Report No. 39710-AO Angola Public Expenditure Review (In Two Volumes) Volume II: Sectoral Review December 20, 2007 A Joint Report by: The World Bank and European Commission (EC) Africa Region Food and Agriculture Organization (FAO) Macroeconomics I United Nations Children's Fund (UNICEF) United Nations Development Program (UNDP) World Health Organization (WHO) Document of the World Bank Currency Equivalents Currency Unit: Angolan Kwanza (Kz) USD= 79.845 Kwanza (Exchange Rate Effective as of April 30, 2007) Weights and Measures Metric System Fiscal Year of Budget: January 1 ­ December 31 Vice President: Obiageli K. Ezekwesili Country Director: Michael Baxter Sector Manager: John Panzer Task Team Leader: Francisco Galrão Carneiro Report No. 39710 - AO ANGOLA PUBLIC EXPENDITURE REVIEW (In Two Volumes) Volume 2: Sectoral Review DECEMBER 20, 2007 A Joint Report by: The World Bank and European Commission (EC) Africa Region Food and Agriculture Organization (FAO) Macroeconomics 1 United Nations Children's Fund (UNICEF) United Nations Development Program (UNDP) World Health Organization (WHO) Document of the World Bank Currency Equivalents Currency Unit: Angolan Kwanza (Kz) USD= 79.845 Kwanza (Exchange Rate Effective as of April 30, 2007) Weights and Measures Metric System Fiscal Year of Budget: January 1 ­ December 31 Vice President: Obiageli K. Ezekwesili Country Director: Michael Baxter Sector Manager: John Panzer Task Team Leader: Francisco Galrão Carneiro TABLE OF CONTENTS INTRODUCTION ................................................................................................................... 1 A. WHAT IS THE OBJECTIVE OF THIS REPORT .............................................................................1 B. WHAT IS THE RELATION OF THE PER WITH THE PEMFAR............................................................2 C. HOW THE REPORT IS ORGANIZED .......................................................................................2 D. THE REPORT'S FIVE CROSS-CUTTING MESSAGES .....................................................................3 Budgeting within a Multi-Year Perspective..........................................................................3 Improving the Efficiency of Public Spending .......................................................................4 Enhancing the Quality of the Data on Social Indicators and the Budget..................................5 Equity in the Budget .......................................................................................................6 Further Progress with Decentralization ..............................................................................6 CHAPTER I ........................................................................................................................... 8 MACROECONOMIC CONTEXT AND UPDATE ON THE REFORM OF FISCAL MANAGEMENT........ 8 A. MACROECONOMIC CONTEXT: BENEFITING FROM A WINDFALL .......................................................8 B. GROWING REVENUES WILL REQUIRE BETTER MEDIUM-TERM PLANNING......................................... 12 C. THE ROLE OF THE GOVERNMENT'S POVERTY REDUCTION STRATEGY.............................................. 15 D. THE BI-ANNUAL ECONOMIC PROGRAMS .............................................................................. 17 E. THE 2006 AND 2007 BUDGETS ....................................................................................... 20 F. FUNCTIONAL CLASSIFICATION OF EXPENDITURES ................................................................... 22 G. ECONOMIC CLASSIFICATION OF EXPENDITURES ..................................................................... 24 H. REFORM OF FISCAL MANAGEMENT: THE REMAINING CHALLENGES................................................ 26 Budget Preparation....................................................................................................... 29 Budget Execution ......................................................................................................... 29 Budget Monitoring........................................................................................................ 31 Ring Fencing and Phasing Out of QFAs ............................................................................ 32 CHAPTER II........................................................................................................................ 37 PUBLIC SPENDING IN EDUCATION .................................................................................... 37 A. RECENT PERFORMANCE IN EDUCATION ................................................................................ 37 Literacy, access and enrolment ...................................................................................... 38 Efficiency of Educational System .................................................................................... 41 B. ANALYSIS OF EQUITY ISSUES........................................................................................... 45 Defining equity ............................................................................................................ 45 Repetition and Completion Rates .................................................................................... 46 Gender inequalities....................................................................................................... 48 C. RECENT EXPENDITURE TRENDS ........................................................................................ 51 Trends in government and externally financed education expenditure.................................. 53 Education expenditure by functional and economic classification ......................................... 55 Classification of expenditures per provinces ..................................................................... 58 D. INSTITUTIONAL PLANNING AND BUDGETING .......................................................................... 61 Weaknesses in the budget process.................................................................................. 62 E. THE VALUE FOR MONEY CHAIN......................................................................................... 62 Teachers..................................................................................................................... 63 Effectiveness of the Angolan education system................................................................. 66 F. POLICY RECOMMENDATIONS............................................................................................ 67 CHAPTER III ...................................................................................................................... 78 PUBLIC SPENDING IN THE HEALTH SECTOR ...................................................................... 78 A. THE CONTEXT: MULTIDIMENSIONAL CHALLENGES................................................................... 78 Challenges at the health outcome level ........................................................................... 78 Challenges at the health service coverage level. ............................................................... 82 The country's response to the challenges......................................................................... 82 B. THE STRUCTURE OF THE HEALTH SECTOR ............................................................................ 83 C. THE BUDGET CYCLE IN THE HEALTH SECTOR......................................................................... 85 Proposed planning calendar for the health sector .............................................................. 89 D. THE STRUCTURE OF EXPENDITURES IN HEALTH...................................................................... 91 E. HEALTH FINANCING...................................................................................................... 96 F. DECENTRALIZATION AND SERVICE DELIVERY ........................................................................ 99 G. EFFECTIVENESS AND EFFICIENCY OF SERVICE DELIVERY ......................................................... 101 Scenarios for Scaling Up Service Coverage..................................................................... 112 -i- H. PROPOSED RECOMMENDATIONS...................................................................................... 115 Inter-sectoral allocation .............................................................................................. 115 Intra-sectoral allocation .............................................................................................. 116 The budget preparation process ................................................................................... 117 Financial management ................................................................................................ 118 Monitoring of budget execution .................................................................................... 118 Further decentralization .............................................................................................. 119 Primary network financing ........................................................................................... 119 Agenda for Reform ..................................................................................................... 120 CHAPTER IV ..................................................................................................................... 122 PUBLIC SPENDING IN AGRICULTURE............................................................................... 122 A. SECTOR BACKGROUND ................................................................................................ 122 Productive potential.................................................................................................... 122 Socio-geographic characteristics................................................................................... 124 Cultivated area, production and yielding........................................................................ 125 Supporting services for production, trading and competitiveness ...................................... 126 B. THE BUDGET FOR AGRICULTURE ..................................................................................... 127 Importance of the State in agriculture........................................................................... 127 Concentration of budget execution................................................................................ 128 Programs .................................................................................................................. 129 Public Investment Projects .......................................................................................... 131 Spatial Distribution of Agriculture Investment Projects .................................................... 132 Thematic Distribution of Agriculture Investment Projects ................................................. 132 Investment Projects in Fisheries................................................................................... 133 C. ALIGNMENT OF BUDGET AND SECTORAL PLANS.................................................................... 134 D. POLICY OPTIONS FOR THE FUTURE .................................................................................. 135 CHAPTER V....................................................................................................................... 137 THE DECENTRALIZATION AGENDA................................................................................... 137 A. THE EXISTING LEGAL AND INSTITUTIONAL FRAMEWORK ......................................................... 137 The Political and Institutional Environment..................................................................... 137 The Organization of Civil Society .................................................................................. 139 B. CONCEPTUAL ASPECTS AND INTERNATIONAL EXPERIENCES ...................................................... 139 Why Transfer Resources.............................................................................................. 140 A Brief Account of Other Experiences ............................................................................ 143 C. A PROSPECTIVE DECENTRALIZATION AGENDA ..................................................................... 147 Phase 1: Short-term Administrative Decentralization....................................................... 148 The Legal Framework.................................................................................................. 149 The Political-Institutional Framework ............................................................................ 150 The Fiscal and Financial Framework .............................................................................. 152 Phase 2: Medium-term Administrative Decentralization ................................................... 154 The Legal Framework.................................................................................................. 154 The Political-Institutional Framework ............................................................................ 154 The Fiscal and Financial Framework .............................................................................. 155 Phase 3: Fiscal Decentralization ................................................................................... 156 The Legal Framework.................................................................................................. 157 The Institutional Framework ........................................................................................ 157 The Fiscal and Financial Framework .............................................................................. 158 LIST OF BOXES BOX 1. 1: WHY IT IS NECESSARY TO MOVE TOWARD A MULTI-YEAR APPROACH TO BUDGETING........................ 11 BOX 1. 2: SUMMARY OF PROGRESS ON SELECTED RECOMMENDATIONS OF THE PEMFAR ................................ 28 BOX 1. 3.: PHASES OF THE "RING-FENCING" PROCESS ...................................................................... 33 -ii- BOX 3. 1: ESSENTIAL PACKAGE OF MOTHER AND CHILD CARE AND SERVICES AT THE PRIMARY LEVEL.................. 103 BOX 3. 2: ESSENTIAL PACKAGE OF MOTHER AND CHILD CARE AND SERVICES TO BE APPLIED BY MOBILE AND ADVANCE HEALTH TEAMS.............................................................................................................. 103 BOX 3. 3: ESSENTIAL PACKAGE OF MOTHER AND CHILD CARE FOR COMMUNITY AND NEIGHBOURHOOD VOLUNTARY HEALTH AGENTS. ........................................................................................................... 104 LIST OF FIGURES FIGURE 1. 1: OIL PRODUCTION IN ANGOLA, 2000-2020..................................................................... 10 FIGURE 1. 2: OIL REVENUES UNDER DIFFERENT PRICE SCENARIOS ......................................................... 11 FIGURE 1. 3: FUNCTIONAL DISTRIBUTION OF EXPENDITURES, 2004-07................................................... 22 FIGURE 2. 1: GROSS ENROLMENT RATES FROM 1999 TO 2003 IN THE THREE ............................................ 40 FIGURE 2. 2: REPETITION AND DROP-OUT RATES IN THE 1ST, 2ND AND 3RD LEVELS OF PRIMARY EDUCATION FROM 1999 TO 2002............................................................................................................... 42 FIGURE 2. 3: PRIMARY COMPLETION RATES IN 2002 IN SUB-SAHARAN AFRICA (SSA) COUNTRIES.................... 43 FIGURE 2. 4: PUPILS REACHING GRADE 5 (% OF COHORT) IN 2002 IN .................................................... 44 FIGURE 2. 5: GROSS ENROLMENT RATES IN THE 1ST LEVEL OF PE PER PROVINCE IN 1999 AND 2003. ............... 46 FIGURE 2. 6: GROSS ENROLMENT RATES PER GENDER IN 1999, 2001 AND 2003 PER LEVEL OF PRIMARY EDUCATION ................................................................................................................................. 49 FIGURE 2. 7: GOVERNMENT OVERALL AND EDUCATION EXPENDITURE AS A SHARE OF GDP. ............................. 52 FIGURE 2. 8: PROPORTION OF EXPENDITURE BY ECONOMIC CLASSIFICATION IN TOTAL EDUCATION EXPENDITURE FROM 1999 TO 2003............................................................................................................... 58 FIGURE 2. 9: PRIMARY PUPIL-TEACHER RATIO IN 2002 FOR GROUPS OF COUNTRIES AND SELECTED SUB-SAHARAN COUNTRIES.................................................................................................................... 64 FIGURE 2. 10: VARIATIONS IN PTR'S BETWEEN PROVINCES IN 2001....................................................... 65 FIGURE 2. 11: VARIATIONS IN PTR'S BETWEEN PROVINCES IN 2003....................................................... 66 FIGURE 3. 1: IMR IN ANGOLA AND SUB-SAHARAN AFRICA.................................................................... 79 FIGURE 3. 2: UNDER-FIVE MORTALITY RATES IN ANGOLA AND SUB-SAHARAN AFRICA .................................... 79 FIGURE 3. 3: CAUSES OF UNDER FIVE MORTALITY, LUANDA 2001............................................................ 79 FIGURE 3. 4: UNDER-FIVE MORTALITY BY REGION .............................................................................. 80 FIGURE 3. 5: BREAKDOWN OF THE HEALTH SECTOR LABOR FORCE: 1999 AND 2005 ..................................... 92 FIGURE 3. 6: IMMUNIZATION RATES ............................................................................................ 106 FIGURE 3. 7 : ITN USE ........................................................................................................... 110 FIGURE 3. 8: BREASTFEEDING ................................................................................................... 110 FIGURE 3. 9: ANTENATAL CARE .................................................................................................. 111 FIGURE 3. 10: IMMUNIZATION................................................................................................... 111 FIGURE 3. 11: DELIVERY SERVICES ............................................................................................. 111 FIGURE 3. 12....................................................................................................................... 112 LIST OF TABLES TABLE 1. 1: MACROECONOMIC FRAMEWORK, 2003 - 2007 ....................................................................9 TABLE 1. 2: BILATERAL SOURCES OF FINANCING FOR THE GOVERNMENT'S ECONOMIC PROGRAM ....................... 19 TABLE 1. 3: MACROECONOMIC ASSUMPTIONS USED IN THE 2007 BUDGET................................................. 21 TABLE 1. 4: ANGOLA: FUNCTIONAL CLASSIFICATION OF GOVERNMENT EXPENDITURES, 2003-2005 ................ 23 TABLE 1. 5: ECONOMIC CLASSIFICATION OF CURRENT EXPENDITURES, 2003-06......................................... 25 TABLE 2. 1: ABSOLUTE NUMBER OF PUPILS PER LEVEL OF EDUCATION IN 1999 AND 2003, % ......................... 39 TABLE 2. 2: GROSS ENROLMENT RATES FOR THE 1ST, 2ND AND 3RD LEVELS OF PRIMARY EDUCATION FROM 1999 TO 2003. ......................................................................................................................... 40 TABLE 2. 3: GROSS ENROLMENT RATES FOR SECONDARY, MEDIUM......................................................... 40 TABLE 2. 4: NET ENROLMENT RATES FOR THE 1ST, 2ND AND 3RD LEVELS OF PRIMARY .................................. 41 TABLE 2. 5: REPETITION AND DROP-OUT RATES FOR THE 1ST, 2ND AND 3RD LEVELS OF PRIMARY EDUCATION FROM 1999 TO 2002............................................................................................................... 42 TABLE 2. 6: TRANSITION AND SURVIVAL RATES FOR THE 1ST, 2ND AND 3RD LEVELS OF PRIMARY ..................... 44 TABLE 2. 7: GROSS ENROLMENT RATES FROM 1999 TO 2003 PER PROVINCE IN THE 1ST OF PE. ...................... 45 TABLE 2. 8: REPETITION RATES PER PROVINCE IN THE 1ST, 2ND AND 3RD LEVELS OF PE IN 2000 AND 2002....... 47 TABLE 2. 9: COMPLETION RATES PER PROVINCE IN THE 1ST, 2ND AND 3RD LEVELS OF PE IN 1999 AND 2002. .... 47 TABLE 2. 10: GROSS ENROLMENT RATES PER GENDER IN 1999, 2001 AND 2003 PER LEVEL OF PRIMARY EDUCATION ................................................................................................................................. 48 TABLE 2. 11: REPETITION, DROP-OUT AND COMPLETION RATES FROM 1999 TO 2002 PER GENDER AND LEVEL OF PRIMARY EDUCATION......................................................................................................... 49 TABLE 2. 12: GROSS ENROLMENT RATES IN 2002 PER PROVINCE AND GENDER IN THE THREE LEVELS OF PRIMARY EDUCATION PER PROVINCE. ................................................................................................. 50 TABLE 2. 13: COMPLETION RATES IN 2002 PER PROVINCE AND GENDER IN THE THREE LEVELS OF PRIMARY EDUCATION PER PROVINCE. ............................................................................................................... 51 -iii- TABLE 2. 14: GDP, TOTAL PUBLIC SPENDING, EDUCATION SPENDING IN MILLION USD AND EDUCATION SPENDING AS A PROPORTION OF TOTAL PUBLIC SPENDING AND AS A PROPORTION OF GDP. .................................. 52 TABLE 2. 15: BUDGETS AND ACTUAL EXPENDITURE (MILLION USD) FROM 1999 TO 2003. ............................ 53 TABLE 2. 16: EXECUTION RATE OF BUDGETS BY FUNCTIONAL CLASSIFICATION FROM ..................................... 53 TABLE 2. 17: EXECUTION RATE OF BUDGETS BY FUNCTIONAL CLASSIFICATION FROM ..................................... 53 TABLE 2. 18: EXECUTION RATE OF BUDGETS IN PRIMARY AND SECONDARY EDUCATION BY FUNCTIONAL AND ECONOMIC CLASSIFICATION FROM 1999 TO 2003 (IN %).......................................................................... 55 TABLE 2. 19: TOTAL EXPENDITURE BY FUNCTIONAL CLASSIFICATION FROM 1999 TO 2003 (IN MILLION USD)...... 55 TABLE 2. 20: TOTAL EXPENDITURE BY FUNCTIONAL CLASSIFICATION FROM 2004 TO 2005 (IN MILLION USD) AND PROPORTION IN TOTAL EXPENDITURE BY FUNCTIONAL CLASSIFICATION............................................... 56 TABLE 2. 21: TOTAL RECURRENT EXPENDITURE BY FUNCTIONAL CLASSIFICATION FROM 1999 TO 2003............... 57 TABLE 2. 22: TOTAL RECURRENT EXPENDITURE BY FUNCTIONAL CLASSIFICATION FROM 2004 TO 2005 AND PROPORTION OF RECURRENT EXPENDITURE IN TOTAL EXPENDITURE BY FUNCTIONAL CLASSIFICATION............ 57 TABLE 2. 23: EXPENDITURES BY DETAILED ECONOMIC CLASSIFICATION FROM 1999 TO 2003 (IN MILLION USD). . 58 TABLE 2. 24: SHARE OF TOTAL EDUCATION EXPENDITURE PER PROVINCE FROM 1999 TO 2005. ...................... 59 TABLE 2. 25: EDUCATION UNIT EXPENDITURE, GRADE 1-8, PER PROVINCE FROM 1999 TO 2003..................... 61 TABLE 2. 26: PUPILS PER TEACHER, TEACHER'S PER CLASS AND PUPILSPER ................................................ 64 TABLE 2. 