwanda Public Expenditure Review Agriculture, April 2007. 11 See annex 6.2 in Agriculture PER Report, 2007, for more details 12 See footnote 19 in Agriculture PER Report, 2007, page 35. 13 The Agriculture PER completed in 2007 presents expenditures mostly in nominal terms. However, when expenditures were converted into real prices, 2001 was used as a base year. When this is the case, 2001 is kept as a base year in the data shown in this report since uncertainty exist about specific assumptions made by the original authors of the Agriculture PER. Whenever data was presented in the Agriculture PER 2007 report in nominal terms, this report has converted data into 2007 prices (unless otherwise indicated) to ensure, to the extent possible, consistency with the overall report. 14 The cause of the underexecution in 2003 and 2004 remains unexplained. 15 See footnote 23 in Agriculture PER Report 2007, page 37. 16 Based on the EICV II (20052006) poverty survey. 17 Although exact data for crosssector contributions to agriculture do not exist, a summary of some interventions is listed on page 43 in Agriculture PER report, 2007. 18 See footnote 19 in Agriculture PER Report, 2007. 19 See section 6.1.1 in Agriculture PER Report, 2007. 20 See table 6.2 in Agriculture PER Report, 2007. 21 See annex 6.1 in Agriculture PER Report, 2007. 22 Rule of thumb: Wages should not exceed 60 percent of the recurrent budget. 23 For example, from 2002 to 2003, the development budget changed significantly as a big project came to an end. As ministries are moving toward SWAP, the fluctuations in development budgets are expected to be minimized. 24 Further details are outlined in annex 6.5 of Agriculture PER Report 2007. 25 Some of the fluctuations seen between programs in 2006 and 2007 are attributed mainly to reclassification of projects in this period. Also, districts are now also classifying their recurrent development budgets for agriculture in line with the PSTA programs and subprograms. 26 See page 18 of the Agricultural PER Report 2007 for further details of the key objectives set forth in this strategy. 27 According to NBR Annual Report, 2005 252 World Bank Working Paper 28 See section 6.1.2 in Agriculture PER Report, 2007. 29 Trunc commun is Rwandas lower general secondary school. It takes three years to complete; these are the last three years of a "nine year basic education," after the first six years of primary. 30 The Education PER Report, July 2007, based these calculations on sources from the FTI appraisal in 2006 and the IMF report of March 2007. 31 Based on 2001 constant prices. For further details, see the Education PER Report, July 2007, page 20. 32 Chapter 4 in the Education PER Report, July 2007, discusses education access and related issues in depths. 33 In French, RAMA stands for La Rwandaise d'Assurance Maladie. 34 Public Expenditure Review Government and Donor Expenditure on Social Protection, also referred to as the Social Protection PER, 2006, on behalf of MINECOFIN, by Mick Foster, Mick Foster Economics Limited, November 2006. 35 Fonds National pour lAssistance aux Rescapés du Génocide 36 As used in the Social Protection PER, 2006, on which this chapter is primarily based. 37 As used in the Social Protection PER, 2006, on which this chapter is primarily based. 38 Transfer to FARG does not equal spending for genocide survivors. For details on this, see annex 2 of the Social Protection PER, 2006 and footnote 12 on page 4. 39 For example, EU financing of Ubudehe appears in the development budget for 2005, but planned financing for 2006 and 2007 does not. If included, 2006 the development budget would increase by RwF 27 billion. 40 In 2006 prices as used in the Social Protection PER, 2006. 41 Although formal medical insurers are required to contribute to a National Solidarity Fund, the amount is unknown. 42 See http://dad.synisys.com/dadrwanda. 43 For example, EU Ubudehe was onbudget in 2005, but offbudget in 2006. 44 The lower percentage for 2006 may be due to the data for 2006 being estimates rather than actual. 45 The Social Protection PER, 2006 noted that this decrease was due to classification issues and should not be taken too seriously. 46 Note that many NGOs programs supporting this segment of the population exist but are not included in the figures represented in the table. 47 According to the Social Protection PER, 2006, this raises concern that these programs are inconsistent with longterm solutions and supports for these dependent groups. 48 As noted in Social Protection PER, 2006. 49 At the time of the Water and Sanitation Sector Performance Report 2006, also referred to as the Water and Sanitation PER Report (WSS PER Report, 2006), by Stephen Hitimana, Consultant. 50 WSS PER Report, 2006. 51 AEP stands for Alimnation en Eau Potable which means Potable Water Supply. 52 CEPEX released its first comprehensive report in 2006, and prior years' data are less reliant and complete. Similar estimates in other sections of this report were made for 2004­07, possibly with other underlying assumptions. Hence, the data for 2004 in this report might illustrate a slightly different picture. 53 At the time of writing of the WSS PER Report in 2006, significant portions of these requirements remained unfunded. 54 WSS PER Report, 2006, page 36 does not make fully clear what year the original data in table is referring to. Although the population census 2002 is quoted, this report assumes that the original data provided by MINITERRE for the purpose of the WSS PER report in 2006, which is reflected in the table 10 of WSS PER, is as of 2005. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 253 55 See additional in WSS PER Report, 2006, page 36, chart 5. 56 See WSS PER Report, 2006, table 8. 57 For an elaboration of general points on public finance and the energy sector in SubSaharan Africa, see the Background Note later in this section. 58 A revision of prior unpublished work for the meeting of African Union finance ministers in Accra, May 2007. 59 This is from an energy budgeting perspective. For budgeting of the forestry ministry, revenues from forest products trade, including wood extraction for fuel purposes, can be a significant issue. 60 The term "energy sector" is used here to refer to electricity and its primary resources and to hydrocarbon fuels, because that is the arena for public finance. The definition can be expanded to include other fuels, and energy supply/enduse equipment, if the public finance role is important. 61 With the possible exception of street lighting (although distributional questions arise when street lighting is not uniformly available to different population groups). 62 Public investments in social service sectors are not capitalized, that is, they do not become part of a corporate balance sheet. 