27: EFFECTIVENESS INDICATORS FOR SCHOOL EDUCATION, 2000 AND 2002. ................................ 67 TABLE 3. 1: KEY HEALTH OUTCOME INDICATORS ................................................................................ 78 TABLE 3. 2: HEALTH SERVICE INDICATORS...................................................................................... 82 TABLE 3. 3: BUDGET EXECUTION RATE FOR GOODS AND SERVICES IN MUNICIPAL, ........................................ 88 TABLE 3. 4: BUDGET EXECUTION RATE BY ECONOMIC CATEGORY............................................................. 88 TABLE 3. 5: EXPENDITURES (2000-05) AND BUDGET (2006) PER ECONOMIC CATEGORY............................... 91 TABLE 3. 6: ANALYSIS OF SPENDING BY LEVEL OF CARE....................................................................... 93 TABLE 3. 7: PROVINCIAL EXPENDITURES (US$ MILLION) ..................................................................... 93 TABLE 3. 8: NUMBER OF DOCTORS AND NURSES IN THE NHS PER PROVINCE, 2005...................................... 95 TABLE 3. 9: HEALTH SECTOR SPENDING PER PROGRAM (US$ `000) ........................................................ 96 TABLE 3. 10: GDP, TOTAL PUBLIC EXPENDITURES AND HEALTH EXPENDITURES ........................................... 97 TABLE 3. 11: BUDGET AND TOTAL PUBLIC HEALTH EXPENDITURES PER CAPITA (US$ `000)A, ........................... 97 TABLE 3. 12: PUBLIC HEALTH EXPENDITURES AND HEALTH INDICATORS IN SOME SADC COUNTRIES................... 98 TABLE 3. 13: AVERAGE EXPENDITURES INCURRED BY INDIVIDUALS FOR A FULL EPISODE OF A COMMON ILLNESS (IN KWANZAS)..................................................................................................................... 99 TABLE 3. 14: PROPORTION OF SPENDING IN THE PROVINCES, LUANDA, AND CENTRAL UNITS.......................... 100 TABLE 3. 15: KEY HIGH IMPACT HEALTH INTERVENTIONS BY SERVICE DELIVERY LEVEL ................................. 107 TABLE 3. 16: COST AND IMPACT OF SCALING UP SERVICE COVERAGE .................................................... 113 TABLE 4. 1: PRODUCTIVE POTENTIAL ­ ANGOLA AND SELECTED COUNTRIES............................................. 123 TABLE 4. 2: ARABLE LAND FOR DRY CROPS ­ ANGOLA AND SELECTED COUNTRIES ..................................... 123 TABLE 4. 3: ARABLE LAND ­ ANGOLA AND SELECTED COUNTRIES ......................................................... 123 TABLE 4. 4: EVOLUTION OF RURAL AGRICULTURE PRODUCTION OF ANGOLA.............................................. 126 TABLE 4. 5: ECONOMIC CLASSIFICATION OF EXPENDITURES IN AGRICULTURE ­ 2003 - 2005 ....................... 128 TABLE 4. 6: IMPLEMENTED PROGRAMS IN AGRICULTURE, FISHERIES AND ENVIRONMENT FUNCTION - 2005 ........ 130 TABLE 4. 7: SOURCE OF FINANCING FOR AGRICULTURAL PROJECTS........................................................ 131 TABLE 4. 8: DISTRIBUTION OF INVESTMENT PROJECTS PER PROVINCE .................................................... 132 TABLE 4. 9: THEMATIC DISTRIBUTION OF INVESTMENT PROJECTS.......................................................... 133 TABLE 5. 1: A POSSIBLE DECENTRALIZATION AGENDA IN ANGOLA WITH 3 PHASES ..................................... 148 -iv- ACRONYMS AND ABBREVIATIONS AUPEC Aberdeen University Petroleum Economics Consultancy BCI Banco de Comércio e Indústria BNA Banco Nacional de Angola BPC Banco de Poupança e Crédito CPI Consumer Price Index CUT Conta Única do Tesouro DHS Demographic and Health Survey DMFAS Debt Management Financial and Analysis System DNC Direcção Nacional de Contatibilidade DNI National Tax Directorate DNPE Direcção Nacional do Patrimonio do Estado DNRH Direcção Nacional de Recursos Humanos DU Dependent Unit EC European Commission ECP Estratégia de Combate à Pobreza EFA Education for All FAO Food and Agriculture Organization of the United Nations FAS Social Action Fund FQ Financial Quotas GAI Gabinete de Apoio Informático GAR Gross Admission Rate GDP Gross Domestic Product GEPE Gabinete de Estudos, Planeamento e Estatística GER Gross Enrollment Rate GEREI Cabinet of Studies and Foreign Relations HIS Health Information System IBAM Brazilian Institute of Municipal Administration IMN Instituto Médio Normal IMT Instituto Médio Técnico INE Instituto Nacional de Estatística INEDA Instituto Nacional de Educação de Adultos INF National Inspectorate of Finance MAPESS Ministério de Administração Pública, Emprego, e Segurança Social MCR Modified Completion Rates MDG Millennium Development Goals MEC Ministry of Education and Culture MECANAGRO National Agricultural Mechanization Company MICS Multiple Indicator Cluster Survey MINADER Ministério da Agricultura e Desenvolvimento Rural MINFIN Ministry of Finance MINPET Ministry of Petroleum MINPLAN Ministry of Planning MOH Ministry of Health MTA Ministry of Territorial Administration MTEF Medium-Term Expenditure Framework NHS National Health System OGE Orçamento Geral do Estado PASS Programa de Apoio ao Sector da Saúde PE Primary Education PEMFAR Public Expenditure Management and Financial Accountability Report PHD Provincial Health Directorate PIP Public Investment Program PSA Production Sharing Agreement PTR Pupil-Teacher Ratio QFA Quasi-Fiscal Activity SADC Southern Africa Development Community SAI Supreme Audit Institution SIGFE Sistema Integrado para a Gestão das Finanças do Estado SINGERH Integrated System for the Management of Public Employees SSA Sub-Saharan Africa UNDP United Nations Development Program UNICEF United Nations Children's Fund WHO World Health Organization of the United Nations -v- WDI World Development Indicators PRSP Poverty Reduction Strategy Paper PPMRRP Priority Phase of the Multisector Rehabilitation and Reconstruction Program OECD Organization for Economic Cooperation and Development PRC People's Republic of China GOA Government of Angola CEM Country Economic Memorandum DPF Delegações Provinciais de Finanças DAGO Departamento de Administração e Gestão do Orçamento MMR Maternal Mortality Rate NBD National Budget Directorate INCA Instituto Nacional do Café UNESCO United Nations Education, Scientific and Cultural Organization PEDR Rural Extension and Development Program ADI Agrarian Development Institute -vi- ACKNOWLEDGEMENTS This report is the result of collaboration between the World Bank and a group of development partners of the Government of Angola that includes the United Nations Development Program (UNDP), the European Commission (EC), the Food and Agriculture Organization of the United Nations (FAO), the World Health Organization of the United Nations (WHO), and the United Nations Children's Fund (UNICEF). Useful information and insights were provided by government officials in many departments in the Ministry of Education, Ministry of Health, Ministry of Agriculture and Rural Development, Ministry of Territorial Administration, Ministry of Planning, and Ministry of Finance. The collaboration between the World Bank and the development partners was formally agreed through a Memorandum of Understanding signed by the heads of the delegations in Angola in June 2006. The main counterpart in the Government was Mr. Manuel Neto da Costa, Head of the Cabinet of Studies and Foreign Relations of the Ministry of Finance (GEREI ­ MINFIN). The joint PER team extends its appreciation to Ms. Ana Dias Lourenço, Minister of Planning, Mr. José Pedro de Morais, Minister of Finance, Mr. Manoel Neto da Costa, Head of the Cabinet of Studies and Foreign Relations of the Ministry of Finance (GEREI ­ MINFIN), and to the Directors of Planning of the Ministries covered in this report for their support and active collaboration. For the World Bank, the task was managed by Francisco Carneiro (Sr. Country Economist, AFTP1), who was also the primary author of the report. The task team included Stefania Abakerli (Local Development Specialist, LCSSO), Jean-Jacques De St. Antoine (Lead Operations Officer, AFTH1), Feng Zhao (Health Specialist, AFTH1), Xiaoyan Liang (Sr. Education Specialist, AFTH1), Eduardo de Sousa (Sr. Economist, AFTS1), Maria Teresa Benito-Spinetto (Research Analyst, AFTP1). Contributions were made by the Bank consultants Mafalda Duarte (Education), Rui Fuschini (Education) José Roberto Rodrigues Afonso, Thereza Lobo, and Rafael Barroso (all Decentralization). Mr. Victor Hugo and Ms. Henda Ducados (Executive Director and Deputy Directors of FAS, respectively) offered useful comments and suggestions on the background paper on decentralization. Useful discussions on decentralization were also held with Serdar Yilmaz (Sr. Social Development Economist, SDV) and Rodrigo Serrano-Berthet (Local Development Specialist, HDNSP). The team is grateful to the Social Development Department (SDV) for funding the consultants who worked on decentralization. The UNDP, jointly with UNCDF, provided important contributions in the area of decentralization and this report draws partially on conclusions and recommendations which are contained in the report entitled "Diagnóstico da Descentralização Fiscal em Angola" of August 2006 authored by Maria Cristina Mac Dowell, Érika Amorim Araújo, Alexandre Sobreira Cialdini, and Nicoletta Ferruglio. The counterpart official from the UNDP in the PER team was Mr. Alfredo Teixeira. -i- The European Commission contributed for the review of public expenditures in the health sector. The counterpart official from the EC in the PER team was Ms. Tamar Bello. The chapter on health draws substantially on the analysis and policy recommendations which are contained in the report entitled "Angola: Despesa Pública no Sector da Saúde 2000-2006" prepared by the Programa de Apoio ao Sector da Saúde (PASS) for the Ministry of Health and published by the Ministry of Health early in 2007. Additional contributions to the review of public spending in the health sector were provided by staffs from the World Health Organization (WHO), and the United Nations Children's Fund (UNICEF). The counterpart official at WHO was Mr. Ole Frank Nielsen and the counterpart official at UNICEF was Dr. Guy Clarisse. The Food and Agriculture Organization (FAO) contributed inputs for the review of public expenditures in the sector of agriculture. The chapter on agriculture draws on a background report prepared by Mr. Guillermo Woods (Consultant, FAO) entitled "Revisión de los Gastos y Inversiones Públicas Destinadas al Sector Agropecuário". Mr. Francisco Chimuco (FAO) provided useful assistance in the contacts with the authorities and mission preparation for the work associated with the agriculture sector. The counterpart official at FAO was Mr. Frits Ohler. The several missions that were held to Angola in connection with this report were fully supported by the development partners. In the case of the analysis on health issues in particular, Dr. Guy Clarisse (Head of Health Team, UNICEF) and Dr. Fatoumata Binta T.Diallo (WHO Angola Representative) provided constructive advice and input in critical stages of the report. Mr. Lluis Vinyals (Consultant of the EU PASS team) provided documents and data needed by the mission. Dr. Alice Otiato (USAID) shared her knowledge and views on health service delivery. The mission appreciates the help provided by Dr. Kusunga (Director of Bengo Province Health Directorate) during the mission's field help. Dr. Maria Julia Grave (HAMSET Project Office, MOH) offered instrumental help to set up meetings and field visits. The report was prepared under the overall supervision of Emmanuel Akpa (Sector Manager, AFTP1) who offered overall conceptual guidance, provided critical analytical advice and ensured quality control and management support. Michael Baxter, Country Director (AFCCS2), supported the process and provided the major guidelines. The peer reviewers, Jeffrey Lewis (DECVP), Delfin Go (AFRCE), and Sameh El-Saharty (MNSHD) provided valuable comments and suggestions in different stages of the report. The colleagues from the Country Office in Angola, especially Alberto Chueca-Mora (Country Manager, AFMAO), Olivier Lambert (Sr. Country Officer, AFMAO), and Christopher Porter (Consultant, AFMAO), provided superb support to the task team. Maria Margarida Baessa Mendes and Domingas Pegado provided excellent logistical support for the missions. Ligia Irias-Castillo went beyond the line of duty and showed outstanding commitment to make sure that the editorial quality of the report was kept at the highest standards. -ii- INTRODUCTION PUBLIC SPENDING IN PERSPECTIVE IN ANGOLA A. WHAT IS THE OBJECTIVE OF THIS REPORT 1. The main objective of this Public Expenditure Review is to examine the allocative efficiency and equity of public spending in Angola. The World Bank routinely undertakes economic and sector work--including economic memorandums, poverty assessments, and reports on specific topics such as decentralization and the environment. A public expenditure review analyzes public sector issues in general and the public budget in particular. This report investigates the composition and structure of public spending across sectors in the Angolan context and identifies areas where trade offs and synergies are to be found. The Angola PER also discusses options to support the government to prioritize public interventions from its budget in terms of efficiency and equity (impact on poverty), and reviews the milestones achieved in the governance and transparency reform agenda. 2. The primary motivation for this PER is to call attention to the opportunities to improve the quality of public expenditures that have been missed in the past and suggest ways to do better in the future. The proposed approach involves (a) an assessment of recent progress and remaining challenges in public finance management; and (b) a critical review of the government's current operations and its use of public resources. The working hypothesis of the PER is that there can be no improvement in the efficiency of public spending without improvements in governance and transparency associated with the budgetary process. Governance is understood, in the context of this PER, as set of 3 elements which include: (i) soundness of policies and regulations; (ii) quality and efficiency of government public expenditure and management capacity; and (iii) robustness of institutional oversight.1 These are issues that are relevant to all levels of government in Angola and that will deserve increased attention in the near future as the country is expected to accumulate sizable oil windfall revenues over a relatively short period of time. 1Good governance is widely recognized as critical to successfully addressing the Paradox of Plenty. Good governance is variously defined, but it is clearly multi-dimensional and should include, among other things: clear and stable laws and regulations; rule of law; a high level of competence in government; fiscal, budgetary and monetary discipline; public-private sector balance in the economy; an open dialogue between government and civil society; and a high degree of transparency. - 1 - B. WHAT IS THE RELATION OF THE PER WITH THE PEMFAR 3. The World Bank published in 2005 the PEMFAR - a comprehensive report that analyzed the main characteristics of the budget process in Angola. The underlying work that led to the production of the PEMFAR was carried out during 2003/04 and focused on assessing: (i) the budget process; (ii) governance issues in the oil sector and the relationship between Sonangol and the Ministry of Finance; and (iii) public procurement practices in Angola. The report, which was discussed with the authorities in 2005, identified weaknesses in the public financial management system in Angola, including the existence of a complex relationship involving Sonangol, the Central Bank, and the Ministry of Finance, and offered policy recommendations on how to strengthen the budget process and on how to develop accountability mechanisms. 4. The PER builds on the analysis and the policy dialogue started by the PEMFAR and focuses on the factors that constrain the efficiency of public spending in Angola. An important feature of the PER is that it presents an assessment of the current macroeconomic context and an update of the reform of fiscal management, based on the findings of the PEMFAR. For example, the report highlights that the link between planning and budgeting is very weak in Angola and that, as a consequence, public expenditure policies ­ including fuel price and utility tariff subsidies ­ are established in a largely ad hoc manner, and without the underpinning of a medium- term spending plan or strategy. Therefore, the foundations for sound public expenditure policymaking become very shaky as well. However, the report makes the point that it is not advisable to move to the outright implementation of a Medium- Term Expenditure Framework (MTEF) at this point in Angola as MTEFs cannot work if a solid budgetary management process ­ particularly as regards budget execution ­ is not in place. 5. On the governance and budgetary process front, the report calls attention to the fact that the deficiencies in public expenditure management and the problems associated with the omnipresent role played by Sonangol in the economy are of greater concern now that the country is expected to accumulate sizeable windfall revenues over a relatively short period of time. The PER argues that this situation weakens transparency and accountability, and impairs planning, ultimately affecting the efficiency of public spending. C. HOW THE REPORT IS ORGANIZED 6. This Public Expenditure Review is organized in two volumes and covers the sectors of Education, Health, Agriculture, and the area of decentralization. The first volume is a policy briefing and the second volume brings the complete analysis by sector. These sectors were chosen because of their importance and their potential to contribute to poverty reduction through enhanced service delivery (Education and Health) and generation of employment and incomes (Agriculture). The scope of the - 2 - inquiry is limited due to data limitations. Infrastructure, although important, is omitted because information is scant and difficult to reconcile at this point in time in Angola. It is hoped that as the budget process is strengthened and statistical capacity building takes place, future PERs may have a broader scope and coverage. 7. The PER is a joint product of the World Bank and several development partners of the Government of Angola. The Government of Angola supported the process underlying this Public Expenditure Review since its inception in June 2006. The report is the result of collaboration between the World Bank and several development partners that include the European Commission (EC), the Food and Agriculture Organization of the United Nations (FAO), the World Health Organization of the United Nations (WHO), the United Nations Development Program (UNDP), and the United Nations Children's Fund (UNICEF). 8. In broad terms, the PER stresses the fact that the post-conflict status of the country has associated with it huge needs in terms of physical and human capital investments. The report acknowledges that deficiencies in public financial management are being addressed at a gradual pace, and that mineral wealth is growing fast with favorable scenarios in the diamond and oil sectors. However, the report also emphasizes that this poses a challenge to the authorities at this point as accumulation and management of abundant windfall revenues becomes particularly difficult when country systems are not properly developed to absorb growing financial flows. D. THE REPORT'S FIVE CROSS-CUTTING MESSAGES 9. Five specific issues came up during the analysis of public spending for the three sectors covered in this report (Education, Health, and Agriculture) and the broad area of fiscal management. They are treated here in a consolidated fashion with the objective of providing an overview of the analysis that follows in the next chapters: The need to prepare public spending plans within a multi-year perspective; The need to improve the efficiency of public spending; The need to improve the quality of the data on social indicators and budget out-turns; The need to observe equity considerations in the preparation of the budget; The need for steady progress with decentralization. Budgeting within a Multi-Year Perspective 10. Angola has experienced a huge improvement in fiscal performance on the account of oil. Economic growth in 2006 reached 15 percent and preliminary data indicate a fiscal surplus of 9.4 percent of GDP in the government accounts. International reserves have grown rapidly and the strength of oil revenues and international reserves has reduced the risk of debt distress, which is assessed now at a moderate level. The government has taken advantage of the positive macroeconomic - 3 - outlook and has contracted a substantive amount of non-concessional loans to finance its investment needs. It is estimated that Angola now holds some US$15 billion in non-concessional terms with non-OECD development partners. 11. The amount of revenues that is expected to be accumulated by the government in the short to the medium term is substantial. According to estimates prepared by the IMF, international reserves are forecast to reach US$30 billion by 2010, the equivalent to 16 months of non-oil related imports. In order to manage and spend growing oil revenues in the most efficient way, a multi-year approach to budgeting is needed. It will also be necessary to strengthen the link between public expenditure policies and the budget process, especially at the sectoral level. At this point, public expenditure policies are established in a largely ad hoc manner and without the underpinning of a medium-term spending plan or strategy. In very broad terms, Angola needs a stronger link between policies and budgeting, but most importantly it will need also to consider how to frame its public expenditure policymaking within a medium to long term framework akin to the theory of permanent income. 12. At the sectoral level, investment planning is done at a very centralized level and is not preceded by a needs and financial efficiency analysis. During consultations with the authorities, there were repeated references to the fact that there exists insufficient coordination between the central and provincial governments on investment needs and planning. Perhaps because of this lack of coordination in the planning stages of the budget process, it becomes difficult to prepare an investment plan within a multi-year perspective. As a result, the annual state budgets include ambitious forecasts for capital expenditures that are actually reflecting projects that are supposed to be executed over a number of years. Absorptive capacity is usually low and so are the execution rates of the planned projects. Improving the Efficiency of Public Spending 13. The introduction of a multi-year approach to budgeting does not necessarily guarantee that money will be well spent. It is also necessary to get the policies right and build well-functioning institutions to avoid waste and develop accountability mechanisms. In order to get the most out of each kwanza spent by the public sector, the authorities need to pay attention to three areas where performance can be enhanced: macroeconomic discipline, strategic priority setting, and efficient public service delivery. Reforms in these areas involve principles of political economy, public economics, budgetary and financial management, and other institutional and regulatory reforms. 