63 Much of the equipment in the supply chains of petroleum and electricity is imported, as are most of the use appliances, from vehicles to machinery. As economies grow, imports of such supply and use equipment grow. 64 Most energy trade is in oil and gas, both within Africa and from Africa to rest of the world; coal trade is confined to exports from South Africa to other continents. CHAPTER 6 Looking Forward: Outstanding Challenges and Concluding Remarks Recent Achievements and Outstanding Challenges Attaining the MDGs Rwanda has made substantial progress on many of the MDG targets. Impressive results were achieved in the social sectors: net primary school enrollment reached 95.8 percent in 2007, and completion rates increased to 75 percent in 2006. Immunization rates, at 96 percent, were among the highest in SubSaharan Africa. Use of insecticide treated bed nets increased from 4 percent to more than 70 percent of the population from 2004 to 2007. HIV prevalence, at 3 percent, has decreased. Unique in Africa, Rwanda scaled up access to health insurance (through local schemes referred to as mutuelles) from 7 percent to 75 percent of the population between 2003 and 2007, leading to increased use of health services (for example, assisted deliveries). Between 2004 and 2007, an additional 1 million people (14 percent of the population) gained access to clean drinking water, which has helped to reduce underfive mortality by 25 percent from 2000 to 2005. With the help of development partners, Rwanda has developed fiscal transfer mechanisms and contracts to ensure a sustained flow of resources from the central government to service delivery points, but to reach the MDGs, a sustained and increased flow of funding for service delivery in education, health (including HIV/AIDS), water, and energy is needed. Infrastructure Constraints to Growth and Agricultural Transformation The government recognizes that infrastructure is important to all sectors of the economy and society, and that an effective system of infrastructure and services is crucial for agricultural productivity and poverty reduction, a determinant of business investment, instrumental to human development, and the foundation for private sector development. There is a need to plan for, provide, and manage physical infrastructure that is efficient, costeffective, financially sustainable, and that supports the social and economic development priorities. Much has already been achieved in recent years in this regard. In particular, the government was successful in managing the energy crisis and is also successfully using solar, wind, methane gas, and biogas energy as power 254 Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 255 production alternatives. The primary challenge that remains is to focus on investments in the energy and transport sectors that would support Rwanda's competitiveness. Recent progress in the area of agricultural transformation includes development of an operational policy and improvements in mediumterm planning and budgeting, which included a firsttime resultsbased MTEF, a strategic plan, and the piloting of fiscal decentralization. These translated into improvements in budget execution rates. However, the economy remains largely agrarian, and rapid population growth, low agriculture productivity, and high pressure on the land remain key challenges to agricultural transformation. Agricultural production remains erratic. Irrigation needs to be improved. Extension services are practically nonexistent. The food shortages in early 2006 highlight the country's vulnerability to exogenous shocks and the need for productivityenhancing measures. Remaining challenges in this regard include continued agricultural practices with minimum value addition that align with competitive markets, and random settlement patterns that limit the opportunities for optimal land use, particularly for mechanized agriculture. Synergy across Sectors Synergy across sectors, especially between infrastructure and other sectors (for example, export promotion policy and access to basic services such as water, electricity, and transport) remains a challenge. For instance, the encouraging progress made in the area of agricultural transformation should be combined with enhanced human and institutional capacitybuilding initiatives. Another example is the need to formally link the export promotion strategy with the PSTA to ensure that it does not come at the expense of food security, but also with access to basic services such as water, energy, and transport. In general, infrastructure services need to be better linked with other sectors. Government Revenues Rwanda is implementing an ambitious development plan; however, the levels of development assistance are not in line with international commitments or with the average increase observed for Africa. Given the level of aid dependency and the government's debt sustainability concerns, the share of aid in the form of grants would need to increase further. Internal revenues increased in 2007 with increases in VAT and despite a reduction in trade taxes resulting from COMESA membership. Moreover, development assistance increased, including a larger share of aid provided as grants, as well as increased amounts of aid provided in the form of budget support. Despite debt relief under HIPC and the MDRI, Rwanda remains at a high risk of debt distress, mainly because of its high degree of reliance on external financing and its narrow export base. Debt sustainability analysis also indicates that Rwanda remains highly vulnerable to external shocks. Concluding Remarks Despite donors' commitments to scale up aid in line with the 2002 Monterrey Consensus and the 2005 Gleneagles Declaration, the outcome has not been consistant. Official development assistance, according to OECD data, declined in real terms by 256 World Bank Working Paper roughly 5 percent in 2006--the first drop since 1997. In 2007 major donors provided US$103.7 billion in development assistance, but this was a decrease from 2006 of 8.4 percent in real terms. Much of the difference is explained by two factors, namely (a) the end of debt relief for Iraq and Nigeria, which translated into exceptionally high levels of development assistance in 2005 and 2006 (US$107 billion and US$104 billion, respectively); and (b) small increases in other categories of official development assistance. With regard to resources needed to achieve the MDGs, donors are not on track to meet their commitments to increase levels of development assistance, which according to the 2005 G8 Summit would raise from the 2004 level of US$80 billion to US$130 billion in 2010. Donors would need to more than double their current rate of increase in development assistance, excluding debt relief, over the remaining three years to reach the 2010 target. Moreover, ODA as a percentage of donors' gross national income fell to 0.3 percent in 2006, after slowly edging its way up to 0.33 percent in 2005--still well short of the UN target of 0.7 percent. Currently, the only countries to exceed the UN target of 0.7 percent of gross national income are Denmark, Luxembourg, the Netherlands, Norway, and Sweden. Bilateral development assistance channeled to SubSaharan Africa, excluding debt relief, increased by 10 percent in 2007. Donors have yet to meet the rates of growth in development assistance that would support the Gleneagles G8 Summit intention to double their assistance to Africa by 2010. In Rwanda development assistance increased as well in 2007, although at a lower rate. Nevertheless, Rwanda remains highly aid dependent and is prone to fiscal risks associated with aid dependency, for example, fiscal risks, vulnerability to exogenous shocks, and debt sustainability. As such, aid predictability and in particular budget support predictability are important for the planning and uninterrupted implementation of development programs. In the case of budget support, the government adjusts to unexpected underdisbursements through higher domestic financing and reductions in domestic investment spending. To make matters worse, these adjustment needs are further exacerbated by the fact that they are sometimes accompanied by tax revenue shortfalls and current expenditure overruns. To smoothly absorb unexpected variations in external financing, the government has designed and implemented several innovative programs that are flexible, in that they are readily adjustable in terms of their scale (for example, capitation grants in the education sector or contributions to the CDF) to adjust to fiscal risks. Such flexible programs need to be developed in more sectors to better mitigate fiscal risks. The government much appreciates the increased alignment of donor activities around the government's priorities and its priority sectors. Rwanda, as one of the first three budget support Harmonization and Alignment pilots, now serves as a good practice example for both development partners and recipient governments. In light of the fact that (a) the government has demonstrated allocative efficiency of national and donor resources in the form of budget support; (b) substantive progress was made in terms of results, especially in priority sectors; and (c) Rwanda stands at a disadvantage in terms of increases in aid as compared to the average increase of 10 percent for Africa, and as compared to average commitments made by major donors at the 2005 G 8 meetings, the government, in its efforts to address outstanding challenges, would Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 257 welcome enhanced collaboration and additional resources from its development partners. Within the existing framework of partnership and mutual accountability, discretionary resources in the form of budget support would better contribute to enhancing aid effectiveness, as this modality not only represents the lowest transaction costs in terms of aid management, but also enhances collaboration and mutual benefit from increased collaboration. Additionally, to allow for better programming and budgeting and to reduce unnecessary duplications, the government iterates its preference for onbudget development assistance, as opposed to resources channeled through nongovernmental entities. Rwanda's high dependence on development resources increases its vulnerability to exogenous shock. Low intraannual as well as multiyear predictability of donor resources only exacerbates this vulnerability in addition to hindering the effectiveness of any mediumterm planning and budgeting process. Appendixes 259 Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 261 Appendix A. Government Reforms 1995­2003 OVERVIEW OF REFORMS 1995--2003 Governance National Tender Board created to oversee procurement. Independent Office of the Auditor General established. FARAP Action Plan adopted and under implementation. Expenditure Expenditure monitoring system (monthly "flash reports" on actual and budgeted expenditures) Management established in 2003; establishment of Central Public Investments and External Financing Bureau (CEPEX) to improve the coordination of investment projects at end-1998; initiation of the MTEF in 2000; Pilot Joint Monitoring System implemented in 2003; Strategic Planning and Poverty Reduction Monitoring Department (SPPRMD) launched to monitor poverty outcomes and assure coherence between the budget and the PRSP. Budget allocations to the social sectors increased significantly in real terms since 1998. These allocations are now protected from budget cuts. Revenue Adoption of fiscal measures (increase turnover tax to 15 percent in 1998, increase excise tax, and subjecting public enterprises to income tax in 1999) and the establishment of the autonomous Rwanda Revenue Authority (RRA) at end-1997. Civil Reform A civil service census carried out late 1998 provided the basis for civil service rationalization. Public Enterprise Reform The law providing the legal framework for the privatization of public enterprises was passed in 1996. Central Bank and A revised central bank statute giving the National Bank of Rwanda (NBR) independence in Monetary Policy Reform conducting monetary policy was adopted in mid-1997. At end-1998, Treasury bill auctions were introduced, giving the NBR an indirect instrument for controlling monetary aggregates and influencing interest rates. Financial Sector Reform A new banking law, promulgated in August 1999, provides for the effective, prudential regulation of commercial banks. Other regulations provided under the law were issued in 1999 and 2000. Comprehensive audits of all five commercial banks were carried out in 1999, providing the basis for these banks to be restructured to comply with the new prudential regulations. Social Security System The Caisse Social du Rwanda (CSR) was audited and an actuarial study was carried out. A large part of the Government's debt to the CSR was consolidated. Regulatory Framework The National Assembly adopted the Rwanda Investment Promotion Act and a new labor code for Private Sector was submitted to the National Assembly in 1999. In late 1999, an independent Private Sector Development Federation was established, paving the way for the abolition of the Government-controlled Chamber of Commerce. Gender An amendment to the civil code giving women the right to inherit and own property was promulgated in late 1999. Gender policy was adopted by Cabinet in 2003, and the Strategic Plan for the Ministry of Gender and Women in Development was finalized and adopted along with the related action plan, which includes a Comprehensive Legal Action Plan to eliminate gender disparities. Trade Liberation The maximum tariff rate was reduced in three steps, from 100 percent before 1995 to 25 percent from end-1999, and the number of non-zero tariff bands was reduced from 40 to 3. Trade restrictiveness index fell from 8 in 1995-97 to 2 by mid-2000. The coffee export tax was eliminated in early 1999. Rwanda joined COMESA in 2003. Exchange Rate Regime A fully liberalized and market-determined exchange rate system was adopted, foreign exchange bureau were licensed, and current account restrictions abolished. Agreements under Article VIII of the IMF's Articles of Agreement were accepted at end-1998. Domestic Prices and Price controls were eliminated for all but a handful of commodities (cement, electricity, water, and Marketing telecommunications). A flexible retail petroleum pricing mechanism was implemented in 1999. Existing laws are being revised to reflect the policy of price liberalization and provide a transparent framework for Government interactions where necessary. Demobilization Government has demobilized almost 10,000 soldiers, end-1997 and end-1998. In addition, about 15,000 ex-FAR (soldiers from the former Government) have been incorporated into the national army since early 1998. 