14. The PER calls attention to the need to maintain the public spending envelope within a level that is consistent with the country's absorptive capacity. This involves planning future spending within a multi-year framework and making investment decisions taking into account future recurrent costs. This is related to the need to identify strategic priorities which in turns requires a technocratic approach that gives policymakers the information they need ­ such as tradeoffs between competing claims on the budget ­ to make informed decisions. The report also points out that - 4 - while a balanced aggregate sectoral allocation of resources is necessary, it is equally important that there are no egregious biases in the composition of intrasectoral expenditures. Having the right budget allocation is essential for meaningful public expenditure reform, but focusing on allocative policies alone will do little unless such policies are efficiently executed and effectively managed. 15. Another important factor affecting the efficiency of public spending is related to the need to have well-functioning institutions. Public institutions such as tax collection agencies, law enforcement agencies, regulatory bodies, accounting and auditing units that lack proper incentives (e.g., low pay, unfair recruitment and promotion) often become a breeding ground for corruption and low morale, eroding the effectiveness of public institutions. In Angola, the legacy of a centralized administration with a Marxist orientation poses a real challenge to the functioning of institutions. But the recent boom in the oil sector places the country in a comfortable position now to reform the incentives in the public sector and create the conditions to increase the efficiency of public spending. Enhancing the Quality of the Data on Social Indicators and the Budget 16. One of the recurrent problems encountered during the preparation of this report was concerned with the quality and completeness of the data on social indicators and budget out-turns. In many occasions the necessary information was either not available, or when available was of limited coverage, quality and usefulness, in particular in what concerns the sectors of Education and Agriculture. There are problems with the classification of public expenditures in most of the sectors and the existing social indicators date back from the period of the war when it was difficult to collect information from the rural areas and conflict-affected zones of the country. 17. The lack of good data constrains the authorities' ability to assess the effectiveness of public expenditures and limits the capacity of policy makers to use technical criteria to define budget allocations. Furthermore, inadequate and out of date data preclude a rigorous analysis of links between public spending and outputs and outcomes. Overall, the situation in Angola is still very fragile in the sense that the public accounts system remains under-developed and information on actual spending (as opposed to the budgeted amount) is available only after a considerable lag. 18. The Government now has a long-term plan to address this problem and it is hoped that data quality and coverage will improve substantially in the near future. In the very short term, the authorities need to improve further the quality of public financial management so that the government can gain a better understanding of its public spending and increase the effectiveness of financial planning and programming. Key to that end is the steady commitment to finalize the implementation of the SIGFE and the need to improve the classification of public spending. Once these steps are concluded, it will be easier to design and implement with more clarity a pro-poor public spending program. - 5 - Equity in the Budget 19. The budget is prepared without due attention to equity considerations. The allocation of revenues to the different sectors is made on ad hoc basis and does not follow any technical criteria. This is actually recognized as a problem both at the provincial and at the national levels. In all the sectors covered in this report, there were strong indications that the policy process does not always involve consultations with line ministries and/or provincial governments on their priorities for the purpose of defining budget allocations. 20. There are no technical criteria for the allocation of resources from the central government to the provinces. A recent UNDP analysis estimates that coastal provinces, including Luanda and those in the West, South, and some in the East, received a flat allocation equivalent to US$200 per capita in the 2007 budget (US$175 in 2006), while those in the North and Center South were allocated the equivalent to US$135 per capita in the 2007 budget (US$100 in 2006). It is important to notice that the latter region (Center South) includes the poorest provinces in Angola, according to the 2001 MICS. For a better geographical distribution, population and poverty indicators should be factored in the decision process. The authorities could also consider using criteria based on "positive discrimination" to support the provinces that suffered the most during the war to catch up with the others. 21. Inequality of opportunities and access to public services, particularly in health and education, are extremely important in Angola. The country is marked with geographical disparities that are visible between and within provinces as well as in municipalities and communes, and rural and urban locations. Despite this, there are no criteria related to the size of the population that guide the distribution of resources by the Ministry of Finance. For example, when the price of oil increased in 2005, each province received an additional amount of US$20 million for investment, regardless of their needs or relative size. This lack of objective criteria favors provinces with a smaller population and could perpetuate existing distortions. Further Progress with Decentralization 22. In the area of decentralization, the PER recommends a gradual approach. The analysis of the Angolan context suggests that it would be important to consolidate administrative deconcentration and delegation of responsibilities first and only then move on towards fiscal decentralization per se. The authorities have actually taken the initiative to start a program of deconcentration and administrative decentralization. A new law (No. 02/07) has recently been passed and refers to a reduction of the degree of dependence of the provinces from the central government and a larger division of responsibilities. So far, the strategy of reform has been in terms of deconcentration (delegation of some responsibilities, while retaining fiscal control with the central government). This gradualist approach seems appropriate especially in the light of the dearth of capacity in public administration that is especially acute outside Luanda. - 6 - 23. From a service delivery point of view, the administrative system remains highly centralized. Some progress has been made with deconcentration to the provincial and commune levels in education and health, but not so much in the case of agriculture. The new legislation also carries some confusion as regards concurrent competencies. Concurrence of competencies poses a serious risk of wastage of efforts and resources, difficulty to determine accountability for the provision of services, and institutional conflicts provoked by the absence of boundaries between different powers. 24. The government should continue to gradually decentralize administrative responsibilities and devise a plan to start introducing some fiscal decentralization in the system accompanied by an increase in the amount of training in public administration at the local level. Full decentralization may take years to materialize, but it is important to think strategically and follow principles of good governance while the existing legal framework is reformed and new laws and decrees are created. 25. The following chapters of this report deal with these cross-cutting issues in a more detailed and disaggregated fashion. The next chapter discusses the macroeconomic context and presents an update on the reform of fiscal management. The following chapter deals with the case of Education and identifies key institutional and organizational constraints that hamper efficiency and quality in service delivery. It also presents an analysis of the trends concerning the budget allocation to education, its composition and impact on education sector goals. The following two chapters deal with the sectors of health, and agriculture. A final chapter discusses alternative scenarios to deepen administrative and fiscal decentralization in Angola. - 7 - CHAPTER I MACROECONOMIC CONTEXT AND UPDATE ON THE REFORM OF FISCAL MANAGEMENT A. MACROECONOMIC CONTEXT: BENEFITING FROM A WINDFALL 26. Recent growth performance has been extraordinary, fuelled by growing oil production and recovery from the war period. Economic growth in 2006 reached 15 percent. In addition to minerals, output of public services, agriculture, construction, and public utilities has grown strongly, increasing by an average of about 10 percent during 2004­06. However, growth in the non-mineral economy is expected to slow to a more sustainable 5­7 percent over the medium term once the economy moves from the post-war recovery phase to a broader development stage. The economy is expected to expand by 30 percent in 2007, largely reflecting continued good performances in the oil and diamond sectors and recovery in the non-mineral sectors. Oil-related GDP growth is expected to reach 41 percent in 2007 while the non-oil economy is projected to grow by 26 percent over 2006. GNI per capita under this assumption would reach US$2,994 at end-2007 placing Angola on the average level of per capita income registered for middle income countries ($2,640 in 2005, WDI). 27. There has been improved fiscal performance on the account of oil. The fiscal situation remained highly favorable in 2006 and is expected to remain comfortable over the next 4-5 years mostly due to high international oil prices and rising domestic oil production. Fiscal policy has been consistent with maintaining macroeconomic stability and containing inflation, which declined to 12 percent in 2006. Preliminary data indicate a fiscal surplus of 9.4 percent of GDP in 2006. The non-oil fiscal deficit deteriorated by 8 percent reaching 69 percent of non-oil GDP on account of larger fiscal spending fuelled by rising capital outlays. In addition to introducing a conservative price of oil and revising downwards the projected oil production figures for 2006 (to minimize pro-cyclical spending), the government has also increased substantially the level of its international reserves to cover financing needs in case of an oil shock and has repaid all principal arrears to the Paris Club ($2.3 billion) in December 2006. 28. The strength of oil revenues and international reserves has reduced the risk of debt distress. Based on the joint low-income country debt sustainability framework of the World Bank and the IMF from June 2006, Angola is assessed to be at moderate risk of debt distress. Its debt ratios have improved substantially in the last two years, and will continue to do so under the baseline scenario used in the debt sustainability - 8 - analysis, owing to higher oil revenues. Foreign exchange reserves are also expected to continue to rise, reducing the likelihood of Angola facing repayment difficulties on its external debt. Apart from the ratio of NPV of debt-to-GDP in 2006, all three debt- burden indicators are expected to remain below their policy-dependent thresholds. With its heavy reliance on oil, however, Angola remains vulnerable to reductions in oil prices. Going forward, Angola will be at much lower risk of debt distress if it abstains from further non-concessional external borrowing. Table 1. 1: Macroeconomic Framework, 2003 - 2007 2003 2004 2005 2006 2007 Est. Proj. Proj. (Percentage change, except where indicated) National income and prices Real GDP 3.3 11.2 20.6 14.6 30.2 Oil sector -2.2 13.1 26.0 15.0 40.9 Non-oil sector 10.3 9.0 14.1 13.8 13.7 GDP per capita (in U.S. dollars) 959 1,322 2,129 2,780 3,614 GNDI per capita (in U.S. dollars) 848 1,157 1,866 2,449 3,082 Consumer price index (annual average) 98 44 23 13 8 Consumer price index (end of period) 77 31 19 10 7 Money and credit (end of period) Net domestic assets 2/ 12 -97 -9 -102 -52 Broad money 2/ 67 50 60 43 29 M2 velocity (non-oil GDP/average M2) 3.35 3.55 3.33 2.92 2.65 Base Money in real terms (percent change) -0.5 19.1 40.2 30.0 16.0 (Percentage of GDP, except where indicated) Fiscal accounts Total revenue 37.9 36.9 38.0 38.0 37.9 Of which : oil 27.9 28.4 30.1 30.0 30.5 grants 0.8 0.5 0.2 0.3 0.2 Total expenditures 44.3 38.5 31.2 35.7 32.5 Overall balance (accrual basis) -6.4 -1.6 6.8 2.2 5.4 Non-oil fiscal balance (accrual basis) -35.1 -30.4 -23.6 -28.0 -25.3 Overall balance (cash basis) -5.6 -3.7 6.0 1.6 3.1 External sector Current account balance (including transfers; deficit -) -5.1 3.5 12.9 8.8 12.4 External debt (in billions of U.S. dollars) 10.2 10.8 12.6 15.0 16.3 External debt-to-GDP ratio 73.1 54.5 38.5 34.1 27.7 Debt service-to-net-export ratio 3/ 16.5 16.4 10.5 4.8 6.4 (In millions of U.S. dollars, except where indicated) Net international reserves (end of period) 4/ 790 2,023 4,140 9,252 13,920 Gross international reserves (end of period) 4/ 800 2,034 4,147 9,261 13,927 Memorandum items: Gross domestic product (in millions of US$) 13,956 19,800 32,810 44,103 59,019 Official exchange rate (kwanzas per U.S. dollar; end-of-period) 79.1 85.6 80.8 ... ... Gross domestic product (in billions of kwanzas) 1,041 1,652 2,860 3,539 4,839 Oil production (thousands of barrels per day) 875 989 1,247 1,434 2,019 Price of Angola's oil (U.S. dollars per barrel) 28.2 36.4 50.1 56.6 57.4 Non-oil fiscal balance/GNDI -38.8 -34.3 -26.6 -31.5 -29.4 Sources: Angolan authorities and IMF and World Bank staff estimates and projections. 1/End of period. A positive sign denotes appreciation. 2/As percentage of beginning-of-period M3. 3/In % of exports net of oil-related expenses such as oil-related imp. of goods and services and oil companies' remittances. 4/Includes government deposits in overseas accounts. - 9 - 29. The macroeconomic outlook remains favorable, but sustainability will require better medium-term expenditure planning. The external current account recorded another surplus at 11.4 percent of GDP in 2006 and gross international reserves reached about 5.7 months of non-oil imports at end-2006. Fiscal revenues should continue to grow robustly with growing oil production until 2010-11 (see Figures 1.1 and 1.2 below). They are currently the equivalent to 46.7 percent of GDP and are forecast to decline to 32.5 percent of GDP by 2011. Fiscal spending, on the other hand, is currently the equivalent to 37.5 percent of GDP (2006) and is expected to reach 33.2 percent of GDP by 2011. Under this scenario, that does not assume new oil discoveries, fiscal surpluses would disappear in the medium-term, as oil revenues decline as a percentage of GDP. The first priority for the government in this context is to prepare a clear strategy to manage the oil revenues in the medium to long-term (see Box 1.1). Figure 1. 1: Oil Production in Angola, 2000-2020 - 3,000 2,500 2,000 1,500 1,000 500 - 200020012002200320042005200620072008200920102011201220132014201520162017201820192020 Current production Pending production - 10 - Figure 1. 2: Oil Revenues under Different Price Scenarios ( 45000 40000 35000 High price 30000 n ioll 25000 mi Base price $ 20000 S U 15000 Low price 10000 5000 0 1990 1995 2000 2005 2010 2015 2020 2025 2030 Box 1. 1: Why it is Necessary to Move toward a Multi-Year Approach to Budgeting Over the next few years, Angola's revenues from oil will be subject to three additional sources of variation, in addition to crude price volatility: (i) the changing composition of production; (ii) the increasing volume of extraction; and (iii) the increased importance of PSAs and its implications for the behavior of the State's profit oil. Thus, more than desirable, it is very necessary that Angola adopts an approach to budgeting that goes beyond the one-year framework of the OGE, to avoid unnecessary disruptions in cash management and to correctly forecast the Government's fiscal stance. In particular, the Government should take advantage of tailor-made instruments, such as the recently adopted Oil Revenue Forecasting Model, to improve the quality and reliability of oil revenue forecasts, over a multi-year period. The need for a multi-year approach to budgeting is even greater when it comes to establish sound public expenditure policies. The link between the latter ­ especially at the sectoral level ­ and the budget process is weak in many developing countries, and Angola is no exception. Public expenditure policies ­ including fuel price and utility tariff subsidies ­ are established in a largely ad hoc manner, and without the underpinning of a medium-term spending plan or strategy. Therefore, not only the links between policies and budgeting are weak, but the foundation for sound public expenditure policymaking is very shaky as well. The adoption of a medium-term approach to budgeting, therefore, has to be accompanied by a medium-term approach to public expenditure policy as well. The adoption of such medium-term approaches should not imply the outright implementation of a Medium-Term Expenditure Framework (MTEF). As pointed out by recent literature (see Le Houerou and Taliercio (2002)), MTEFs cannot work if a solid budgetary management process ­ particularly as regards budget execution ­ is not in place. Insofar as Angola's public financial management remains characterized by a dichotomy between formal and "non-conventional" mechanisms of budget execution, adopting an MTEF at this point in time will be difficult, if not impossible. But this does not prevent the Government from moving towards the conditions required for a successful MTEF. - 11 - B. GROWING REVENUES WILL REQUIRE BETTER MEDIUM- TERM PLANNING 30. The recent improvement in the fiscal stance requires a better medium-term public spending framework. If constraints on absorptive capacity and deficiencies in public expenditure management can be overcome, the authorities should be in a comfortable position to increase public spending that could lead to major improvements in the social and economic infrastructure and alleviate constraints on the supply side of the economy. According to the latest estimates (from the 2006 IMF Article IV mission), with rising oil revenues, international reserves are projected to rise to about US$30 billion, the equivalent to 16 months of non-oil-related imports, by 2010. But to use these resources in an effective way, the authorities must improve their capacity to plan public spending using a medium-term framework. 31. The migration towards a Medium-Term Expenditure Framework (MTEF) requires careful preparation and steady political commitment. The successful adoption of a medium-term approach to budgeting and policy needs to rest on three main pillars2: · Expenditure control mechanisms, particularly with respect to budget execution, need to be strengthened; · The budget process needs to be guided by public expenditure policy, not the reverse; and · Revenue forecasting capacity needs to be enhanced. 32. An MTEF is not a panacea and the process to get there should follow a building-block approach. The first pillar above is particularly concerned with the strengthening of the formal budgetary execution system, as discussed in the PEMFAR report of 2005. The second one ­ public expenditure policy driving the budget process ­ requires a sound articulation between short- and medium-term policy and planning instruments, as well as improved coordination between MINFIN and MINPLAN. And much progress can be made on the third one if the current capacity development activities currently ongoing in the National Tax Inspectorate to implement the oil revenue forecasting model are adequately implemented (see paragraph 58). 33. A possible model of articulation of these three pillars can be mirrored on the example of Mozambique. The model, illustrated in Chart 1, which is reproduced from the PEMFAR, advocates the establishment of clear links between four basic medium- term programming instruments ­ the PIP, sectoral/provincial plans, the PRSP, and an MTEF ­ with the two basic yearly operational plans ­ the Bi-Annual Economic 2Cf. Schiavo-Campo and Tommasi (1999), p. 88. - 12 - Programs and the State Budget (OGE). The main difference is that in Mozambique all six instruments are under the responsibility of a single entity, the Ministry of Finance and Planning. In Angola, MINPLAN is in charge of the PIP, the sectoral/provincial plans, and the Bi-Annual Economic Programs, while MINFIN is responsible for the OGE. The PRSP process has also been assigned to MINPLAN, while an eventual MTEF would likely be part of MINFIN's work program. 34. Some good principles that could guide the adoption of an MTEF include: (i) start with the introduction of a medium-term fiscal framework; (ii) do not launch MTEF in selected sectors until there are medium-term ceilings provided by the Ministry of Finance that provide the context for the sectoral MTEFs; (iii) the integration at a measure pace of capital and recurrent budgets; (iv) incorporation in the budget process of performance information (outcomes, outputs) in the first year ­ these might be introduced in the second or third year of an MTEF; (iv) gradual delegation of more flexibility for spending ministries to allocate their resources across programs, subprograms, and activities both within the year and for the budget as the overall control environment improves; (v) once budget classification and execution monitoring are fully operational in the SIGFE pressure on ministries to find resources within current expenditures may be more successful based on the existing information on program and activity costs. 