262 World Bank Working Paper Appendix B. Joint Analytical Work Summary 2004­07 Table B.1. Joint Analytical Work by Donors and/or Country--Rwanda 2004­07 Delivered-- Program-- Delivered-- C: Program-- C: Combined Combined A: prepared A: prepared Total Coordinated prepared jointly w/ prepared jointly w/ Delivered AAA jointly w/ other jointly w/ other Fiscal Output (incl. Delivered other donors & other donors & Year Type Description suppl) (incl. suppl) donors Country donors Country 2004 Report Energy & 1 Environment 2004 Report CPAR (LKD) 1 2004 Report Country 1 2005 Report Fiscal Year 2005 1 1 1 1 2005 Report Assessment 1 2005 Report Rwanda 1 1 1 1 2006 Report Diagnostic Trade 1 1 1 1 Integ. Study 2007 Policy Note Policy Note 1 1 1 1 2007 Report Fiscal Year 2007 1 1 Source: World Bank. Coordinated Analytical Work refers to: (a) Analytic work conducted jointly (in form of a cash and/or inkind contribution) or in significant coordination with at least one multilateral or bilateral donor. In addition: the document clearly signals that the donor(s) accept the study as a product prepared jointly or in coordination with them. (b) Participatory analytical work with the country, for which the country contributes significantly in cash and/or inkind. (c ) Both (a) and (b). The FY0004 data on coordinated AAA offered only one option Prepared jointly with other donor(s); hence, considerable caution should be exercised in any comparison with postFY05 data, except if restricted solely to donor participation. Table B.2. Joint Technical Assistance Work by Donors and/or Country--Rwanda 2004­07 Delivered-- Program-- Delivered-- C: Program-- C: Combined Combined A: prepared A: prepared Total Coordinated prepared jointly w/ prepared jointly w/ Delivered AAA jointly other jointly other Fiscal Output (incl. Delivered w/other donors& w/other donors& Year Type Description suppl) (incl. suppl) donors Country donors Country 2005 Knowledge Integr. 1 Sharing Gender - Forum HIV/Aids 2006 How-To FSAP 1 Guidance Follow-up 2006 ESMAP 1 Tariff Policy Source: World Bank. Coordinated Analytical Work refers to: (a) Analytic work conducted jointly (in form of a cash and/or inkind contribution) or in significant coordination with at least one multilateral or bilateral donor. In addition: the document clearly signals that the donor(s) accept the study as a product prepared jointly or in coordination with them. (b) Participatory analytical work with the country, for which the country contributes significantly in cash and/or inkind. (c) Both (a) and (b). The FY0004 data on coordinated AAA offered only one option Prepared jointly with other donor(s); hence, considerable caution should be exercised in any comparison with postFY05 data, except if restricted solely to donor participation. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 263 Appendix C. Case Study of District of Rulindo 2006­08 Progress and Challenges Following Decentralization Introduction The district of Rulindo is located north of the city of Kigali, in the Northern Province. As part of the decentralization process in 2006, Rulindo was created as a district out of three provinces (KigaliNigali, Byumba, and Ruhengeri). Although Rulindo is located only about 60 minutes from the city of Kigali, it is considered a rural district and lacks much of the benefits of the urbanization that exists in Kigali and its surroundings. Rulindo is home to 261,000 people of the 9.2 million living in Rwanda. Ninetytwo percent of the population in Rulindo lives off of agriculture, and 38 percent live under extreme poverty. Of the government's transfers to districts in 2006 and 2007, Rulindo received about 3 percent of the total transfer. The 30 districts in Rwanda each received between 1 percent and 5 percent of the total transferred to the districts. Since the reorganization that followed the decentralization in 2006, Rulindo and all of the districts in the country have seen many changes. For example, since 2006 the districts' MTEF has been streamlined, and through this process, the 2008 MTEF is now aligned with the government's programs. This improvement followed after the government worked closely with districts to develop an MTEF that reflected both the needs of the people in the districts and the need to align budgeting, planning, and reporting with policy objectives and with the government's centrallevel programs. Rulindo, together with all other districts, took part in a major workshop in May 2007 where the districts were given a voice in the proposed new MTEF. District officials say that the process and the improvements made by the new MTEF have been welcomed. Rulindo was selected to participate in many piloting programs, several of which are now scaled up on a national level. These include mutuelles, Imihigo, Ubudehe, Umuganda, Ibimina, HIMO, and others. The district reports that the introduction of IMIHIGO, the contract between the president and the district, has proven a tremendous incentive for the district and its representatives to mobilize and commit to meeting its objectives. District officials speak of the contract as a source of inspiration that has helped create a sense of competitiveness among districts and within its own administration, thereby creating ownership in the process and progress and pride as the district manages to improve in various areas. The following section will (a) look at Rulindo's different sectors, to provide a sense of the district's progress to date compared with Rwanda overall; (b) review the planning processes, funding of sources, funding allocations, and expenditures in the district for some key sectors; and (iii) highlight successes and remaining challenges in the district. 1. Rulindo Sector Review Health Sector Rulindo currently has 17 health centers and 1 hospital, the Rutongo Hospital, employing 154 nurses and 6 doctors. Following the introduction of mutuelles, the coverage rate in the district reached 95 percent in 2007, the second highest coverage rate in the country following the district of Gasabo, and well above the national 264 World Bank Working Paper average coverage rate of 51 percent. The district also provides assistance for those households that cannot afford to pay for mutuelles. In 2007, the district supported 26,500 (10 percent) vulnerable families to pay for mutuelle coverage. Coverage of assisted deliveries has increased; at the end of 2007 it was 39 percent, up from 37 percent in 2006. In the beginning of 2008, the district set its goal for assisted deliveries to reach 42 percent. National average at the time was 51 percent. The introduction of bed nets has significantly decreased deaths related to malaria, and around 29,200 families are currently using bed nets--about 70 percent of the population of Rulindo, at par with the national average of 70 percent in 2007. Furthermore, 85 percent of the district's children were vaccinated. Education Sector Rulindo currently has 78 primary schools, of which 1 is private, 12 trunc commun, and 22 secondary cycle schools (22 public and 5 private). Rulindo also has 2 technical/vocational training centers, though it does not have any youth centers or sport centers. The district's student/teacher ratio in primary school is 56:1, below the national average of 71:1. Infrastructure Sector Public Markets: Rulindo constructed 2 new markets in 2007 3 more were under construction in 2008, and the longerterm plan is to complete 19 markets, depending on future resources being made available mainly though