35. Various improvements can be made to the budget process that move in the direction of the ideal MTEF incrementally. The principles of multi-year forecasts, policy and resource allocation alignment, costing activities, spending ceilings for line ministries, and prioritization of competing claims for new resources can be implemented in various ways. An MTEF needs to be customized to the country, including initial conditions in Public Expenditure Management (PEM), human and information technology capacity. Table A.1 in the annex to the this Chapter provides a multi-year perspective that correlates with the three-level framework of public financial management objectives and identifies the responsible agency, necessary pre- conditions and assesses Angola's readiness. 36. Improving the coordination between the investment and current expenditure programming remains as a key challenge in Angola. While merging MINFIN and MINPLAN might facilitate such articulation, this would be neither a necessary nor a sufficient condition for a successful integration of the medium- and short-term processes. As noted by Schiavo-Campo and Tommasi (op. cit., p. 96), "the real issue (...) is the lack of integration between investment and current expenditure programming, and not the formally separated processes in themselves. This is important, because to misspecify the issue would lead (and often has) to `solving' the problem by a simple merger of two ministries ­ even while coordination remains just as weak". This is an issue which still deserves the focused attention of the authorities in Angola, as the articulation between the PIP and the OGE has yet to improve. - 13 - Budget Plans bi- of objectives Social or (OGE) Policy Actions and (annual Consistency with and Operational Budget Instruments State of Economic Programs annual) Annual Articulation Tools Proposed the goals Limits Objectives of sector Budget stem: of process of and Sector - Sy and Goals Short-Term Expenditure Investment 14- Integration and Coordination monitoring Integration Objectives Macro Scenarios Budgeting and and Medium- Sector and Planning, of (internal policy allocation resources) (PIP) and limits Policy, framework (ECP) 1: Expenditure external Harmonization Strategy Program Programming Strategy Resource Inter-sectoral Investment and Chart Provincial Term Plans (MTEF) and Reduction sector PCRRP) Investment policy Resource of Sector Strategic and Poverty Medium Framework Medium-Term Public (including Priority Allocation Integration strategy 37. The importance of the PRSP process in strengthening the links between policy and budgeting should not be underestimated. A sufficiently open process in the refining and future updating of the current draft of the Angolan I-PRSP could increase the odds that the line ministries would take more ownership of the policies envisaged in the poverty reduction strategy, and could contribute to reduce the degree of MINFIN discretion in devising public expenditure allocation across sectors and provinces. Therefore, the PRSP could be converted into the ultimate instrument for articulating policies and financing in Angola. In addition, as the public financial management process is strengthened, and the "seeds" for a medium-term approach to budgeting continue to be cultivated, the government should consider a careful move towards a Medium-Term Expenditure Framework as one of the pillars in an eventually revised PRSP. C. THE ROLE OF THE GOVERNMENT'S POVERTY REDUCTION STRATEGY 38. The Estratégia de Combate à Pobreza (Strategy to Combat Poverty, ECP - Angola's PRSP) was approved by the Council of Ministers in December 2003 and circulated to the donor community in January 2004. Although the ECP has been approved by the Council of Ministers, to this date it has not yet been formally submitted to the Boards of the Bank and of the Fund for a number of different reasons. These reasons range from disappointment with the lack of prompt assistance from donors at the end of the war in 2002, to a rapidly changing economy leading to more than doubling of the country's income per capita (from $848 in 2003 to an estimated $1,980 in 2006), thus moving Angola rapidly towards IDA graduation. 39. The government also has an official rehabilitation and reconstruction program that complements the ECP. The first phase, called the priority phase (Priority Phase of the Multisector Rehabilitation and Reconstruction Program, PPMRRP), is being implemented during 2003­08, and covers about 70 percent of the costs identified in the ECP. The government plans to define and prepare a second phase at the end of implementation of the priority phase. The overall purpose of the PPMRRP is to consolidate peace and reestablish state administra- tion throughout the country, and to initiate the medium and long-term process of reconstruction and economic rehabilitation. Specific objectives are (a) to address the immediate needs of the poorest people and displaced population; (b) to rapidly improve the quality of life of people living in both rural and urban areas by expanding access to vital public services; (c) to reestablish critical transportation links throughout the country and, in particular, to reconnect the provinces most affected by the war--Bié, Huambo, Moxico, Lunda Norte and Lunda Sul, Uige, Cuando Cubango and Malanje; (d) to enhance food security and start to revive the rural economy, especially in the provinces most affected by war; and (e) to strengthen capacity of government at all levels to formulate, prepare, implement, and manage medium and long-term development programs. The government intends to fund some 60% of the PPMRRP from its own resources and expects the donors to assist in supporting the rest. 40. The ECP proposes to implement a public expenditure program focusing on ten priority areas, namely: (i) social reinsertion; (ii) demining; (iii) food security - 15 - and rural development; (iv) HIV/AIDS; (v) education; (vi) health; (vii) basic infrastructure; (viii) employment and vocational training; (ix) governance; and (x) macroeconomic management. The estimated costs of the ECP for the 2003-2006 period total US$ 2.6 billion (at end 2005 values). The ECP costs were originally presented as part of a multi-year fiscal framework, which was supposed to be the basis of an effort at establishing an MTEF (which has not yet materialized). 41. The authorities are aware that the original strategy, which covered the period 2003-2006, now shows a stark disconnect with the rapidly changing economic realities in Angola. First of all, the macroeconomic framework that is used as background for government interventions in the different priority areas elected in the ECP is no longer forward-looking and is now largely outdated. For example, the estimate for total revenues projected for 2005 and 2006 in the current draft of the ECP was of $8.6 billion in 2005 and $9.9 billion in 2006. Taking into account the rapidly growing oil production that has taken place over the last two year and rising international oil prices current official estimates for total revenues are at least 50% higher than those used in the ECP at $12.5 billion in 2005 and $16.1 billion in 2006. The same is true for total expenditures and, in particular, capital expenditures, which were projected at $2.4 billion in 2005 and $1.7 billion in 2006 in the current draft of the ECP, as opposed to current official estimates in the revised 2006 state budget of $1.7 billion in 2005 and $7.6 billion in 2006. 42. A second challenge is to present a detailed costing analysis of the proposed interventions per priority area, per sector, per activity, and per year. The annual state budgets are not in coordination with the expenditures envisaged in the ECP. In 2006 alone, for example, the revised budget submitted to parliament in July 2006 projects total capital expenditures at more than $7 billion while the total estimated cost of all the public spending associated with the ECP is estimated at $2.6 billion for the whole period 2003-2006. Even if one considers that budget execution rates have historically been below 60% in Angola due to limited absorptive capacity, and that budgets are not prepared using a medium-term fiscal framework due to insufficient technical capacity to plan beyond a 2-year time horizon, there is a clear disconnect between the original objectives of the ECP and the fiscal envelope currently available to the government. 43. Third, the ECP is not anchored on a monitoring and evaluation strategy that outlines annual targets and expected outcomes per priority area and does not include a list of indicators (with their respective sources) to be used to assess progress in its implementation. For example, the matrix in the current draft of the ECP puts forward ambitious targets in the area of health such as a 75% reduction in the rates of infant mortality and maternal mortality by 2015 (which is well beyond the period of coverage of the ECP). As the National Institute of Statistics (INE) does not produce statistics on infant mortality and maternal mortality rates on a regular basis (the latest available social statistics are from 2001 and are based on a curtailed sample that did not cover the whole of the country due to the ongoing war at the time), it is not clear how progress towards achieving these targets will actually be monitored during the period of implementation of the ECP. 44. Finally, there is still a dearth of knowledge on the livelihoods of the poor in Angola which hampers the design of an appropriate and targeted pro-poor public spending program. Overall, it is difficult to have a full assessment of the - 16 - existing social programs given that these do no have at the outset an expected number of beneficiaries. This failure comes from the lack of reliable data on poverty to which programs could refer to in order to assess appropriateness and effectiveness in their targeting and the impact of their activities. A sensible way of tackling this limitation would be to strengthen the National Statistical Institute (INE) and establish a program of annual monitoring of poverty indicators and disseminate these to the public at large. 45. The authorities are working to improve the meaningfulness of the ECP. As described in the text of the ECP, implementation was supposed to be initially targeted to immediate post-conflict rehabilitation needs and to facilitating the return of internally-displaced persons and refugees to their communities, which has been happening with some degree of success with the support of the World Bank. Upon completion of that priority phase (expected for 2007), attention should turn to the stabilization and recovery phase (2007-20), with a focus on reconstruction of the destroyed infrastructure and support to economic growth. While the phasing of the strategy is right, the current draft of the ECP appears largely underbudgeted and unrealistic. Government is finalizing an updated version of the ECP that covers 2006­08 and includes a forward-looking macroeconomic framework. D. THE BI-ANNUAL ECONOMIC PROGRAMS 46. In the absence of a medium- to long-term expenditure framework, the government has been designing on a rolling basis bi-annual economic programs to support the preparation of the annual state budgets. Originally, the ECP was supposed to be the basis for a long-term public expenditure policy that should be implemented through two yearly interventions ­ the bi-annual economic programs and the annual state budgets. However, as the ECP has not been revised substantially since its first draft of 2002 while the bi-annual economic programs and the annual state budgets have been driven by short-term prioritization and constraints, there has a been a growing disconnect with the ECP as time goes by. 47. Recent economic programs forecast massive physical investments in infrastructure and a substantial scaling up of social spending. In the 2005-2006 economic program that was submitted to the National Assembly in November 2004 the government prioritized five main broad objectives: (i) re-launching development investments (including rural development and other productive sector activities, and upgrading of war affected physical infrastructure); (ii) improving service delivery; (iii) enhancing social assistance and safety net programs; (iv) institutional reforms; and (v) upgrading housing and urban infrastructure. The cost of the 2005-06 program has been estimated at Kz 364 billion ($4.3 billion), which is more than double on an annual basis what is programmed in the draft PRSP ($2.6 billion over the 2003-2006 period). 48. A significant part of the program is being financed by non-traditional OECD donors such as China and Brazil. The larger part of the costs will be directed to re-launching the economy (63.4%) and improving service delivery (19.8%). Government hopes to finance this program largely through bilateral and multilateral credits. China will provide about 80% of the funds with Brazil offering another 10%. Table 1.2 below provides a snapshot of the financing - 17 - available from some bilateral partners. At the multilateral level, the government is also expecting financing over the two-year period from IDA (1.7%), the European Union (1.5%), and the African Development Bank (0.2%), with the remaining 7.6% being financed with resources from the National Treasury. 49. The funding from the People's Republic of China (PRC) involves a credit facility estimated at some $12 billion. The Chinese funds are supposed to support infrastructure and social spending by means of a rolling line of credit of $2 billion per year over the next 6-7 years (average interest rate of LIBOR plus 1.5%). The first of these credit lines was approved by the Angolan Cabinet on March 17, 2004, and since then there has been official confirmation of a jumbo credit line in the amount of $9.7 billion signed early in 2006 The terms of these loans are similar and the purpose of the facility is to fund a series of projects, mainly infrastructure, each executable over a maximum period of 18 months. Funding of a minimum of $9 million each would be provided for individual projects, and would typically cover a 12-year repayment period, with project-specific grace periods. Each transaction would require an up-front GOA contribution of 10%. The credit facility will also support commercial trade financing between Angola and the PRC in minimum transactions of $l0 million each. The line of credit is guaranteed by shipments of oil to the PRC. 50. An extension of the 2005-2006 program has been prepared covering the years 2007-2008. The government has decided to treat the 2007-08 program as a continuation of the 2005-06 program to guarantee the maintenance and consolidation of the good economic and social performance that has been observed in Angola since 2005. The new program envisages as opportunities for growth in 2007-08 a number of areas that involve: (a) the consolidation of peace and national reconciliation; (b) the current positive macroeconomic performance; (c) the attraction of private foreign investment; (d) the rebuilding of infrastructure; (e) the relative populational re-equilibrium in the country; (f) the significant improvement in the government accounts; (g) the availability of lines of credit; (h) the creation of the national development bank of Angola; and (i) the opening of the stock exchange. The program also identifies threats to enhanced growth that are attributed to less than expected progress in reducing poverty and the lack of an integrated transport system. 51. The 2007-08 program relies on public investment for the reactivation of the economy and has been costed at US$9.5 billion. The projects that support the economic program for 2007-08 are included in the Public Investment Program (PIP). The government claims that some 76.6 percent of all of the public investment included in the PIP will be executed in 2007-08, and that the remaining will be left for the following years. The amount forecast to be spent in 2007 alone is equivalent to some US$6.5 billion. Infrastructure and productive sectors will absorb 63.2 percent of the total budget for 2007-08. Social spending will consume 13.3 percent of the resources and the remaining sectors, including institutional capacity building and reforms, will account for 11 percent of the funds. The remaining 12.5 percent will be allocated to the provincial programs to expand and improve the supply of basic social services. - 18 - Table 1. 2: Bilateral Sources of Financing for the Government's Economic Program Lender and Date Credit Limit Purpose Terms and Conditions Interest Paid Services and Commitment Guarantee rate Commission LR Luminar US$750 Operations in the Libor +1.5% Management commission Contract of Petroleum Finance Limited, million­ public sector (US$10,000 yearly up to the final supply Israel revolving reimbursement) expenses of (No date preparation and negotiation provided) (US$50,000) Immobilization commission (0.5% per year, paid quarterly) Credit Insurance Eur 150 million Public project Euribor+1.25% Installation commission (1% flat) Ministry of Finance Hermes Germany investment Management commission (1% (No date flat) provided) Deutsch Bank, US$500 million Public project Libor + (1% to 5%) Installation commission (1% flat) Ministry of finance Spain investment Management commission 1% November 2003 flat) Variable Insurance Premium Export Import US$40 million 5 contracts for the 1.75% Ministry of Finance Bank of India supply of equipment August 2004 for Moçamedes Railway China EximBank US$2 billion Public Project Libor 3 months + Management commission (0.3%) Contract of petroleum March 2004 Investment 1.5% Installation commission (1%) supply Immobilization Commission (0.3%) Portugal­Cosec Eur 300 million Public project Euribor + (0.4% to Management commission (0.1%) Ministry of Finance Nov. 2004 investment 0.6%) China Fund US$9.8 billion Projects managed by Libor 3 months + Management commission 0.3% 2005 government 1.5% Immobilization commission 0.3% Export Import US$31.4 Rehabilitation project 0.60% Management commission (0.10% Ministry of Finance Bank, Korea million of cotton in Sumbe above each disbursement) December 2005 Satander Bank, Eur 100 million Public Project Libor 6 months + (1­ Management Commission (0.5% Ministry of Finance Spain Investment 1.5%) flat) March 2005 Variable insurance premium Commitment commission (0.25% flat) Fortis Bank, Spain Eur 250 million Public project Libor 6 months + Management Commission (2% Ministry of Finance September 2005 investment (0.75­ 1%) flat) Variable risk Commitment commission (0.5% flat) Brasil Proex US$580 million Public project Libor Management commission (0.5%) Contract of Petroleum 2006 investment Immobilization commission supply (0.5%) Brazilian US$750 million Public project Libor + 1% Management commission 1% Contract of Petroleum Development investment Immobilization commission supply Bank (0.5%), to be confirmed 2006 Export Import US$10 million Acquisition contract of Libor 6 months + Management commission (0.5% Ministry of Finance Bank of India 599 tractors "SAME" 2.50% per year) July 2006 Immobilization commission (0.5% flat) Source: Government of Angola, Ministry of Finance. 52. The budgets for 2006 and 2007 are ambitious, but absorptive capacity remains a binding constraint. Taking advantage of a sudden increase in the availability of financing, both at the domestic level with the rapid growth of oil revenues as well as from accessing substantive credit lines, the government is - 19 - signaling an enormous expansion in government spending. The budgets for 2006 and 2007 forecast increases in capital expenditures of huge magnitudes (US$10 billion in 2006 and US$8 billion in 2007, approximately). But these forecasts should be seen with care given that they include projects that are supposed to be executed over a number of years and for which the total cost has been uploaded in the annual budgets in which they were conceived. Another source of concern is the limited capacity to plan and execute all projects that are part of the government's PIP. The next sections discuss the orders of magnitude in the 2006 and 2007 budgets, and how expenditure execution rates have evolved up to 2005 (due to data availability). E. THE 2006 AND 2007 BUDGETS 53. Total revenues for 2006 were budgeted at Kz 1.411 trillion (US$16 billion) amounting to about 41 percent of GDP. Government revenues are predominantly dependent on petroleum (80 percent, up from 67 percent in 2005), with the bulk of the remainder from income and trade taxes and duties. The 2006 budget was prepared based on an estimated production of some 597.1 million barrels of oil (435.2 million in 2005) at an average per unit price of US$45 ($53.8 in 2005), which reflects the differential price applied to Angolan oil due to variations in quality. The budget also projected a growth rate of 27.9 percent (against an actual growth rate of 15 percent in 2006), compared with 15.5 percent in 2005, led mainly by a 37.2 percent forecast expansion in the oil sector. The non-oil economy is expected to grow by 11.9 percent in 2006. 54. Total expenditures in 2006 were estimated at some Kz 1.64 trillion (US$18.6 billion), representing roughly 48.3 percent of GDP. Of this amount, recurrent expenditures were expected to account for about 54.8 percent, with capital expenditures absorbing the remaining 45.2 percent. The budget earmarked 22.8 percent of expenditures for administrative and related expenditures, 14.1 percent for supplies, 14.6 percent for transfers, and 3.4 percent for interest payments. It also projected a deficit in the region of 6.6 percent, which was basically reflecting the conservative implicit price of oil used in the preparation of the budget ($45/barrel). In terms of sectoral allocations, social sector spending was forecast at almost a quarter (24.7 percent), the highest proportion in the past three years, followed by administration (13.3 percent), the lowest level over that same period. Financial charges (internal and external debt service) were budgeted at 11.4 percent. Spending on the economic sector was projected to increase to 27.2 percent, while defense and security expenses were expected to remain constant at 17 percent. 