transfers such as CDF. Bridges and Roads: The district has 783 bridges, of which 496 are considered in bad condition. The district possesses 42 km of roads in good condition and 868 km of roads in bad condition. Water and Sanitation: The district has 160 water tanks and 838 public water taps, and while 40 percent of the population has access to clean water (below the national average, according to WSS PER 2006), 65 percent has access to proper sanitation, above the national average. The district reports that water and sanitation have seen major improvements in the last couple of years, primarily thanks to five newly constructed canals. Energy: The district uses solar energy to power six of its sector's offices. In terms of electrical power, the district has 850 km of electrical supply. Coal can be found in the Bahimba Swamp, but is not currently being used. Firewood is commonly used in the district instead, but the price is high (one stere costs RwF 4,000). Also, biogas is used in one secondary school in Shyorongi, Stella Matutina. The district reports that it does not generally experience blackouts, and the price from Electrogaz is standardized. Prisons: The district does not have a prison. The greatest needs in the infrastructure domain include road rehabilitation, improvements in the water supply, and school buildings. Furthermore, the district needs funding for its project of building the Kinihira Hospital, two new health centers, and four socalled postes de santes communautaires. In addition, the district is lacking a proper district headquarters and sector and cell offices. Agriculture The average family in Rulindo has 0.67 ha of land The government has initiated some programs to improve the district's agriculture sector, including land consolidation, Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 265 agroforestry, soil erosion protection, PSTA, and initiatives aligned with the EDPRS. The district reports that many changes have occurred since 2006, although many challenges remain. For example, the district reports that it has seen an increase in the number of technicians in its sectors, but the lack of capacity still remains. Rulindo has constructed 1,360 hectares of radical terrace and 1,960 water poles, but soil erosion remains a major problem, and although progress in terracing has been made, much more work is needed to sufficiently address the problem. Other improvements include the protection of 60 percent of land, modernization of harvest systems, and implementation of HIMO activities in three sectors, socalled foodforwork programs for terracing. Also, the district faces challenges with traditional livestock, although the government's OneCowPerFamily program has brought 40,000 new cows of a better breed to the district as of 2008. Sectors that receive some funding as well have chosen to provide 130 pigs to each sector. Other challenges within the agriculture sector in the district include the lack of technicians, weak income from agricultural production and harvest caused by lack of manure, and limited land to be used for agriculture and harvesting. Currently, the district reports that this sector and related climate change issues remain underfunded. Social Protection Program The District of Rulindo supports several social protection programs, including support of genocide survivors, OVCs, and families too poor to pay school fees and/or mutuelles. Genocide Survivor Support: Rulindo has built 223 houses to date for genocide survivors. It also helps pay school fees and mutuelle fees for genocide survivors. Orphans and Vulnerable Children Support: Rulindo pays school fees for vulnerable children, helps integrate them into families, and supports them by paying for the mutuelle. Education Fund Support: In 2006, more than RwF 14 million of district funds assisted students of poor households by paying school fees. In 2007, RwF 24.8 million was used for the same purpose. Public Works Support: Public works has decreased in the district, but own initiatives, including private sector initiatives, are being pursued. The Ubudehe program has helped address the issue of poverty and rural development, and the district reports that the program is deemed successful and has helped empower citizens to solve their own community problems. Productive and Private Sectors Productive sectors remain small. The district has only one mining site, Masoro. The district lacks markets where its citizens can conduct commerce. In 2007, two new markets were constructed, and three more are planned for the year 2008. The district plans on eventually having 19 markets to meet the district's needs, a plan that is heavily dependent on future funding. Furthermore, the district has some key industries that are providing valuable job opportunities. For example, it has a large tea factory and a large tea plantation. Only three or four districts in the country have similar tea production. The district also has a coffee washing station and a major juice producing factory, one of the largest in the country. This factory is evidence of the importance of access to capital as the factory, SINA Gerard, was started with funding 266 World Bank Working Paper from BRD and with a tax break from the government that helped the company, which employs 4,000 people, get off the ground. Besides juice, the company has developed and is selling several other goods, such as cassava powder and other food products. 2. Planning Processes, Funding Sources, Allocations, and Expenditures Comments on Budget Reforms and Plans Districts are required to submit a fiveyear district development plan, a threeyear plan (an MTEF), and annual action plans. In the years following decentralization, the district reports that the planning processes have improved and that MTEF changes and related budget reforms have been welcomed. However, the district also reports that the MTEF being submitted to the government for approval of funding and the MTEF actually being implemented are different, as insufficient funding requires reallocation of funds throughout the year. Civil Society and Joint Action Forum The district of Rulindo currently has 28 members on its Joint Action Forum (JAF), including members from NGOs, cooperatives, faithbased organizations, and the district itself. Last year a coordination committee was elected (consisting of a president, vice president, council members, and a secretary), and the forum's first steps to an integrated district plan were taken. Last year, international NGOs' contributions to the district were, for the first time, reflected in the district's budget. Committees were created to coordinate various initiatives in various sectors, and collaboration is under way among all JAF members to prepare next year's annual action plan. Last year was the first time efforts were coordinated among JAF members to prepare a joint action plan; part of this first action plan included the organization of the JAF itself. Today the JAF is organized around three committees: Committee for Good Governance, Committee for Social Affairs (includes health and education), and Committee for Economic Affairs. SNV, the Dutch NGO, has provided valuable support to the district in the efforts to establish the JAF. The district reports that the JAF has served the district well as it facilitated the planning and budgetallocation process and improved dialogue between the district and other stakeholders, such as civil society. However, greater collaboration is still desired, and some stakeholders still are not fully engaged in the process. An additional problem encountered in the planning phase is that many faithbased organizations and local NGOs do not know the budget for the year, making budget allocation and coordination difficult. Transfers Block Transfers: Block transfers are nonearmarked funds from the central government to support the district's administrative work and facilitate in the district's daytoday operations. Rulindo currently uses 70 percent of the funds to pay wages and salaries for its staff and the remaining 30 percent to pay for different development programs that were not funded by the CDF or other sources. The district reports that the block transfers are not sufficient to pay for all the needs of the district to run its operations effectively and efficiently and that local resources and revenues are not sufficient to make up the difference. However, prior to reform the situation was significantly worse, Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 267 and the district reports that the ongoing dialogue between districts and MINECOFIN is good and that funding has increased since decentralization. Earmarked Transfers: Earmarked transfers are being used for specific programs, such as capitation grants. The district reports that the earmarked transfers were very successful, and much has improved within certain areas thanks to the support of these earmarked funds. Common Development Fund: CDF funding serves an important function in supporting development projects, but the district reports that only half of the promised CDF funding is actually received. Of the projects submitted to the CDF for approval each year, only some are selected, and the district has to come up with other sources of funds to support its projects or the projects are put on hold. Vision2020 Umurenge is a welcomed initiative, but 50 percent of CDF funding going to districts will now be earmarked for Vision2020related initiatives, further undercutting funding for other development projects. Currently CDFsupported projects include solar energy to six sector offices, terracing, constructions of two roads, construction of two sector offices, construction of four bridges, and construction of two markets. The following projects were submitted for approval for next year: infrastructure program, Ubudehe program, and a program for radical terracing. Revenue and Resources The district generates revenue through decentralized taxes, such as patent taxes, rental taxes, property taxes, and other administrative charges. Also, a main source of resources is donors and the central government. Although the district reports that a major challenge is to effectively and consistently ensure that its limited tax base is paying its taxes and that its staff receives proper training, the district has seen an improvement in its ability to generate revenues since 2006. But still, revenue sources are unreliable as the district lacks a complete database of all taxpayers, and delays in donor and government funds further hinder proper planning and budgeting. Rulindo Budget Allocation Snapshot Rulindo allocated 12 percent and 30 percent of its recurrent resources in 2006 and 2007, respectively, to the education sector, according to its MTEF. This was lower than the national average of 22 percent in 2006, but higher than national average of 24 percent in 2007. Rulindo allocated 33 percent and 12 percent of its recurrent resources in 2006 and 2007 to the health sector, according to its MTEF. This is a higher than at the national level (where the government allocated 7 percent and 8 percent of recurrent expenditures in 2006 and 2007). Rulindo allocated 5 percent and 2 percent of its recurrent resources in 2006 and 2007 to the infrastructure sector, according to its MTEF. Although development spending data are limited, in 2007 90 percent of its MTEF development spending was allocated to infrastructurerelated projects. Major projects were construction of classrooms and building of houses for genocide survivors and vulnerable families. About 2 percent of this spending went to energyrelated projects. Rulindo allocated 10 percent of its recurrent resources in 2006 and 2007 to the agriculture sector, according to its MTEF. This is higher than at the national level (where the government allocates about 2 percent of recurrent spending to agriculture). 268 World Bank Working Paper 3. Successes and Remaining Challenges at the District Level The Greatest Needs and Challenges The greatest challenges of decentralization at its initial phase included the increase in district responsibilities, but with insufficient funds to fully and properly execute those responsibilities. Although this remains a challenge, the dialogue with government has led to increased funding in some areas, including block transfers, increased earmarked transfers, and in 2008 the addition of certain subsidies. But development projects remain underfunded, and the CDF funding is not sufficient to meet needs. Consequently, 30 percent of block transfers are currently being used to support underfunded development projects. Unpredictable expenses, such as those related to natural disasters, are hard to foresee and also often put extra strain on the ability to execute properly. For example, following the earthquake in early 2007, several homes were destroyed. Severe flooding resulting in 28 deaths has also caused unexpected needs when more than 35 households were destroyed. The district also needs further help to train its staff properly. Urgent areas of capacitybuilding needs include training in taxes and auditing. Progress and Improvements to Date Despite many challenges, the improvements since decentralization are noted in several areas. The IMIHIGO has served the district well, and strong dialogue with the central government has led to improvements in certain areas, including increased funding to support increased responsibilities on the district level. The population is more fully engaged in its own decisionmaking process on district, sector, cell, and umudugudu (village) levels. Coordination has increased with the central government and within the district itself, and consequently, processes were streamlined. A simple, practical example includes the reduction of meetings that took away capacity from daytoday implementation efforts. Sectors were strengthened through increased financing and staffing capacity. Although many challenges still remain, including lack of funding and lack of capacity, the overall feedback from the district is that decentralization was successful in empowering its population and its representatives. Hence, positive results in several areas in the short period of time since decentralization are clearly noted. This case study is based on data from the District of Rulindo and a group interview with Vice Mayor of Economic Affairs, Mzamwita Deo; Director of Good Governance, Rutware Joseph; Director of Health, Gender, and Children, Fecericie Uwamurera; Director of Land, Urbanization, Habitat, and Infrastructure, Joseph Manzi Migeri; Planning Officer, Augustin Dusengimana; District Land Officer, Teophile Mutanganda; and Director of Finance, Emmanuel Rutaremara. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 269 Appendix D. Map of Rwanda Source: World Bank, January 2010. 270 nk World Ban Working Paper E. ucture in Rw Appendix E Decentralization Stru wanda am, v. Aug 2007, p Source: Vision 2020 Umurenge Progra page 4. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 271 Appendix F. Available Data Sets These data sets, exclusively based on MINECOFIN sources, contain the core data used for this written report. Because of the comprehensiveness of these data sets, they are not included in the report, but they can be obtained upon request in a separate Excel file. They can be used as a printed document or for additional analysis when used in their original Excel format. Additional data used for the public expenditure reviews (chapter 5) can be obtained upon request from the responsible ministry. Data Set 1a: SUMMARY PAGE of All Line Ministries' Expenditures in Real Terms Data Set 1b: MASTER FILE ­ in Nominal Terms Data Set 2: MINISTRY RECURRENT ­ 2004 Data Set 3: MINISTRY RECURRENT ­ 2005 Data Set 4: MINISTRY RECURRENT ­ 2006 Data Set 5: MINISTRY RECURRENT ­ 2007 Data Set 6: DISTRICT BUDGET ANALYSIS ­ CLASSIFICATION of EXPENDITURES 2004 Data Set 7: DISTRICT BUDGET ANALYSIS ­ CLASSIFICATION of EXPENDITURES 2005 Data Set 8: DISTRICT BUDGET ANALYSIS ­ CLASSIFICATION of EXPENDITURES 2006 Data Set 9: DISTRICT BUDGET ANALYSIS ­ CLASSIFICATION of EXPENDITURES 2007 Data Set 10: SUMMARY OF CAPITAL EXPENDITURES 2004­07 ­ Budget and Actuals Data Set 11: DEVELOPMENT BUDGET LAW (ORIGINAL) 2004 Data Set 12: DEVELOPMENT BUDGET LAW (ORIGINAL) 2005 Data Set 13: DEVELOPMENT BUDGET LAW (ORIGINAL) 2006 Data Set 14: DEVELOPMENT BUDGET LAW (ORIGINAL) 2007 Data Set 15: DEVELOPMENT BUDGET (Revised and Actual) 2004 Data Set 16: DEVELOPMENT BUDGET (Revised and Actual) 2005 Data Set 17: CEPEX ­ DEVELOPMENT BUDGET (Revised and Actual) 2006 Data Set 18: CEPEX ­ DEVELOPMENT BUDGET (Revised and Actual) 2007 Data Set 19: SUMMARY OF RESOURCES NET 2004­07 Data Set 20: RESOURCES NET 2004 Data Set 21: RESOURCES NET 2005 Data Set 22: RESOURCES NET 2006 272 World Bank Working Paper Data Set 23: RESOURCES NET 2007 Data Set 24: CEPEX Sorted by DONOR 2006 Data Set 25: CEPEX Sorted by DONOR 2007 Data Set 26: CEPEX ­ List and Overview of Multilateral Donor Projects 2006­07 Data Set 27: CEPEX ­ List and Overview of Bilateral and UN Donor Projects 2006­07 Data Set 28: BUDGET SUPPORT OVERVIEW 2004­07 Data Set 29: GENERAL BUDGET SUPPORT Commitment vs. Disbursements 2004­07 Data Set 30: CDF Overview (partial) Data Set 31: SUMMARY OF ALL CATEGORIES OF NET RESOURCES Data Set 32: KEY MACROECONOMIC INDICATORS Data Set 33: GDP Nominal 1999­2007 Data Set 34: GDP Real 1999­2007 Data Set 35: GDP Deflator 1999­2007 Data Set 36: CPI and GDP Deflator Indexes References Agriculture PER (2007). Rwanda Public Expenditure Review Agriculture Martin Fowler, Claver Garisabo, Sam Kanyarukiga, Augustin Mutijima. Kigali: Ministry of Agriculture and Animal Resources. April 2007. Akroyd, S. and Smith, L. (2006). Review of public spending to agriculture. Joint DFID/World Bank study. January final draft. Oxford Policy Management. Celasun O. and Walliser J. (2007). Predictability of Aid: Do Fickle Donors Undermine Economic Development? Washington, DC: International Monetary Fund and the World Bank. CEPEX (2006). Executed Development Budget. Kigali: Ministry of Finance and Economic Planning. CEPEX (2007). Executed Development Budget. Kigali: Ministry of Finance and Economic Planning. DHS (2005). Rwanda Demographic and Health Survey 2005. National Institute of Statistics of Rwanda (NISR) and ORC Macro. 2006. Calverton, Maryland, U.S.A.: NISR and ORC Macro. DHS (2007). Rwanda Interim Demographic and Health Survey 200708. Ministry of Health (MOH) [Rwanda], National Institute of Statistics of Rwanda (NISR), and ICF Macro. 2009. Calverton, Maryland, U.S.A.: MOH, NISR, and ICF Macro. EDPRS (20082012). Economic Development and Poverty Reduction Strategy, Kigali: Government of Rwanda, 2008. Education PER (2007). Rwanda Education Public Expenditure Report, Russell Craig and Paul Blay, Claudine Mbaraka, and Douglas Kigabo for Ministry of Education. Kigali. July 2007. EICV I (20002001). The First Household Living Conditions Survey, Kigali: National Statistics Institute of Rwanda, 20002001. EICV II (20052006). The Second Household Living Conditions Survey, Kigali: National Statistics Institute of Rwanda, 20052006. Fan, S., and N. Rao (2003). Public spending in developing countries: trends, determination and impact. EPTD Discussion Paper No. 99. IFPRI. Fan, S. and Saurkar, A (2006). Public spending in developing countries: trends, determination and impact (mimeo). Foster M. and Heller P.S. (2007). Managing the risks associated with aid increases in Rwanda. Kigali: Department for International Development. Health PER (2008). Rwanda Public Expenditure Review for the Health Sector 20062007 Data and Trends from 20022007. Kigali, Ministry of Health. Draft, April 2008. 273 274 World Bank Working Paper IMF Finance and Development (2007). Finance and Development News Letter. Washington D.C.: International Monetary Report, December 2007, Volume 44, Number 4, page 20. Gupta S., Powell R. and Yang Y. (2006). Macroeconomic Challenges of Scaling up Aid in Africa. Washington D.C.: International Monetary Fund. Heller P.S. (2005). Understanding Fiscal Space. Washington D.C.: International Monetary Fund, IMF Policy Discussion Paper PDP/05/4. IMF (2002). Rwanda: 2002 Article IV Consultation and Requests for a New Poverty Reduction and Growth Facility Arrangement and for Additional Interim Assistance Under the Enhanced Initiative for Heavily Indebted Poor Countries. Washington D.C.: International Monetary Fund, Country Report No. 02/204. IMF (2005a). Rwanda: Enhanced Initiative for Heavily Indebted Poor Countries Completion Point Document. Washington D.C.: International Monetary Fund, Country Report No. 05/173. IMF (2005b). Rwanda: Fourth Review Under the ThreeYear Agreement Under the Poverty Reduction and Growth Facility. Washington D.C.: International Monetary Report, Country Report No. 05/171. IMF (2005c). Rwanda: Fifth Review Under the ThreeYear Arrangement Under the Poverty Reduction and Growth Facility. Washington D.C.: International Monetary Report, Country Report 05/333. IMF Finance and Development (2007). Finance & Development News Letter. Washington D.C.: International Monetary Report, December 2007, Volume 44, Number 4, page 20. IMF (2007a). Rwanda: 2006 Article IV Consultation, First Review Under the ThreeYear Arrangement Under the Poverty Reduction and Growth Facility, and Request for Waiver of Nonobservance of Performance Criteria. Washington D.C.: International Monetary Fund, Country Report No. 07/80. IMF (2007b). Second review under the threeyear arrangement under the poverty reduction and growth facility, request for waiver of nonobservance of performance criteria. Washington D.C.: International Monetary Fund, Country Report No. 07/233. JEEAR (1996). "Joint Evaluation of Emergency Assistance to Rwanda", Steering Committee for the Joint Evaluation of Emergency Assistance to Rwanda, March 1996 Rwangombwa (2005) "Budget Support Trends", power point presentation in 2005 by John Rwangombwa, Secretary General and Secretary to the Treasury at the development partners retreat. Akagera: Ministry of Finance and Economic Planning. Mackinnon J., Thomson A., Hakizinka I. and Rugwabiza L. (2003). The impact of increases in public expenditures on poverty in Rwanda ­ PSIA study. Kigali: Ministry of Finance and Economic Planning. Mehta, M. (2004). Sector Finance and Resource Flows for Water Supply, A pilot Application for Kenya. ------ (2004). National Rural Water Supply and Sanitation Program. MINAGRI/NEPAD Secretariat. (2006). CAADP Implementation in Rwanda under EDPRS: review and stocktaking report ongoing development efforts and their alignment with CAADP targets and principles. Kigali: Ministry of Agriculture and Animal Resources and New Partnership for Africa's Development. Mimeo. Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform 275 MINAGRI/RADA. (2006b). Strategy for increasing fertiliser use in Rwanda. Mimeo. Kigali: Ministry of Agriculture and Animal Resources and Rwanda Agriculture Development Authority. Draft. MINECOFIN (2004a). Organic Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2004b). Executed Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2004c). Rwanda Development Indicators. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2005a). Organic Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2006). Joint Budget Support Review III Report. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2006a). Organic Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2006a). Draft law determining the state of finances for the 2007 fiscal year. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2006b). Executed Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2007). Draft Economic Development and Poverty Reduction Paper. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2007). Unpublished preliminary MINECOFIN GDP estimates. Mimeo. Draft for discussion. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2007a). Organic Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2007b). Executed Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN (2007e). Revised Budget Law. Kigali: Ministry of Finance and Economic Planning. MINECOFIN Macroeconomic Tables (2007). Kigali: Ministry of Finance and Economic Planning. MINECOFIN Flash Reports (200206). Government Operation Tables 200206. Kigali: Ministry of Finance and Economic Planning. NBR (2006). National Bank of Rwanda Annual Report 2005. Kigali: National Bank of Rwanda. June, 2006. NHA (2006). National Health Accounts Rwanda 2006 with HIV/AIDS, Malaria, and Reproductive Health Subaccounts. Health Systems 20/20, June 2008. Bethesda, MD: Health Systems 20/20 project, Abt Associates Inc. NISR (2006). Preliminary poverty update report: integrated living conditions survey 2005/06. First draft for consultations. December. Kigali: National Institute of Statistics of Rwanda. NISR (2007). National Accounts, 19992005.Estimates based on 2001 benchmark. Kigali: National Institute of Statistics of Rwanda. OECD (2007). www.oecd.org. Organisation for Economic Cooperation and Development. 276 World Bank Working Paper Rwanda General Population and Housing Census (2002). Kigali: Ministry of Finance and Economic Planning and the National Census Commission. Sayinzoga, Kampeta (2007). "Should Fiscal Policy Make Space for Aid? A Case Study of Rwanda (2000­2006)." Based primarily on IMF data and government of Rwanda data. University of Nottingham. Social Protection PER (2006). Public Expenditure Review Government and Donor Expenditure on Social Protection, Mick Foster, Mick Foster Economics Limited for Ministry of Local Government, Good Governance, Community Development, and Social Affairs. Kigali. November 2006. World Bank (2005). Poverty Reduction Support Grant 2: Program Document. Washington D.C.: World Bank, Report No.33798RW. World Bank (2006). Poverty Reduction Support Grant 3: Program Document. Washington D.C.: World Bank, Report No.37959RW. World Bank (2007a). Fiscal Policy for Growth and Development: Further Analysis and Lessons from Country Case Studies. Washington D.C.: World Bank. World Bank. (2007b). Promoting propoor agricultural growth in Rwanda: challenges and opportunities. Draft final dissemination workshop. Sustainable Development Network, Africa Region. World Bank and IMF (2006). Rwanda Joint Debt Sustainability Analysis. Washington D.C.: World Bank. WSS PER (2006). Water and Sanitation Sector Performance Report, Stephen Hitimana, Consultant, for Ministry of Lands, Environment, Forestry, Water, and Natural Resources. Kigali. June 2006. Eco-Audit Environmental Benefits Statement The World Bank is committed to preserving Endangered Forests and natural resources. We print World Bank Working Papers and Country Studies on postconsumer recycled paper, processed chlorine free. The World Bank has formally agreed to follow the recommended standards for paper usage set by Green Press Initiative--a nonprofit program supporting publishers in using fiber that is not sourced from Endangered Forests. For more information, visit www.greenpressinitiative.org. In 2008, the printing of these books on recycled paper saved the following: Net Greenhouse Trees* Solid Waste Water Total Energy Gases 289 8,011 131,944 27,396 92 mil. *40 feet in height and Pounds CO2 6­8 inches in Pounds Gallons BTUs Equivalent diameter Budgeting for Effectiveness in Rwanda: From Reconstruction to Reform is part of the World Bank Working Paper series. These papers are published to communicate the results of the Bank's ongoing research and to stimulate public discussion. This paper reviews Rwanda's general budget support rele- vance, rationale, and outstanding challenges by providing a historical background of budget support, assessing progress in budget support-related processes and practices, reviewing Rwanda's economic and structural reforms and budget support predictability trends, and assessing the net resources available to the government of Rwanda. In this context, the paper reviews resource allocations and spending among the government's ministries, including its transfers to districts, summarizes in-depth studies undertaken in priority sectors, addresses challenges, and offers conclusions. This working paper was produced as part of the World Bank's Africa Region Health Systems for Outcomes (HSO) Program. The Program, funded by the World Bank, the Government of Norway, the Government of the United Kingdom and the Global Alliance for Vaccines and Immunization (GAVI), focuses on strengthening health systems in Africa to reach the poor and achieve tangible results related to Health, Nutrition and Population. The main pillars and focus of the program center on knowledge and capacity building related to Human Resources for Health, Health Financing, Pharmaceuticals, Governance and Service Delivery, and Infrastructure and ICT. More information as well as all the products produced under the HSO program can be found online at www.worldbank.org/hso. World Bank Working Papers are available individually or on standing order. The World Bank Working Paper series is also available online through the World Bank e-library (www.worldbank.org/elibrary). ISBN 978-0-8213-8558-6 THE WORLD BANK SKU 18558