55. The budget for 2007 amounts to the equivalent to US$22.5 billion, averaging nearly US$1,500 per capita. Fiscal revenues, of which 80 percent are derived from oil, are forecast at Kz 1.7 trillion (US$21 billion) and are expected to average 37.5 percent as a share of GDP. Total expenditures are estimated at Kz 1.8 trillion (US$22.5 billion), or the equivalent to some 40.8 percent of GDP. Current expenditures are estimated to consume 64.3 percent of total expenditures while capital expenditures are projected to consume the remaining 35.7 percent. Social expenditures are expected to average 28.1 percent of the total. There is also - 20 - a substantive increase forecast for financial charges to 22.4 percent of the total reflecting mostly the regularization of payment arrears to the Paris Club. As compared to 2006, there is also a significant increase forecast for administration from 13.3 to 22.6 percent. Spending on the economic sector was projected to decline to 14.1 percent down from 27.2 percent last year, while defense and security expenses were expected to account for 12.7 percent, down from 17 percent in 2006. Table 1. 3: Macroeconomic Assumptions Used in the 2007 Budget Indicator 2005 2006 2007 Annual Inflation (%) 18.5 10.0 10.0 Oil Production (million barrels/year) 363.5 551.4 736.7 Price of Angolan Oil in the Budget (US$) 36.90 56.00 50.00 Gross Domestic Product Nominal Terms (Kz billion) 1,656.9 2,726.7 4,483.3 Real growth rates 11.7 19.5 31.2 Oil Sector 13.9 21.2 33.6 Non-oil sectors 9.1 17.2 27.9 Source: Ministry of Finance, OGE - 2007 56. The macroeconomic assumptions used to prepare the 2007 budget assume stable inflation and growing oil production. The 2007 budget assumes that inflation will stay at the 10 percent annual level and that oil production will increase by 33.6 percent over the level registered in 2006 to reach 736.7 million barrels by year end. The implicit price of oil in the budget estimated by the government is of US$50 a barrel, down from the estimated US$56 a barrel used in the 2006 budget. The budget also assumes that the economy will grow by some 31.2 percent over 2006, with the oil sector growing by 33.6 percent and the non- oil sectors expanding by 27.9 percent in real terms (see Table 1.3). As oil revenues account for some 80 percent of all government revenues, if international oil prices decline causing the price of Angolan oil to decline to levels below US$50 a barrel next year, the government would face a significant reduction in its revenues. According to a recent UNDP analysis, a decline of US$10 in the price of oil would lead to a resource gap of roughly US$5.2 billion in the value of oil production and exports implying a loss of revenues equivalent to $3 billion, or half the PIP for 2007. 57. The current mechanism for allocating budget funds for sub-national governments does not promote equity among the different provinces. There are no technical criteria for the allocation of resources from the central government to the provinces. A recent UNDP analysis estimates that coastal provinces, including Luanda and those in the West, South, and some in the East, received a flat allocation equivalent to US$200 per capita in the 2007 budget (US$175 in 2006), while those in the North and Center South were allocated the equivalent to US$135 per capita in the 2007 budget (US$100 in 2006). It is important to notice that the latter region (Center South) includes the poorest provinces in Angola, according to the 2001 MICS. For a better geographical distribution, population and poverty indicators should be factored in the decision process. The authorities could also consider using criteria based on "positive discrimination" to support the provinces that suffered the most during the war to catch up with the others. 58. The effective implementation of government plans calls for a reprioritization of public expenditures in favor of social sectors. The description so far has been based on what the government is planning to do with the growing - 21 - oil revenues. An analysis of what has actually happened in terms of the execution of the budget shows that spending in social sectors remain at low levels that do not match government ambitious to meet the MDGs. Figure 1.3 in fact shows that the notional allocation of funds to social sectors as a share of the total budget in the budget proposal for 2007 is lower than what had been proposed in 2005 and 2006 whereas there has been a significant increase in the proposed allocation for general public services and interest payments (the latter to account for the payment of interest arrears to the Paris Club). The following section shows that, as in other countries in the run up to elections, public spending in Angola remains heavily focused on general public services, consumption, higher wages and benefits. F. FUNCTIONAL CLASSIFICATION OF EXPENDITURES 59. Public spending on consumption and defense has so far grown faster than social spending. The latter have actually increased lately, but they are still low as a share of GDP (Table 1.4). Expenditures on education have increased by 13 percent and 47 percent, respectively, in 2004 and 2005, while spending on health grew by 15 percent and 17 percent over the same period, in dollar terms. Expenditures on social security, welfare and housing also increased, growing by 10 percent and 160 percent over the same period. In 2005, expenditures on education represented only 2.1 percent of GDP, health 1.4 percent of GDP, and social security welfare and housing 5.4 percent of GDP. Spending on general public services and defense, however, have by far outgrown social spending (i.e., education and health combined). They have risen, respectively, by 222 percent and 558 percent in 2005 alone. As a share of output, general public services represented some 7.8 percent of GDP while spending on defense and internal security represented 7.3 percent of GDP in 2005. Curiously, spending on energy, agriculture, mining, and transportation declined by 49 percent in dollar terms in 2005 which was enough to reduce its share in GDP to 2.2 percent down from 7.1 percent of GDP in 2004. Figure 1. 3: Functional Distribution of Expenditures, 2004-07 40 35 ng dine 30 Spla otT 25 of gea ntecr 20 Pe 15 10 5 0 Defense and internal General Public Services Social Sectors Economic Sectors Interest Payments security 2004 19.2 35.8 25.4 7 12.5 2005 26.2 16.9 29.3 11.2 16.4 2006 13.3 17.4 30.8 27.2 11.4 2007 22.6 12.7 28.1 14.1 22.4 - 22 - Table 1. 4: Angola: Functional Classification of Government Expenditures, 2003-2005 2003 2004 2005 Prel. (In percent of GDP) General public services and other economic affairs 6.7 3.9 7.8 Defense and internal security 2.9 1.8 7.3 Education 3.1 2.3 2.1 Health 2.4 1.9 1.4 Social security, welfare, and housing 4.5 3.3 5.4 Energy, agriculture, mining, and transportation 10.2 7.1 2.2 Interest payments (committed) 1.8 2.2 5.6 Unclassified 13.9 17.4 ... Total 45.5 39.9 31.9 (In millions of U.S. dollars) General public services and other economic affairs 921 797 2572 Defense and internal security 402 366 2408 Education 424 478 702 Health 329 379 443 Social security, welfare, and housing 617 678 1763 Energy, agriculture, mining, and transportation 1414 1437 734 Interest payments (committed) 255 458 1833 Unclassified 1923 3538 ... Total 6284 8131 10455 Sources: Angolan authorities; and IMF/World Bank staff estimates. 60. When taken together, sectoral spending on general public services and defense relative to total spending were responsible for nearly half of the total expenditures in 2005. Education and health when considered together, on the other hand, remained responsible for roughly 10 percent of total expenditures while there was a doubling of the share of spending on social security, welfare, and housing in 2005 alone. The sectoral allocation relative to total spending for energy, agriculture, mining, and transportation declined steadily since 2003 from 22.5 percent to just 7 percent in 2005. This pattern of public spending is common in countries with multiparty systems with presidential elections as governments in the run up to elections tend to increase spending with the objective of expanding employment in order to generate political support in pre-election periods. 61. In practical terms, the increased spending on general goods, consumption, higher wages and benefits, means that the government may have to face trade offs in terms of social spending. In the case of Angola, given the low level of social indicators and the challenges to be faced by the authorities in their reconstruction of the country, the existence of such trade offs is clear, since spending on - 23 - education and health have remained at low levels while expenditures with general public services and defense have increased much faster. Therefore, the low level of social indicators, partly as a result of decades of war and internal conflicts, call for a quick reprioritization of public spending in favor of social sectors, especially education and health. G. ECONOMIC CLASSIFICATION OF EXPENDITURES 62. Angola should avoid the temptation to increase spending on recurrent expenditures in the run up to elections. By looking at the economic classification of recurrent expenditures in dollar terms, the hypothesis of increased spending on recurrent expenditures in a pre-electoral period is confirmed. As seen in Table 1.5, total recurrent expenditures increased by 20 percent and 36 percent in dollar terms in 2004 and 2005, respectively. Excluding spending on interest payments, expenditures with personnel and goods and services increased by 38.6 percent and 50.2 percent, respectively, in 2005 alone. This pattern conforms with theoretical and empirical models that predict that pre-election spending tends to be biased toward recurrent expenditures such as wages and salaries, and transfers or subsidies (see Rogoff, 1990). 63. Improvements in the recording and classification of public spending are called for if the authorities want to adopt a meaningful pro-poor public expenditure policy. The authorities argue that the increase in personnel expenditures include the necessary hiring of new teachers and nurses, plus the annual wage adjustment for public servants which helps to inflate the wage bill. In addition, the authorities also argue that some expenditures on health and education are not classified as such, which contributes to keep the statistical level of social spending at a low level. This points to the need to improve further the quality of public financial management so that the government can gain a better understanding of its public spending and increase the effectiveness of financial planning and programming. Key to that end is the steady commitment to finalize the implementation of the SIGFE and the need to improve the classification of public spending. Once these steps are concluded, it will be easier to design and implement a pro-poor public spending program. 64. Expenditures with transfers and subsidies have remained constant as a share of GDP. As oil production has been rising fast and international oil prices remain high, GDP growth has been strong over the last few years and therefore the proportion of public spending as a share of GDP has declined since 2003. Total expenditures, for example, have dropped from 44 percent of GDP in 2003 to a much lower level of 31 percent of GDP in 2005. Total current expenditures have declined from 36 percent of GDP to 25 percent of GDP over the same period. Expenditures with personnel, and goods and services have also declined as a share of total output while expenditures on subsidies have averaged 5 percent of GDP during the 2003-05 period, which actually means that the latter have increased in absolute dollar value terms (see Table 1.5). In fact, total transfers have more than doubled in dollar terms since 2003 reaching $2.1 in 2005. Within total transfers, spending with fuel and utility price subsidies are the most important, totaling $1.3 billion in 2005. - 24 - Table 1. 5: Economic Classification of Current Expenditures, 2003-06 2003 2004 2005 2006 Proj. Proj. (Billions of US$, unless otherwise indicated) Expenditures 6.2 7.6 10.2 15.8 Current expenditure 5.1 6.1 8.3 10.3 Personnel 1.7 2.0 2.8 4.1 Goods and services 2.2 1.9 2.8 3.1 Interest payments due 0.3 0.5 0.6 0.7 Domestic 0.0 0.1 0.2 0.2 External 0.2 0.4 0.4 0.5 Transfers 0.9 1.7 2.1 2.4 Of which subsidies (oil & utilities) 0.7 1.3 1.3 1.5 (In percent of GDP; unless otherwise indicated) Expenditures 44.3 38.5 31.2 35.7 Current expenditure 36.3 30.6 25.2 23.4 Personnel 12.4 10.3 8.6 9.3 Goods and services 15.7 9.5 8.6 7.1 Interest payments due 1.8 2.3 1.7 1.5 Domestic 0.1 0.4 0.5 0.4 External 1.7 1.9 1.2 1.1 Transfers 6.4 8.5 6.3 5.5 Of which subsidies (oil & utilities) 4.8 6.6 4.0 3.3 Sources: Angolan authorities and IMF/World Bank staff estimates and projections. 65. Fuel prices in Angola are one of the lowest amongst African countries and other oil-rich countries elsewhere. Fuel subsidization is common practice not only in Angola, but in many other oil rich countries, such as Iraq, Iran, Saudi Arabia, Libya, Egypt, Algeria, Indonesia, Malaysia, Venezuela and Bolivia. Angola ranks fourth in a list of African countries with the lowest prices of gasoline and diesel. In 2004, for example, the US gasoline and diesel retail prices per liter were 54 and 57 cents, respectively, while in Angola, gasoline and diesel retail prices per liter in US dollar cents were 39 and 29. The size of the subsidies in Angola is even higher if one considers the pricing mechanism used by Sonangol, sometimes exceeding actual prices by over 300 percent for gas, gasoline, diesel and kerosene.3 66. There are also equally important subsidies in the provision of water and energy through the public utility companies. The rationale for the governmental subsidy transfer in the electricity and water sectors has to do with the inability of the public utility companies to raise sufficiently high revenue from sales to meet 3A detailed analysis of fuel price subsidies in Angola can be found in World Bank (2005). - 25 - their costs. This phenomenon has not only been the product of utility companies' inefficiencies in properly metering, billing customers for service delivered and preventing the surge in illegal connections, but also of the fact that electricity and water prices have been set by the Ministry of Finance at levels below the long run marginal cost. In the electricity sector, for example, these are affirmed to be US$0.11 per kWh, but the prices practiced by the utility are currently close to US$0.005 per kWh. In the water sector, in order to cover the financial loss of about US$6 million, the utility company has received price and operational subsidy amounts of about US$ 7.4 million and US$0.9 million, respectively. 67. The vast majority of the population considers all public services to be important but of poor quality. Mostly due to deterioration of physical and human capital during the long war, public service delivery of water and energy is neither constant nor comprehensive and is therefore of low quality. The provision of other important services by the government (such as public education, transportation, sanitation, and health) is also considered to be of low quality while at the same time considered to be highly important by the population. Poor households usually find it difficult to have access to formal public service delivery and have to recur to secondary markets for some services (such as water and some types of fuel) where they pay a high price. These households are potentially the beneficiary of gas and kerosene subsidization but not all can afford them and/or have easy access to these fuels. Fuel price subsidization overall has been found to benefit mostly the rich and not the poor. 68. The main beneficiaries of utility and fuel price subsidization are the richest households and not the poorest. The profile of the households who benefit from utility and fuel price subsidies reveals that the main beneficiaries are homeowners who reside in large houses and apartments, who own generators, refrigerators and computers, and who are in the third to the fifth income quintiles. More specifically, gasoline expenditure appears to be relatively more important for households within the fifth income quintile than for households within the other four quintiles because this energy source is highly used by well off households in the generation of lighting or to fuel their automobiles. These well off households are the ones who will be most directly affected by the removal of fuel price subsidies in the first instance while the poorest households could become better off if the amount spent on subsidies was more directly spent on improving their welfare. H. REFORM OF FISCAL MANAGEMENT: THE REMAINING CHALLENGES 69. Important weaknesses in public financial management were identified in the PEMFAR of 2005. The PEMFAR identified two parallel systems for expenditures: a "conventional" system coordinated by the National Treasury Directorate (DNT) at the Ministry of Finance and a "non-conventional" system, anchored on fiscal operations executed outside the budget by the national oil company, Sonangol. The PEMFAR pointed out that the dual structure had large transparency and efficiency costs. The report recognized that the dual structure had arisen as a result of the civil war, when Sonangol could lend expedience for a - 26 - variety of government expenditures. With the end of the war in 2002, however, Sonangol's continued significant fiscal involvement is no longer justified and could only blur the fiscal picture and deflect resources that Sonangol could better use in its own commercial activities. Furthermore, because Sonangol's fiscal expenditures were largely unbudgeted and were recovered through tax offsets and other noncash mechanisms, it was difficult to properly record and reconcile the transactions in the Treasury's Single Account (CUT) at the BNA, rendering recording, monitoring, and auditing costly and far from transparent. This situation remains largely unchanged to date. 70. Progress in implementing measures included in the PEMFAR action plan has been uneven. A summary of the progress observed in selected indicators is presented in Box 2 below. There has been progress in some important areas, such as in improving realism of macroeconomic assumptions in the budget preparation stage, and in the rollout of the SIGFE. The former is already an important step forward that helps policy makers with macroeconomic programming, while the second is a welcome development that contributes to increase transparency in the budget process. The reform in other areas has not yet benefited from the degree of political buy-in that would be necessary to guarantee steady cooperation from the different actors involved in the process. In particular, those elements that are more politically sensitive appear to have garnered less support for steadfast implementation than those that are of a purely technical nature. - 27 - Box 1. 2: Summary of Progress on Selected Recommendations of the PEMFAR Priority Actions Recommended in the PEMFAR Agency in Preliminary Assessment of Progress Charge Budget Preparation Increase realism of macroeconomic assumptions MINFIN Satisfactory ­ Strengthened macroeconomic (GEREI) management has helped projections to come closer to outturns. Adopt Oil Diagnostic's Financial Model to generate more accurate MINFIN (DNI, Moderate ­ Oil revenue forecasting model adopted by revenue projections DNO, GEREI) DNI as of April 1, 2006. Staff being trained to run model. Improve coordination between MINFIN and MAPESS on payroll MINFIN Partial ­ There is coordination, but information on stock information; establish a single cadastre for all public sector (DNO) and of civil servants continues deficient. personnel MAPESS Establish interface between SIGIP and SIGFE; turn the Public MINFIN (DNC) Unsatisfactory ­ Interface has not yet been established. Investment Program into a multi-year exercise and MINPLAN Multi-year PIP not yet adopted. Budget Execution Ensure consolidation of the Treasury Single Account (CUT) MINFIN, BNA Moderate ­ There are still significant delays in feeding revenue information to the CUT. Use financial programming projections as the ceilings for MINFIN (DNT) Moderate ­ Ceilings are being used but do not apply to commitments all budget units; Sonangol's QFA have no ceilings. Train key budget units staff to use the new version of the SIGFE MINFIN (DNC) Moderate ­ Training is ongoing and covers mostly staff working on the expenditure side. Enforce the correct sequencing of the budget execution process MINFIN (DNC, Moderate ­ Roll-out of SIGFE is progressing as planned (cabimentação -> liqüidação -> pagamento) , by making full use of DNT, INF) and is expected to be completed by 2007. Sonangol's SIGFE's automated controls QFA to follow correct sequencing before end-2006. Accounting and Reporting Reduce time allowed for submission of CGE to Tribunal of MINFIN (DNC) Satisfactory ­ CGE are closed in the first quarter, Accounts to six months after the end of the fiscal year approved by Council of Ministers and sent to Tribunal in the second quarter. Provide training on accounting and reporting practices on a MINFIN (DNC) Moderate ­ Training is ongoing, especially in relation to regular basis for relevant staff in the UOs use of SIGFE. Establish accounting and financial management procedures and MINFIN Moderate ­ Acquisition of non-financial assets recorded routines for non-financial assets (DNPE) in SIGFE since 2004; assessment of stocks as of end- 2003 yet to be finalized. Complete adoption of accounting manual as well as the new Chart MINFIN (DNC) Satisfactory ­ Both are now adopted. of Accounts. Ring-Fencing of Sonangol QFAs Ensure that the value of these activities is estimated in advance MINFIN Moderate ­ Values included in the budget ex post, but and included in the annual budget (DNO), in not yet estimated in advance. cooperation with Sonangol Agree what quasi­fiscal activities are to be identified and MINFIN (DNO, Unsatisfactory ­ There are frequent disputes about the measured within Sonangol for recovery (that is subject to the DNI, DNT), in amounts presented by Sonangol to MINFIN. No clear plan profit oil and tax offsets) cooperation to phase them out. with Sonangol Include such activities as a separate task within the independent Sonangol Unsatisfactory ­ Concessionaire activities not yet audits of Sonangol separated in corporate audits. Submit an independently audited analysis of unbudgeted "non- Sonangol Unsatisfactory ­ Not accomplished. Approval process is conventional"- type costs, in accordance with a Cabinet Minute to not straightforward. simplify and accelerate the approval process within MINFIN Extend the scope of the tax audits to include an examination of MINFIN (DNI) Unsatisfactory ­ Auditor reconciles tax filings with the calculation of net taxes due to the Treasury after netting off revised tax assessments and with payments made, but offsets and the value of quasi-fiscal activities netting off and QFAs of Sonangol not included. Formalize the foregoing in a new procedure approved by Cabinet Council of Unsatisfactory ­ No formal procedure to that effect is in Ministers place. Establish formal reporting mechanisms from Sonangol to MINFIN BNA, MINFIN Moderate ­ Formal reporting mechanisms do exist, but and to BNA (DNI, DNC, there is no clear plan and a calendar to phase out DNT, INF), Sonangol's QFAs. and Sonangol Eliminate fuel price subsidies through periodic price adjustments MINFIN Moderate ­ Policy to gradually eliminate subsidies implemented in 2004 and 2005, but halted with oil price increases in 2005. - 28 - Budget Preparation 71. The authorities have adopted a comprehensive model to forecast oil revenues. In April 2006 the National Tax Directorate (DNI) in the Ministry of Finance formally adopted the Oil Revenue Forecasting Model developed by Aberdeen University Petroleum Economics Consultancy (AUPEC) in the context of the Oil Diagnostic Study. A resident consultant from AUPEC is working full time with staff from DNI since April updating the model with information provided by oil companies and providing in situ training on the mechanics of the model. There have also been training trips by DNI staff to Aberdeen to learn more about Petroleum Economics. The AUPEC model represents a more effective tool to forecast oil revenues than the previous ad hoc process developed based on empiricism by DNI ­ a step in the right direction that is of critical importance as oil inflows grow and the macro fiscal stance is increasingly exposed to external shocks. 72. The integration of investment expenditures with recurrent spending has yet to be articulated. The Ministry of Planning designs the public investment program (PIP) while the budgeting of recurrent expenditures is the primary responsibility of the Ministry of finance. An interface between MINPLAN's information system for public investment (SIGIP) and the SIGFE has not yet been developed. Moreover, budgeting is increasingly complicated by the contraction of a large number of foreign lines of credit, each executed under different conditions and procedures, and managed out of different offices. The massive US$9.7 billion line of credit with the People's Republic of China, for example, is managed by the Office for National Reconstruction (Gabinete de Reconstrução Nacional ­ GRN), while the US$2 billion line of credit with China, contracted in 2004, is managed by an office created specifically to manage this loan in the Ministry of Finance (GAT ­ Gabinete de Apoio Técnico). This adds complexity to a system that is already characterized by weak capacity to program and assess investment projects. The consequence is that very often the full amount for multiyear projects are budgeted within a single year, while the cash required from the Angolan government to start up the projects is under budgeted. 73. There are lingering deficiencies in the budgeting of personnel expenditures. The budget is formulated based on the payroll database maintained by MAPESS. The information on personnel entered in the SIGFE is provided by the Informatics Support Office (Gabinete de Apoio Informático - GAI) at the Ministry of Finance. The GAI updates the MAPESS database by adding information on newly hired civil servants. The interface between the GAI database and the SIGFE is poor, and data are still entered into the SIGFE manually. SIGFE does not maintain detailed information on wage costs, which are contained in the two payroll databases; and the system can only access financial information on wages by spending units. In practice, thus, the government is using two different databases in parallel. Moreover, neither includes the military. Budget Execution 74. There has been progress in the rollout of SIGFE. The SIGFE became operational in 2004 covering the province of Luanda and by mid-2007 the system had expanded to the 18 provinces of Angola. The authorities' plans are to have the - 29 - system rolled out to missions abroad, and to some autonomous budget units by 2008. The rollout to the provinces has already yielded positive results. The authorities report that the spending units seem to be improving their compliance with the stages of budget execution envisaged in Angola's Organic Budget Law. As the Organic Budget Law requires, the SIGFE also includes a double-entry accounting system that is already operational, moving away from the old cash- based, single entry accounting system. 75. While recent progress has concentrated on the expenditure side, coverage of revenue inputs remains incomplete. The SIGFE is supposed to be fed with information on tax receipts at three stages of the tax collection process: (1) tax liabilities self-assessed by taxpayers; (2) tax liabilities that are then paid to the banking system; and (3) tax payments transferred from the banking system into the Treasury's single account (CUT) at the BNA. However, the SIGFE currently receives information only on the last stage--tax payments transferred to the CUT--and with a 30-day delay. Fuller data are managed by the Tax Directorate on electronic platforms that are incompatible with the SIGFE. Accurate information on accrued tax payments is crucial for financial programming and monthly cash plans for the budget execution units. If this is information is not available or is not accurate, it becomes very difficult to keep expenditures within the budget ceilings. This situation makes it urgent to develop an automatic interface between the systems or a tax-control system that might be linked directly to SIGFE. 76. The operations performed by Sonangol are now being recorded in the budget, but with a delay of up to 3 months. Government has started to strengthen the capacity of the Ministry of Finance to control expenditures and ring-fence the operations of Sonangol on behalf of the treasury, which are now recorded in the budget, but with a delay of 90 days. There have been attempts to install a SIGFE terminal at Sonangol to register these operations in real time, as it is the case with the other ordinary government expenditures, but this has not yet materialized. The Ministry of Finance acknowledges that it is necessary to ring-fence and phase-out these operations and also that it is necessary to separate the concessionaire and operator roles of Sonangol. However, government and Sonangol have both indicated that there will be no change on this configuration at least until 2010 due to institutional and technical limitations in the Ministries of Finance and Petroleum 77. Weaknesses in the management of financial asset stocks into the SIGFE also have to be overcome. There are significant problems with entering external and domestic debt transactions into the system, which threatens the adequacy of debt service. The two databases for debt management that co-exist at the BNA (DMFAS for external debt and GEMA for domestic debt) are incompatible with the SIGFE. Information provided the DNT are used by the BNA to update these databases. The result of this is a report produced by the BNA in a format that is incompatible with the SIGFE and a data-sharing process that tends to be subject to delays of up to three months. This situation will only be resolved when a debt control system that is directly compatible with the SIGFE's needs at the DNT is implemented. - 30 - 78. The CUT has yet to be fully consolidated. The mechanism of the CUT, explained in detail in the PEMFAR report, remains far from being consolidated. The main issues impairing its consolidation are related to some of the operations carried out by Sonangol and the lingering existence of the compensation mechanisms involving Sonangol and the Ministry of Finance. In principle, profit oil, Sonangol's taxes, the Kwanza-equivalent to the dollar-value of taxes paid by the foreign oil companies, and all other taxes transit through the CUT. However, since the CUT records only actual cash transactions, taxes and profit oil from Sonangol are often not recorded, due the above mentioned compensation mechanism. In addition to its complexity, this cash management system impairs planning and does not prevent over commitments. Budget Monitoring 79. There have been progresses in the management of non-financial assets and liabilities, but further improvements are necessary. Non-financial assets and liabilities management is under the responsibility of the National Directorate for the Management of Non-Financial Assets of the State (Direcção Nacional do Patrimonio do Estado, DNPE). A key concern in this area is to ensure that quality information on the non-financial assets of the State is generated in order to increase coverage and accuracy of the General State Accounts (Conta Geral do Estado, CGE). Indeed, lack of timely and accurate data on public debt and on State physical assets inventory prevents the production of reliable assets and liabilities accounting (contabilidade patrimonial). The SIGFE now allows DNPE to maintain an inventory of all new government non-financial assets at both central and provincial levels, but its records on old property and non-financial assets before the computerization of the system remain largely out of date. To update the inventory with past stock a task force will need to be created to collect the information all over the country and feed the SIGFE. There is also a need to strengthen compatibility and linkages between the budget accounting classification for expenditures and revenues and those for non-financial assets. 80. The internal control functions have shown some progress. In Angola, the internal audit function is carried out by the National Inspectorate of Finance (INF) while the external audit function is the responsibility of the Tribunal of Accounts. The Tribunal is the Supreme Audit Institution (SAI) whose sole function is to audit State accounts. Capacity for ex-post monitoring and auditing is still very poor, although there has been some improvement in the quality of returns from the provinces connected to the SIGFE. The Accounts Tribunal has officially adopted the double-entry Accounting Manual and Chart of Accounts underlying the SIGFE, which should help make more consistent the expenditure execution system and its ex-post auditing and monitoring functions. 81. Bolder improvements are needed on the auditing side, which remains weak. The major problem remains in the realm of the relationship between Sonangol and the Ministry of Finance. Sonangol, at the Government's request, undertakes a wide range of activities or tasks on behalf of the Government. In the past, the Government has sometimes failed to reimburse Sonangol, or has been delayed in doing so. Sonangol's reaction has been to reduce the tax and profit oil payments it owes to the Government by the amount of the costs it has incurred on Government's behalf. The consequences of this practice for effective revenue - 31 - management are adverse. Disputes arise because in the past there has not been clarity on which activities qualify for offset treatment, and because expenditures under qualifying categories have not been audited. Ring Fencing and Phasing Out of QFAs 82. Encouraging steps have been taken to incorporate Sonangol's quasi-fiscal operations in the budget, but major weaknesses have yet to be addressed. The budget includes estimates for Sonangol's key fiscal operations, including expenses on oil subsidies, outlays on goods and services, debt service, and accrued tax and non-tax receipts that the company owes or collects on the government's behalf. These are offset against each other in a procedure that is cumbersome, opaque, and subject to dispute. Because no cash payments are actually made, it causes substantial swings in arrears from month to month. The Tax Directorate staff is not yet fully able to assess the accuracy of the revenue information Sonangol presents. 83. The Government has moved to incorporate at least some of Sonangol's quasi-fiscal operations into the budget, but needs to go much farther. In 2005, the expenditure reports Sonangol sent to the MinFin were entered manually into the SIGFE. Sonangol's reports covered both budgeted and unbudgeted outlays; the latter were classified as unpaid "floating debt" when they could not be accommodated as regular budgeted expenditure types. However, the reports were not submitted every month but were sent together toward the end of the year. This prevented the MINFIN from assessing Sonangol's fiscal activities often enough, caused arrears to build up, and hampered intra-year fiscal programming. Moreover, MINFIN staff reportedly had problems classifying spending items that were not always clearly justified in Sonangol reports. On the way forward, an agreement on allowable offsets and planned audits could help a great deal, to start with. Ultimately the quasi-fiscal activities should be moved out of Sonangol. Although this cannot happen immediately, in the meantime all-quasi fiscal activities could be "ring-fenced" within Sonangol for better monitoring and control and ease of eventual transfer. The phases of a possible ring-fencing process proposed in the PEMFAR is described in Box 1.3. - 32 - Box 1. 3.: Phases of the "Ring-Fencing" Process First phase · First step - ensure that the value of these activities is estimated in advance and included in the annual budget. (This has started with the 2003 Revised Budget). · Second step ­ agree what quasi activities are to be identified and measured within Sonangol for recovery (that is subject to the profit oil and tax offsets). · The improved measurement of the associated costs would include the introduction of time writing techniques and cost allocation systems. · Third step ­ the foregoing transactions would be included as a separate task within the proposed independent audit of Sonangol. Second phase Submission by Sonangol of an independently audited analysis of unbudgeted expedient type costs, in accordance with a Cabinet Minute would simplify and accelerate the approval process within MINFIN. · The present independent audit of taxes paid should be extended to include an audit of the calculation of net taxes due to GOA after netting off offsets and the value of quasi-fiscal activities. · This would result in the whole cycle of tax flows being audited from the taxes paid by Sonangol to the final receipt of these taxes by MINFIN after the deduction of the offsets and quasi-fiscal activities. · Thus an independently audited full reconciliation between taxes originally due and net taxes finally received by MINFIN would be available. Third phase · Another important aspect would be that approval of Sonangol's tax returns should be strengthened by centralizing both the approval of gross taxes due and the value of offsets to arrive at the net taxes due and paid within MINFIN's Tax Directorate. · The foregoing should be formalized in a new procedure approved by Cabinet. Fourth phase Formal reporting mechanisms from Sonangol to MINFIN, and from Sonangol to BNA, need to be established, particularly regarding external debt service, tax and profit oil offsets, and oil exports undertaken by Sonangol. 84. The ring-fencing process should also aim at addressing the issue of conflict of interests associated with the activities of Sonangol. Certain aspects of Sonangol's role as concessionaire, notably its approval of all major procurement contracts create a significant potential for conflict of interest and loss of revenues to the Treasury. Further, the Government lacks the capacity to oversee effectively Sonangol's revenues and investment programs although their scale is enormous and has clear macroeconomic relevance. Tempered by expedience and by the need to build capacity in oversight agencies, consideration should be given to transferring Sonangol's concessionaire roles to the Ministry of Petroleum, where they are traditionally found in international practice. Ring-fencing these activities first within Sonangol would help prepare the transfer (see Box 1 above). The urgency of proper oversight of Sonangol is such that GoA/MINFIN might be well- - 33 - advised to engage qualified consultants sooner rather than later to assist in performing this function. 85. A well defined and time-bound plan to achieve normalization in this process should be adopted by the authorities. "Normalization" refers in this context to a situation in which the "non-conventional" or "expedient" activities, to use the language of the PEMFAR, would have been eliminated over an agreed time period. This would represent the arrival point of a transition process involving the key institutions - MINFIN, BNA, and MINPET. However attention must be paid as part of the transition process to upgrade the resources, skills, training, internal procedures and reporting and communication within and among these institutions in order for them to be ready to take back ownership of the Treasury-like activities currently performed by Sonangol. These changes would involve improvements to salary structure, employee career planning and benefits, all of which would require commitment to institutional reform. - 34 - Table A. 1: Multi-year Perspective to the Adoption of an MTEF Item Multi-year Feature PEM Responsible Intended Effect Requires Angola Objective Agent 1 Macroeconomic Macrofsical Finance Provides strategic framework for Forecasting model, capacity, multi-year Yes for the oil forecast discipline Ministry setting fiscal and monetary macro variable time series, or access to sector. policy, and allows proactive multiple non-governmental forecasts Rudimentary for fiscal adjustment to changing the real sector. economic trends. 2 Multi-year revenue, Macrofiscal Finance Sets upward bound for Forecasting models. Stronger if Yes, but no debt sustainability discipline Ministry expenditures, limiting deficits, relationships between macro growth, cabinet-level analysis and debt inflation, and currency income distribution, and revenues discussion, policy, yielding depreciations; supports understood and modeled. Debt approval. Basis expenditure sustainable fiscal policy, and sustainability analysis/model, or hard for debt envelope realistic expenditure planning rule on debt/deficit limits. More effective sustainability within the expenditure envelope; in changing behavior if aggregate analysis supports focus on adequate expenditure ceilings/policy debated and uncertain. revenue mobilization. approved by cabinet. Insufficient treatment of revenue policy. 3 Multi-year forecast Macrofsical Finance Provides broad indicator of MoF provided inflators for pay, non-pay, No. Bi-annual of spending under discipline, Ministry, or in future cost of current spending and clear guidance for projecting costs. economic current policy or Allocational some cases trends, identification of potential Can be automated. Can be budget year programs exist, current level of efficiency spending risk areas, and proactive, only, but more effective over several but there is no services, by ministry (sector) ministries with measured, more rational fiscal years. detailed multi- or program clear guidance adjustment. Provides baseline year forecast of for evaluating policy spending spending. choices, including pay increases, relative to Item 2, and measuring sectoral reallocation decisions.. Some incentive to examine cost- drivers for programs/sectors. 4 Multi-year ceilings Allocational Finance Provides framework for sector Can be developed from Item 3, or from No. Depends on for sector ministries efficiency Ministry ministry budget preparation, item 2 expenditure limits proportionally progress made (sector), enabling more realistic planning allocated, but these would not reflect on item 2. Macrofiscal by sectors, designing reallocation decisions. More credible if discipline, appropriate policies; provides reflecting policy choices, which requires Operational more incentive for reviewing some explicit policy directions on efficiency existing programs for reallocation. Even more credible in effectiveness, making trade-offs terms of providing multi-year and reallocations within sectors. expenditure planning targets for At center of Government, allows spending ministries if they reflect explicit decisions on trade-offs understanding of cost inefficiencies and between sectors. sectoral marginal returns/outcomes, but this requires more detailed understanding of sectors. More effective in changing behavior if approved by cabinet or parliament. 5 Multi-year sector Allocational Spending Sector strategic plan that may Strategic planning capacity at sector Some ministries strategy efficiency Ministry link outputs/outcomes with ministry, information on may have (sector), inputs in multi-year framework. outputs/outcomes of programs, and begun to Operational Effective with respect to relationship to activities and inputs. develop these efficiency, macrofiscal discipline and (Health, Macrofiscal allocational efficiency only if Education, discipline prepared within multi-year Agriculture), sectoral resource ceiling. though in the absence of multiyear resource ceilings. 6 Multi-year estimates Operational Spending Identifies multi-year implications Requires guidance/training for spending Partially. of cost of new efficiency, Ministries of new initiatives relative to their ministry staff, and spending ministry Spending policies or Allocational objectives, and assessment of staff capacity; MoF provision of ministries have programs efficiency whether they can be financed common inflators for use by ministries their own cost (recurrent), or (sector), from within existing sectoral (pay rates, non-pay, capital costs). Less estimates, but expansion of Macrofiscal ceilings or even within aggregate effective in the absence of multi-year these are used existing programs, discipline spending ceilings, and if they are sector ceilings (item 4) and cost of for their own prepared by sector financially sustainable over time. existing programs (item 7) and sector consumption as ministries strategy (item 5). there is no consolidated multi-year spending plan. - 35 - 7 Multi-year estimates Operational Spending Similar to Item 3, but prepared Can begin at program level, and Partially. Some of cost of existing efficiency, Ministries by sector ministry. Sensitizes progressively push down for deeper spending policies, programs, Allocational sector ministry to cost drivers, bottom-up approach in future years, ministries have subprograms, or efficiency affordability of existing policy or adding subprogram and then activity their own activities prepared (sector), programs, attention to different costing. Requires trained staff at estimates, but by sector ministries Macrofiscal means of attaining objectives, spending ministries, agencies. Requires these are done discipline unit cost. MoF provision of guidance and common independently inflators for use by ministries (pay rates, and without non-pay, capital costs). Less effective in guidance from the absence of multi-year sector ceilings the Ministry of (item 4) and cost of new programs Finance. (items 6 and 8) and sector strategy (item 5). 8 Multi-year estimates Operational Spending Similar to item 6, but for capital Requires trained staff at spending Partially. of cost of new efficiency, Ministries projects. Many capital budget ministries, guidance on costing, Sectoral projects (capital), Allocational processes already include such understanding of project design and estimates exist or expansion of efficiency estimates, including the work flow, to yield good estimates. Less and are existing projects, (sector), recurrent cost implications of effective in the absence of multi-year included in the prepared by sector Macrofiscal new capital projects. sector ceilings and cost of existing PIP, but the PIP ministries discipline programs (items 6 an d 7) and sector is not yet a strategy (item 5). multi-year instrument. Source: The World Bank - 36 - CHAPTER II PUBLIC SPENDING IN EDUCATION A. RECENT PERFORMANCE IN EDUCATION 86. The education system in Angola is in transition. Between 1999 and 2005, the Angolan education system was under two different Learning and Education Systems enacted legally, namely one from 1977 (approved by the Decree 40/80 from 14th of May) and the second one from 2001 (approved by the Law 13/01 from 31st of December), which started being implemented in 2004. 87. The Learning and Education System from 1977 which was in place until 2004 had the following main features: · A General Basic Education sub-system composed of 8 grades, structured in three levels, the first of which with 4 grades (compulsory and with the official age for enrolment of 6 years) and the other two with 2 grades each, which should also become compulsory when conditions allow that to happen (4+2+2). According to the international classification of UNESCO, the 2nd and 3rd level of primary education in Angola are equivalent to the international 1st cycle of General Secondary Education. This sub-system also contains an initiation grade (classe de iniciação) which aims at adapting the child to a school environment before enrolling in the 1st grade. Throughout this chapter the overall sub-system of basic education and each of its three levels will be named primary education to facilitate the understanding of the issues involved and the establishment of international comparisons; · A Pre-University Education sub-system initially established as a `transition module' between the last stages of the Secondary Education from the colonial system and the new education system, for access to Higher Education. Initially structured in 4 semesters it changed in 1986 to 6 semesters; · A Medium Education sub-system, with 4 years and two fundamental specializations: Technical and Normal, the first one aiming at training intermediate technicians for the productive sector and the second one aiming at training teachers for the General Basic Education; · A Higher Education sub-system, structured in Faculties, with average duration of 5/6 years. - 37 - 88. The new Learning and Education System which started to be implemented in 2004 has the following main features: · A Primary Education sub-system composed of 6 grades, structured in one level (compulsory and with the official age for enrolment of 6 years). This sub-system continues to maintain an initiation grade (classe de iniciação) which aims at adapting the child to a school environment before enrolling in the 1st grade; · A Secondary Education sub-system structured in two cycles, each of three years. The first cycle of secondary education comprises grades 7 to 9 and the second cycle grades 10 to 12. initially established as a `transition module' between the last stages of the Secondary Education from the colonial system and the new education system, for access to Higher Education. Initially structured in 4 semesters it changed in 1986 to 6 semesters; · A Medium Education, with 3 to 4 years and two fundamental specializations: Technical and Normal, the first one aiming at training intermediate technicians for the productive sector and the second one aiming at training teachers for the General Basic Education (in both cases students must already hold the 9th grade). The normal medium education aiming at training teachers for grades 1 to 6 and it's undertaken in the Institutos Médios Normais (IMN) and the technical medium education is undertaken in the so-called Institutos Médios Técnicos (IMT); · A Higher Education, structured in Faculties and different types of degrees (diploma, bacharelato, licenciatura, master and PhD). A Diploma can be obtained with 3 years of higher education, a bacharelato can be obtained after 4 to 5 years and licenciatura after 6 years. In addition, a master degree, depending on the field of specialization can be obtained after 2 or 3 years of post-licenciatura studies and a PhD after 4 or 5 years of post-licenciatura or post-master studies. Literacy, access and enrolment 89. Since the statistical data available from the Ministry of Education is only consistent in terms of enrolment and performance data for the three levels of primary education (EBR 1st, 2nd and 3rd levels), it is not possible to include a consistent analysis of the performance of the education system for the secondary, technical and higher levels of education. For these two sub-systems only general data will be presented. (see the annex for the definition of the ratios used in this chapter). 90. The official documents from the Ministry of Education suggest the following illiteracy rates in the country: · North Region: 65% (Luanda 65%, Bengo 80%, Cabinda 45%, Zaire 70%, Malange 65%, Kwanza Norte 65% and Uíge 65%); - 38 - · Central Region: 61.25% (Benguela 60%, Kwanza-Sul 70%, Huambo 50% and Bié 65%); · East Region: 71.6% (Lunda Norte 70%, Lunda Sul 65% and Moxico 80%); · South Region: 79% (Huíla 60%, Namibe 50%, Kuando-Kubango 80% and Cunene 75%). 91. Table 2.1 presents the absolute number of pupils per level of education in 1999 and 2003, the percentage growth in the period and as a share of total number of students in each level in 1999 and 20034. It is worth noting that the number of pupils more than doubled in all levels of primary education from 1999 to 2003 and that primary education always represented more than 90 percent of the total enrolment in the Angolan education system (92.8 percent in 1999 and 95,6 percent in 2003). The increase from 1999 to 2003 in the proportion of enrolment in primary education compared to total enrolment in the system was mostly due to the increase in enrolment in the 1st level of PE. 92. Secondary education represented 6.5 percent of the total of pupils enrolled in the education system in Angola in 1999 but only 4.0 percent in 2003. Higher education also decreased their importance in relative terms and represented both in 1999 and 2003 less than 1 percent of the total enrolment in the public system of education. Table 2. 1: Absolute number of pupils per level of education in 1999 and 2003, % growth in the period and proportion in total number of students of each level in 1999 and 2003. % Proportion in Proportion in 1999 2003 growth total 1999 total 2003 Total 1.306.223 2.779.436 112.8 Primary education 1.212.227 2.656.928 119.2 92.8 95.6 1st level 983.963 2.172.772 120.8 75.3 78.2 2nd level 152.929 319.502 108.9 11.7 11.5 3rd level 75.335 164.654 118.6 5.8 5.9 Secondary 85.460 109.942 education 28.6 6.5 4.0 PUNIV 17.331 21.771 25.6 1.3 0.8 Medium Normal 28.342 43.945 55.1 2.2 1.6 Medium Technical 39.787 44.226 11.2 3.0 1.6 Higher Education 8.536 12.566 47.2 0.7 0.5 Source: MEC statistics. 93. As far as the gross enrolment rates are concerned, there has been a positive trend in all the three levels of primary education since 1999 to 2003, with a more expressive growth from 2002 to 2003 (see Figure 2.1 and Table 2.2 below), with the end of the civil war. In all levels of PE there was a growth in enrolment of around 90 percent during the period, which means that the proportion of students 4In terms of access rates, since the data available was very scarce and related only to 2003, it is not possible to make an analysis of the Gross Admission Rate (GAR). In what follows, instead, the focus is turned to the Gross and Net Enrolment Rates (GER and NER) from 1999 to 2003, in the first case, and of 2001 and 2002, in the second case. - 39 - enrolled compared to the official school-age population in the period grew considerably and in the 1st level of PE in 2003 even outnumbered the official school-age population for that year. However, it should be noted that even though there was an impressive growth in the GER in all three levels of PE, the differences in the rates between the levels is very substantial, indicating the low levels of access to education above grade 4 and low levels of efficiency of the education system in general. Table 2. 2: Gross Enrolment Rates for the 1st, 2nd and 3rd levels of primary education from 1999 to 2003. 1999 2000 2001 2002 2003 GER 1st level EP 68.7 75.7 82.8 96.0 134.7 GER 2nd level EP 22.4 25.4 29.6 30.8 41.5 GER 3rd level EP 11.5 13.0 14.7 16.1 22.3 Source: MEC statistics. Figure 2. 1: Gross Enrolment Rates from 1999 to 2003 in the three levels of primary education. Gross Enrolment Rates 150 100 1st level EP 2nd level EP 50 3rd level EP 0 1999 2000 2001 2002 2003 Years 94. In relation to the GER in the secondary, medium and higher sub-systems of education, Table 2.3 presents the figures obtained, despite the very strong limitations in the data collected. The very low figures clearly illustrate the squeezed pyramid representing the education system in Angola. Table 2. 3: Gross Enrolment Rates for secondary, medium and higher sub-systems of education for 1999 and 2003 1999 2003 GER in higher education 0.7 1.0 GER in secondary and medium levels 6.7 7.7 GER in PUNIV 1.4 1.5 GER in EMN 2.2 3.1 GER in EMT 3.1 3,1 Source: MEC statistics. - 40 - 95. Regarding the NER, again not much can be assessed since only data for 2001 and 2002 was available (see Table 2.4 below5). Assuming that the data has some degree of consistency, the NERs were very low both in 2001 and 2002. Table 2. 4: Net Enrolment Rates for the 1st, 2nd and 3rd levels of primary education for 2001 and 2002. With age group With single age 2001 2002 2001 2002 NER 1st level EP 49.0 47.0 19.0 16.0 NER 2nd level EP 7.0 6.1 4.0 3.6 NER 3rd level EP 3.6 3.3 2.4 2.0 Source: MEC statistics. 96. The comparison between the GER and NER in 2001 and 2002 for all three levels of PE indicates that the majority of the students enrolled in any level of PE are not of the official school-age for the specific levels. This is most likely the result of a combination of factors such as late entry in the 1st level of PE and high levels of repetition and drop-out rates (to be analyzed in the following section). 97. When compared to countries in the region, the challenges that Angola faces in terms of establishing similar patterns of educational results become clearer. While Angola had a NER in primary education (grades 1-4) of 49 percent and 47 percent in 2001 and 2002, respectively, Botswana had 80 percent of NER in primary education in both years; Cape Verde more than 99 percent in both years; Guinea Equatorial 84,6 percent in 2001; Gabon 78 percent in 2000; Mozambique 60 percent in 2000 (grades 1-5); South Africa 89 percent in 2000 and 2001; Tanzania 69 percent in 2001 and Zambia 66 percent in 2000 and 68,4 percent in 2001. Efficiency of Educational System 98. All three levels of primary education have shown very high repetition and drop-out rates. Whereas the drop-out rates have decreased between 1999 and 2002 in all three levels of PE, the repetition rates have increased in the same period for all the levels of PE (see Table 2.5 and Figure 2.2 below). In addition, the least efficient level of primary education both in terms of repetitions and drop-outs is the 1st level, from grade 1 to 4. The difference in the repetition rates between the 2nd and 3rd levels of PE is considerable, with the 3rd level being slightly more efficient. However, both levels are almost at the same level of efficiency regarding the drop-out rates. 5 The table presents the net enrolment rates considering the specific age group for all grades (e.g. all children from 6 to 9 years from grade 1 to 4 considering children 9 years old in grade 1) and the single specific age for each grade (e.g. in grade 1 only 6 years old children are considered or in grade 4 only 9 years old children are considered). - 41 - Table 2. 5: Repetition and Drop-out Rates for the 1st, 2nd and 3rd levels of primary education from 1999 to 2002. 1999 2000 2001 2002 Repetition Rates 1st level EP 26.7 27.1 29.1 29.4 2nd level EP 24.1 22.7 26.1 28.2 3rd level EP 20.6 20.8 21.0 24.1 Drop-out Rates 1st level EP 21.1 17.2 15.1 14.8 2nd level EP 12.4 11.0 12.2 9.8 3rd level EP 13.2 10.9 12.1 9.4 Source: MEC statistics. Figure 2. 2: Repetition and Drop-out Rates in the 1st, 2nd and 3rd levels of primary education from 1999 to 2002. Repetition Rates Drop-out rates 35 25 30 20 25 1st level PE 1st level PE set 20 15 2nd level PE set 2nd level PE 10 Ra 15 Ra 3rd level PE 3rd level PE 10 5 5 0 0 1999 2000 2001 2002 1999 2000 2001 2002 Years Years 99. Angola ranks low in comparison to its neighboring countries in terms of repetition rates. While Angola had a repetition rate in grades 1 to 4 of 29.4 percent in 2002, Botswana had 3.2 percent, Namibia had 13.1 percent, Zambia 7.6 percent and even Mozambique which also faces serious problems of efficiency in primary education had a repetition rate for grades 1 to 5 of 23 percent. 100. Table 2.6 presents the situation in terms of completion and modified completion rates (CR and MCR) from 1999 to 2002.6 Both the completion and modified completion rates increased from 1999 to 2002, showing a positive trend. However, both the number of students graduating in the final year of each primary cycle of education as a proportion of the population of official graduation age and the total enrolment in the final year of each primary cycle of education as a proportion of the population of official graduation age were very low compared to international standards. 6The definition of completion rate used in this report is the total number of students graduating from the final year of a cycle of education e.g. primary, regardless of age, expressed as a percentage of the population at the official graduation age (see annex 3 at the end of the Chapter). - 42 - Table 2. 6: Completion and Modified Completion Rates for the 1st, 2nd and 3rd levels of primary education from 1999 to 2003. 1999 2000 2001 2002 Completion Rates 1st level EP 25.8 30.6 31.1 33.2 2nd level EP 12.0 15.1 14.0 16.9 3rd level EP 7.4 8.9 8.9 10.0 Modified Completion Rates 1st level EP 40.6 48.3 51.3 56.4 2nd level EP 18.8 21.3 23.3 25.2 3rd level EP 10.9 12.4 12.4 14.0 Source: MEC statistics 101. When compared to countries in the region, in particular for the 1st level of primary education, Angola faces a big challenge in order to attain the regional levels and the international goals. While Angola had a completion rate of 31.1 percent in 2001 and 33.2 percent in 2002 for grades 1 to 4, Gabon, a neighboring country had 77 and 74 percent, respectively, in primary education (grades 1 to 5); Nigeria had a completion rate in primary education of 82 percent in 2001; Zambia had a CR of 60 percent in 2001 and 69 percent in 2002; Botswana had a CR of on average 91 percent in the period from 1999 to 2002; Mozambique had a CR in 2002 of 52 percent (see Figure 2.4). Figure 2. 3: Primary completion rates in 2002 in Sub-Saharan Africa (SSA) countries. Primary completion rates % 2002 91 100 82 90 74 69 80 70 52 60 % 50 33,2 40 30 20 10 0 Angola Botswana Gabon Mozambique Nigeria Zambia Countries Source: UNESCO internet database. Note: The PCR for Nigeria presented in the figure relates to 2001. 102. The transition rate between the 1st level of PE and the 2nd level of PE was 63.3 percent in 2000 and 58.6 percent in 2002 while the transition rate between the 2nd level of PE and the 3rd level of PE was 66.5 percent and 66.9 percent, over the same period 2002 (see Table 2.7 below). This means that in 2002 only 58.6 percent of the students who reached the 4th grade continued to study in the 5th grade and only 66.9 percent of the students who reached the 6th grade continued to study in the 7th grade. It should also be recalled that a large percentage of students will already have dropped out before the 4th and 6th grades and thus one could argue that the transition rate provides a skewed picture (in a positive direction). However, it is interesting to note that the transition rate decreased by 4.7 percentage points from 2000 to 2002 in the 1st level of PE. In addition, the transition rate from the 2nd to the 3rd levels of PE increased slightly by 0.4 percentage points from 2000 to 2002. - 43 - Table 2. 7: Transition and Survival Rates for the 1st, 2nd and 3rd levels of primary education in 2000 and 2002 Percentage point 2000 2002 variation Transition Rates 1st level EP 63.3 58.6 -4.7 2nd level EP 66.5 66.9 0.4 Survival Rates 1st level EP 10.7 8.8 - 1.9 2nd level EP 5.0 3.6 - 1.4 3rd level EP 2.4 1.7 - 0.7 Source: MEC statistics 103. Based on repetition and drop-out data from 2000 and 2002, Table 2.7 shows the cohort survival rates for the different levels of education. Using 2002 figures, the cohort analysis shows that only 88 pupils out of a cohort of 1000 would reach each successive grade of the 1st level of PE without repeating or dropping out until they would complete the 4th grade; only 36 of these would reach and complete the 6th grade without repeating or dropping out and after 8 years only 17 pupils of the initial cohort of 1000 would have reached and completed the 8th grade (after 9 years, 36 pupils and after 10 years 43 pupils). The survival rates are extremely low due to the high repetition and drop-out rates in the country. 104. As a point of comparison, in Zambia, the survival rate to grade 7 was 63 percent in 2000, or 60 percentage points higher than in Angola if we consider the survival rate to grade 8 in 2000. Other more recent examples are the proportion of children reaching grade 5 (% of cohort) in countries like Botswana, Gabon, Namibia, Zambia (neighboring countries of Angola) and even Mozambique which also faces the consequences of a very long period of civil war after independence. Figure 2.5 shows that in 2002, Botswana had 87.6 percent of pupils, from a cohort, reaching grade 5; Gabon 69.3 percent, Zambia 98,5 percent, Namibia 94.7 percent and Mozambique 49.2 percent. In addition, it is of concern to note that the survival rates to have decreased even more from 2000 to 2002 in all levels of education. Figure 2. 4: Pupils reaching grade 5 (% of cohort) in 2002 in Sub-Saharan Africa countries Pupils reaching grade 5 (% cohort) 2002 98,5 94,7 87,6 100 69,3 80 49,2 60 % 40 8,8 20 0 Angola Botswana Gabon Mozambique Namibia Zambia Countries Source: World Bank internet database ­ http://devdata/worldbank.org/edstats - 44 - B. ANALYSIS OF EQUITY ISSUES Defining equity 105. In the context of education, equality of provision of services is defined as equal access to quality education for all sections of the population, independently from where they live, which social and income group they belong to or whether they are male or female. Equity in output/outcomes is measured in terms of basic indicators such as enrolment, repetition, drop-out, completion, survival and literacy rates. 106. Geographical disparities in Angola occur between and within provinces as well as municipalities and communes and rural and urban locations. Angola has 18 provinces and 163 municipalities. The provinces with the highest proportion of school age-children are Luanda (north region), Benguela (central region), Huíla (south region) and Huambo (central region). Provincial Inequalities 107. Regarding the gross enrolment rates per province in the 1st level of PE, the disparities are quite clear and presented in both Table 2.8 and Figure 2.6 below. The same disparities were found for both 2nd and 3rd levels of PE. 108. In 1999, while the highest GER was 141.9 percent in Bengo, the lowest (excluding Namibe), was 29.7 percent in Kuando-Kubango, illustrating very clearly the high level of inequality (112.2 percentage points difference). In 2003, while the highest GER was 323.2 percent in Huíla the lowest was 77.4 percent in Moxico, which shows the increasing provincial inequality as far as this indicator is concerned (248.5 percentage points difference). Table 2. 8: Gross enrolment rates from 1999 to 2003 per province in the 1st of PE. 1999 2000 2001 2002 2003 National 68.7 75.7 82.8 96.0 134.7 Bengo 141.9 126.1 119.4 151.8 158.2 Benguela 62.4 64.9 81.4 74.7 82.3 Bié 54.3 50.4 68.8 77.8 147.8 Cabinda 122.6 167.9 110.9 111.6 137.4 Huambo 34.9 67.1 83.3 192.7 189.4 Huíla 132.3 120.6 130.8 154.7 323.2 Cunene 52.2 56.7 92.4 100.1 196.3 Kuando Kubango 29.7 49.8 56.9 62.3 124.8 Kuanza Norte 92.9 91.8 98.7 95.6 109.8 Kuanza Sul 94.3 81.9 85.0 82.5 160.0 Luanda 72.4 76.9 71.6 74.5 77.5 Lunda Norte 39.4 37.3 50.7 53.0 89.6 Lunda Sul 44.3 43.0 88.9 91.1 121.8 Malange 36.3 75.7 96.0 137.2 215.3 Moxico 46.9 53.7 80.3 77.9 77.4 Namibe 12.0 139.3 162.5 171.8 179.0 Uíge 41.5 53.0 52.8 51.2 99.0 Zaire 69.2 67.8 82.7 81.3 105.2 Source: MEC statistics. Note: In Namibe the difference in the GER in 1999 and other years, is explained by the difference in total number of pupils enrolled in 1999 to 2000 (the information provided says that in 1999 there were 1,953 pupils enrolled while in 2000 there were 23,298 pupils). If we were to consider the latest number in 1999, the GER in Namibe would be 143.5 percent and the national GER would change from 68.7 to 70.2 percent). It was decided to leave the figures as given but the GER from Namibe will not be considered as a province for comparison in 1999. The difference in - 45 - the GER in Huíla from 2002 to 2003 is also explained by an impressive increase in the total number of pupils from one year to the other (from 218,607 in 2002 to 470,434 in 2003). Figure 2. 5: Gross enrolment rates in the 1st level of PE per province in 1999 and 2003. Gross enrolment rates in 1st level of PE 350 300 250 200 1999 150 2003 100 50 0 é bo al a a et l ge eri onali ngoe Bi etr l co bei Huí B uando anzu No anzu Su Su xi Uí Lunda Nor Lunda langea m Za Nat enguela Cabinda Huam Cunene K bangou K K Luanda Mo B K M Na Provinces 109. It is also important to note also that the provinces of Benguela and Luanda showed an insignificant increase in the GER from 1999 to 2003 in the 1st level of PE, compared to the national average increase. The same levels of high inequality are also found in relation to the net enrolment rates. For example, whereas in 2001, in the 1st level of PE Lunda Sul had a NER of 2.0 percent and 3.5 percent in 2002, Bengo, for example, had a NER of 48.4 percent in 2001 and 60.3 percent in 2002. Repetition and Completion Rates 110. There are some intriguing inter-province patterns in what concerns repetition rates. Considering the 1st level of PE in the provinces of Cabinda and Kuando Kubango, for example, it is clear that further analysis should be done to assess the reasons why there is a difference of around 20 percentage points on average in repetition rates in that level of education between those provinces. The same applies when comparing Luanda and Huíla or Lunda Norte in the 2nd level of PE (see Table 2.9). 111. Another interesting questions is understanding the reasons why provinces perform much better in some levels of education than others (examples: Luanda presents lower repetition rates in grades 5 to 8 but very high repetition rates in grades 1 to 4 or Lunda Norte with very high repetition rates in grades 5 and 6 whereas there is a significant lower inefficiency in grades 1 to 4 and 7 and 8), implying that more than an overall structural education system problem there seem to be problems at specific levels of education that need to be solved. - 46 - Table 2. 9: Repetition Rates per province in the 1st, 2nd and 3rd levels of PE in 2000 and 2002 1st level of PE 2nd level of PE 3rd level of PE 2000 2002 2000 2002 2000 2002 National 27.1 29.4 22.7 26.1 20.8 21.0 Bengo 27.6 26.0 20.1 Benguela 26.9 30.2 33.2 35.1 32.8 34.8 Bié 15.9 25.9 28.9 38.3 15.2 49.2 Cabinda 35.9 38.5 35.0 36.4 36.4 32.4 Huambo 20.3 25.0 5.9 25.3 29.2 16.6 Huíla 29.2 32.5 39.3 45.3 51.0 43.1 Cunene 30.6 37.9 30.2 41.9 32.6 46.6 Kuando Kubango 16.9 15.9 33.7 27.3 30.0 34.2 Kuanza Norte 27.4 29.6 36.6 39.8 31.4 40.7 Kuanza Sul 28.2 32.4 31.7 34.8 50.0 29.0 Luanda 27.1 28.7 16.5 17.1 12.7 17.1 Lunda Norte 17.9 26.4 34.3 49.1 9.5 24.7 Lunda Sul 29.8 29.8 25.7 28.7 27.2 37.5 Malange 23.7 17.4 22.5 46.2 12.2 23.8 Moxico 21.0 21.6 33.3 36.5 29.9 38.3 Namibe 32.3 37.1 34.1 28.0 30.0 31.3 Uíge 28.2 32.8 30.1 32.2 18.8 27.3 Zaire 32.9 34.3 29.8 28.2 27.0 29.9 Source: MEC statistics. 112. Regarding the completion rates, once again the inequalities are very clear with differences among provinces higher than 30 percentage points in the case of the 1st and 2nd levels of PE (e.g Kuando Kubango and Luanda in 1999 and Lunda Norte and Luanda in 2002) and higher than 20 percentage points in the case of the 3rd level of PE (e.g. Luanda and Kuando-Kubango in 1999 and Luanda and Cunene in 2002), even though all of the provinces present very low completion rates (Table 2.10). Table 2. 10: Completion Rates per province in the 1st, 2nd and 3rd levels of PE in 1999 and 2002. 1st level of PE 2nd level of PE 3rd level of PE 1999 2002 1999 2002 1999 2002 National 25.8 33.2 12.0 16.9 7.4 10.0 Bengo Benguela 23.7 26.4 7.9 10.9 3.8 4.1 Bié 10.7 20.9 5.7 15.9 3.0 6.1 Cabinda 27.4 41.6 10.2 21.4 6.9 10.9 Huambo 15.8 25.1 4.4 11.5 1.8 6.0 Huíla 37.2 50.5 9.5 11.8 2.9 4.0 Cunene 10.2 18.1 3.0 4.5 1.2 1.3 Kuando Kubango 9.9 15.2 1.2 5.2 0.4 2.5 Kuanza Norte 20.6 29.4 5.4 5.8 1.7 1.7 Kuanza Sul 19.4 21.5 0.7 6.6 1.3 4.7 Luanda 47.8 53.4 31.3 37.3 21.7 26.7 Lunda Norte 12.0 13.5 3.9 3.1 1.5 1.6 Lunda Sul 15.4 17.3 6.4 4.4 1.7 2.0 Malange 0.0 15.9 0 6.1 0.0 3.2 Moxico 7.0 16.3 3.0 6.1 1.2 2.1 Namibe 0.0 52.5 0.0 27.6 0.0 12.5 Uíge 9.7 13.7 4.1 5.2 2.6 1.8 Zaire 25.4 34.2 6.0 12.8 1.8 4.4 Source: MEC statistics. - 47 - Gender inequalities 113. Taking into account the data included in Table 2.11 and Figure 2.7 below it should be noted that, firstly, girls have a more restricted access to primary education in general in relation to boys; secondly, that this inequality in access is stronger in the first level of PE and is less visible in higher grades; and, thirdly, that the differences increased from 1999 to 2003. Table 2. 11: Gross Enrolment Rates per gender in 1999, 2001 and 2003 per level of primary education 1999 2001 2003 Male Female Male Female Male Female GER 1st level of PE 72.4 65.0 90.3 75.3 166.9 102.4 2nd level of PE 24.0 20.8 32.9 26.3 49.3 33.8 3rd level of PE 11.7 11.3 15.9 13.5 27.4 17.3 Source: MEC statistics. - 48 - Figure 2. 6: Gross Enrolment Rates per gender in 1999, 2001 and 2003 per level of primary education Gross Enrolment Rates by Gender 180 160 140 120 1st level of PE 100 2nd level of PE 80 60 3rd level of PE 40 20 0 Male Female Male Female Male Female 1999 2001 2003 114. Regarding the repetition rates, the trend is for girls to repeat less than boys, with few exceptions, and for their rates to show a more positive evolution than those of boys in all three levels of PE (Table 2.12). 115. In relation to the drop-out rates, again girls tend to drop-out less than boys with few exceptions. In terms of evolution, the reverse happens if compared with the repetition rates, which means that the drop-out rates for boys have shown a more positive trend than those for girls. Table 2. 12: Repetition, drop-out and completion rates from 1999 to 2002 per gender and level of primary education. 1999 2000 2001 2002 M F M F M F M F Repetition rates 1st level of PE 14.2 14.6 13.9 14.1 15.2 13.9 13.5 13.3 2nd level of PE 15.0 13.0 14.2 12.4 13.4 11.4 14.5 12.1 3rd level of PE 11.1 11.0 12.3 10.4 11.1 9.4 11.8 10.8 Drop-out rates 1st level of PE 22.8 19.2 18.3 15.9 14.2 16.3 13.6 16.3 2nd level of PE 13.2 11.3 11.3 10.7 12.3 12.1 10.7 8.8 3rd level of PE 14.7 11.8 11.4 10.4 12.5 11.6 9.6 9.3 Completion rates 1st level of PE 29.5 24.3 32.7 28.4 34.4 27.7 37.4 29.0 2nd level of PE 13.4 11.6 16.6 13.6 16.5 13.3 18.3 15.7 3rd level of PE 8.2 7.2 9.2 8.6 9.3 8.5 10.9 9.1 Source: MEC statistics. 116. The analysis of both drop-out and repetition rates provides sufficient information to say that girls are usually more efficient than boys in primary school since they repeat and drop-out less, presenting therefore a higher retention rate than boys. Despite that, the proportion of girls who actually complete the 4th, 6th and 8th grades compared to the official population with graduation age is lower than boys. It seems that even though girls repeat less and drop-out less, more boys than girls graduate as a proportion of the population at the official graduation age. This can only be explained by the fact that girls enrol less than boys in school. If from grade 1 the total enrolment of girls as a proportion of the official female - 49 - school-age population is lower than that of boys even though girls repeat and drop out less (and the differences in the repetition and drop out rates per gender are not as great as the difference in the GER's), then it is understandable that girls have lower CRs than boys. 117. Regarding the evolution of the completion rates, once again (like in the case of the drop-out rates), boys have shown a more positive trend from 1999 to 2002 than girls. 118. As can be seen in Table 2.13 below there is a considerable level of inequity within and between provinces. Examples within provinces are: Kuanza Norte presents a GER for males in the 1st level of PE in 2002 of 127.1 percent but only 64 percent for females; Moxico presents a GER for males in the 2nd level of PE in 2002 of 21.3 percent but only 7.5 percent for females. On the contrary, provinces like Luanda, Benguela or Lunda Sul are much less unequal in terms of gender. Examples between provinces are also very illustrative: whereas Uíge had a GER for males in the 1st level of PE of 59.7 percent, the same GER in Namibe was 178.6 percent. Table 2. 13: Gross enrolment rates in 2002 per province and gender in the three levels of primary education per province. GER 1 st level PE 2nd level PE 3rd level PE M F Total M F Total M F Total National 96.0 30.8 16.1 111.7 80.4 33.1 28.4 17.2 15.0 Bengo 128.9 174.8 151.8 48.0 35.2 41.6 11.2 9.3 10.2 Benguela 76.4 73.1 74.7 26.9 24.5 25.7 10.3 9.2 9.7 Bié 81.0 74.5 77.8 24.1 22.7 23.4 7.3 13.7 10.5 Cabinda 114.2 109.1 111.6 54.2 41.7 47.9 23.7 22.6 23.2 Huambo 297.6 87.7 192.7 29.0 22.9 26.0 12.0 8.6 10.3 Huíla 166.9 142.5 154.7 42.5 32.7 37.6 12.6 10.7 11.7 Cunene 98.2 101.9 100.1 12.4 14.9 13.7 4.0 4.4 4.2 Kuando Kubango 88.5 36.1 62.3 13.7 7.5 10.6 7.0 2.8 4.9 Kuanza Norte 127.1 64.0 95.6 1.2 39.0 20.1 7.7 3.5 5.5 Kuanza Sul 97.3 67.7 82.5 24.5 12.0 18.2 9.6 5.8 7.7 Luanda 75.2 73.9 74.5 51.8 49.9 50.9 38.2 35.2 36.6 Lunda Norte 69.2 36.8 53.0 13.0 6.3 9.7 8.7 5.3 6.9 Lunda Sul 91.8 90.4 91.1 18.4 13.1 15.7 5.7 3.7 4.7 Malange 175.8 98.5 137.2 20.3 12.6 16.5 9.1 3.8 6.4 Moxico 108.1 47.7 77.9 21.3 7.5 14.4 7.7 2.6 5.1 Namibe 178.6 165.0 171.8 64.3 51.3 57.8 29.1 20.9 25.0 Uíge 59.7 42.8 51.2 17.0 7.2 12.1 4.6 2.5 3.5 Zaire 94.4 68.2 81.3 34.0 15.8 24.9 12.2 6.9 9.5 119. In the case of the primary education completions rates, once more there are significant inequalities within and between provinces. Whereas Luanda usually presents the highest completion rates in the country, both for males and females in all the three levels of primary education, Lunda Norte presents the lowest completion rates, with rates more than 20 percentage points lower in all the cases. Not only Lunda Norte, for example, is in a much worse situation in terms of completion rates for males and females when compared with other provinces, it - 50 - also faces gender inequities inside the province. The opposite happened, for example, in Cunene (Table 2.14). Table 2. 14: Completion rates in 2002 per province and gender in the three levels of primary education per province. Completion rates 1st level PE 2nd level PE 3rd level PE M F Total M F Total M F Total National 37.4 29.0 33.2 18.3 15.7 16.9 10.9 9.1 10.0 Bengo 0 0 0 0 0 0 Benguela 28.8 23.9 26.4 11.5 10.4 10.9 4.2 3.9 4.1 Bié 20.5 21.4 20.9 15.4 16.4 15.9 5.6 6.6 6.1 Cabinda 46.4 36.9 41.6 25.2 17.6 21.4 11.4 10.4 10.9 Huambo 27.7 22.4 25.1 12.7 10.3 11.5 6.6 5.3 6.0 Huíla 56.9 44.1 50.5 13.6 10.1 11.8 4.7 3.2 4.0 Cunene 18.7 17.6 18.1 3.8 5.2 4.5 1.4 1.2 1.3 Kuando Kubango 20.8 9.6 15.2 7.0 3.4 5.2 3.5 1.5 2.5 Kuanza Norte 48.8 9.9 29.4 8.4 3.2 5.8 2.6 0.8 1.7 Kuanza Sul 30.1 13.0 21.5 8.7 4.5 6.6 5.9 3.6 4.7 Luanda 55.2 51.5 53.4 38.3 36.3 37.3 28.4 25.0 26.7 Lunda Norte 19.3 7.8 13.5 4.2 2.1 3.1 1.3 1.8 1.6 Lunda Sul 21.1 13.4 17.3 5.7 3.1 4.4 2.4 1.6 2.0 Malange 22.8 9.0 15.9 6.6 5.6 6.1 4.4 2.1 3.2 Moxico 23.9 8.7 16.3 9.2 2.9 6.1 3.5 0.8 2.1 Namibe 57.4 47.6 52.5 32.0 23.3 27.6 15.0 10.1 12.5 Uíge 20.2 7.3 13.7 7.5 2.9 5.2 2.6 1.0 1.8 Zaire 47.6 20.8 34.2 17.1 8.5 12.8 6.5 2.4 4.4 C. RECENT EXPENDITURE TRENDS 120. As can be seen in Table 2.15, from 1999 to 2003 using nominal or current prices, GDP in Angola increased in absolute terms around 125 percent, total public spending around 25 percent and education spending by more than 200 percent. Between 1999 and 2003, GDP in current prices increased from US$ 6,093,200,000 to US$ 13,825,700,000 and education expenditure increased from US$ 152,900,000 to US$ 472,600,000. The government has also demonstrated that it is committed to improving education opportunities by increasing the share of education spending in total state expenditure from 3 to 7.5 percent between 1999 and 2003 and the share of education expenditure as a proportion of nominal GDP from 2.5 to 3.4 percent in the same period. 121. From 2004 to 2005 nominal GDP also increased by more than 100 percent compared to 2003. However, by 2004 education spending (in million USD) was still below the levels of 2003 and by 2005 it was still below the levels of 2002. This is not to say that in absolute terms the education spending in the local currency did not increase. It did in fact, but since the Angolan Kwanzas exchange rate appreciated (devaluation of the currency), compared to previous years, the increase does not become visible. In any case since nominal GDP is also expressed in USD, the comparison is still relevant. Total education spending in 2004 and 2005 expressed as a percentage of GDP was below the levels of 1999. In summary, if from 1999 to 2003, the Government showed clearly a commitment to improve the education levels of the population, that same commitment was not sustained in the following two years. - 51 - Table 2. 15: GDP, Total Public Spending, Education Spending in million USD and Education Spending as a proportion of Total Public Spending and as a proportion of GDP. Unit: 000'000 USD 1999 2000 2001 2002 2003 2004 2005 Nominal GDP 6093.2 8631.1 9473.9 11203.7 13825.7 19600.0 28000.0 Real GDP 1109.6 2115.5 7401.5 10877.3 Total Public Spending 5017.9 5242.7 4385.5 5370.7 6266.8 Total Discretionary Public Spending 4461.9 4746.7 3940.5 5013.7 5940.8 Education Spending 152.9 222.8 292.3 641.3 472.6 457.9 637.4 Proportions % education in GDP 2.5 2.6 3.1 5.7 3.4 2.3 2.3 % education in Total Public Spending 3.0 4.2 6.7 11.9 7.5 % education in Total Public Dis. Spe. 3.4 4.7 7.4 12.8 8.0 % Total Public Spending in GDP 82.4 60.7 46.3 47.9 45.3 Source: 1999 to 2003: IMF Country Report No. 03/292 and MINFIN webpage; 2004-2005: World Bank Country Data Profile, Penn World Tables and MINFIN webpage. 122. In addition, Figure 2.8 illustrates that total government expenditure, as a share of GDP, decreased from 1999 to 2003 and government education expenditure, as a percent share of GDP, increased in the same period. Total government expenditure as a percentage share of GDP decreased from 82.4 percent to 45.3 percent, whereas Government education expenditure as a share of GDP increased from 2.5 percent to 3.4 percent. Figure 2. 7: Government overall and education expenditure as a share of GDP. 90,0 82,4 80,0 70,0 60,7 60,0 46,3 50,0 47,9 45,3 % 40,0 30,0 20,0 2,6 10,0 2,5 3,1 5,7 3,4 0,0 1999 2000 2001 2002 2003 Years Gov. Total Exp. as share of GDP Gov. Educ. Exp. as share of GDP 123. Education expenditure on a per capita basis increased by more than 100 percent from 1999 to 2003 (from US$ 11.8 to US$ 32.2). However, Angola is still very far away from the levels of education spending from its neighboring countries and Sub-Saharan Africa in general. On average, Sub-Saharan African (SSA) countries spent 5.1 percent of GDP on education, whereas Less Developed Countries spent the equivalent to 3.9 percent of GDP in 1997 (World Education Report, 2000). This is considerably higher than the 2.5 percent that was spent on education in Angola as a proportion of GDP in 1999 and even the 3.4 percent spent in 2003. If the levels of investment in education of 2004 and 2005 are taken into consideration, the gap relative to neighboring countries even widened. - 52 - Trends in government and externally financed education expenditure 124. The analysis of expenditure data on education was negatively affected by the lack of consistent classification of spending categories. For example, expenditure data for 1999 to 2003 were classified differently than those for the period 2004 and 2005. The different classification methods impaired the comparison of these two periods which are therefore presented in different tables. 125. The existing data show that there was a general and substantive decrease in the execution rate for all functional expenditures from 2002 to 2003, indicating either that the Government budgeted the amounts but did not disbursed them accordingly (e.g. liquidity problems) or that if disbursed, the institutions did not have the capacity or time to use those resources (Table 2.16). The second point to emphasize is the fact that both the expenditures of secondary, other expenditures, subsidiary services and other services in 2002 were higher than the correspondent budgets (Tables 2.17 and 2.18). Table 2. 16: Budgets and actual expenditure (Million USD) from 1999 to 2003. 1999 2000 2001 2002 2003 2004 2005 Budgets 192.6 312.5 355.1 658.6 1138.1 842.4 958.8 Actual 152.9 222.8 292.3 641.3 472.6 457.9 637.4 expenditure Execution 79% 71% 82% 97% 42% 54% 66% rate Source: MINFIN database. Table 2. 17: Execution rate of budgets by functional classification from 999 to 2003 (in %). Services of education 1999 2000 2001 2002 2003 Pre-primary and primary 73.2 74.6 84.8 95.1 40.0 Secondary 56.3 52.6 67.0 147.0 58.3 Secondary ­ teacher training 46.9 66.0 68.7 76.9 38.9 Tertiary 63.9 67.0 87.9 61.4 36.6 Other expenditures 73.1 45.0 54.2 109.9 52.7 Subsidiary services 95.9 90.8 98.6 175.0 5.6 Other education 84.8 69.5 97.0 83.1 26.0 Other culture 30.1 43.0 86.2 81.0 3.0 Other 82.3 85.5 69.1 137.8 15.4 Source: MINFIN database. Table 2. 18: Execution rate of budgets by functional classification from 2004 to 2005 (in %). Services of education 2004 2005 Primary 55.6 76.6 Secondary 34.4 42.4 Technical-professional training 54.4 23.2 Higher Education 69.2 81.1 Adult education 4.3 99.7 Other services of education 54.0 76.1 Source: MINFIN database. 126. Whereas the budget for secondary education in 2002 was of 87.8 million USD the expenditures reached 129.1 million USD, almost 50 percent higher. This should also be considered as a budgetary or expenditure problem. The main reasons explaining these differences are the insufficient budget for expenditures - 53 - on salaries for several institutions (e.g. Instituto Médio de Saúde of Bengo or Instituto Normal de Educação of Bié) and the level of expenditures of many others ­ more than or almost the double of the budget attributed (e.g. Instituto Médio Normal de Educação Cmdt. Kwenha of Benguela or Instituto Médio Agrário of Huambo). However, the most significant point regarding secondary education in 2002 is the expenditure on the program for reintegration of young people in difficult situations in Cabinda, of more than 60 million USD which was not budgeted for. 127. Regarding the category of other expenditures, the over-spending is related to the 1.1 million USD spent by the Provincial Directorate of Education and Culture from Uíge in teacher training which was not budgeted for. The same applies to subsidiary services in 2002, when around 5 million USD were spent by the Government of Luanda Province on school feeding which were not budgeted for. Finally, in relation to the classification of other services many institutions not linked to the education sector presented expenditure in the education function without being budgeted for. 128. Regarding the period 2004-2005, the main point to emphasize is the increase in the execution rate in all services or functions of education, except for technical and professional training, where the rate decreased by more than the double. 129. Table 2.19 below illustrates in more detail the execution rate of budgets in primary education and secondary education by a more detailed economic classification. According to this table it is interesting to note that, in general, the execution rates for the recurrent budgets have been increasing. In addition, the capital expenditure in primary and secondary education followed very irregular patterns. 130. The very high level of execution in secondary education for capital expenditure in 2002 is again related to the Cabinda Program for the reintegration of young people in difficult situations and the execution rate regarding transfers and subsidies in primary education in year 2002 is related to the fact that the 100,000 USD spent were almost not entirely budgeted for. The high execution rate in secondary education in 2003 in transfers and subsidies is related to a level of expenditure through `Encargos Financeiros do Estado' in Public Institutes and Autonomous Services twice as large as the as the amount that had been budgeted. - 54 - Table 2. 19: Execution rate of budgets in primary and secondary education by functional and economic classification from 1999 to 2003 (in %). Services of education 1999 2000 2001 2002 2003 Primary education Recurrent 73.3 74.4 95.1 95.9 40.5 Salaries and remunerations 78.4 79.4 96.4 96.6 41.1 Goods and Services 45.1 31.7 68.7 67.3 32.6 Transfers and subsidies 71.1 26.2 73.7 390.0 55.5 Capital Investments 63.1 78.5 29.8 80.4 35.2 Secondary education Recurrent 50.1 55.0 74.7 77.3 61.9 Salaries and remunerations 47.3 65.6 75.3 84.9 35.8 Goods and Services 51.7 17.3 70.0 65.6 43.2 Transfers and subsidies 57.6 74.3 101.7 96.6 199.0 Capital Investments 87.1 51.6 54.2 468.3 21.4 Source: MINFIN database. 131. In summary, there seems to exist serious problems in the implementation of the education sector plans insofar as the planning and use of resources is concerned. There are problems in preparing budgets which are reflected in the fact that in some years some institutions do not seem to have budgets even for payment of salaries; there might exist problems in terms of underestimation of resource needs on the side of the cost centers and irrational cuts by central Ministry of Finance and Planning which could explain the big differences between low budgets and high levels of expenditure; and liquidity problems which could explain scenarios like the general very low levels of expenditure compared to the budgets in 2003. Education expenditure by functional and economic classification 132. Primary education has received most of the resources invested in the education sector from 2000 to 2003 and its proportion has been increasing overall (Table 2.20). This is consistent with the Government objectives of achieving primary education for all by 2015. However, it does not mean that the level of expenditure from 2003 will allow that to happen, considering the wastage ratio of the education system in Angola. The second sub-system of education with higher expenditure levels from 1999 to 2001 was higher education and from 2002 to 2003 secondary education (grades 9 to 12). Table 2. 20: Total expenditure by functional classification from 1999 to 2003 (USD mi). Services of education 1999 2000 2001 2002 2003 Total 152.9 222.8 292.3 641.3 472.6 Pre-primary and primary 27.4 92.4 160.5 405.7 311.2 Secondary 3.2 12.3 32.2 131.1 98.6 Secondary ­ teacher training 0.7 3.2 7.9 20.0 18.3 Tertiary 16.0 25.3 32.1 44.8 27.5 Other not by levels 1.4 3.4 10.8 11.5 8.2 Subsidiary services 49.6 30.2 33.9 11.8 1.3 Other education 38.7 37.1 11.2 14.4 7.0 Other culture 1.5 3.8 0.4 0.0 0.0 Other 14.5 15.0 3.4 1.9 0.5 Source: MINFIN database. - 55 - Table 2. 21: Total expenditure by functional classification from 2004 to 2005 (in million USD) and proportion in total expenditure by functional classification. Services of education 2004 % total 2005 % total Total 457.9 637.4 Primary 19.9 4.3 36.7 5.8 Secondary 10.6 2.3 16.7 2.6 Technical-professional training 17.8 3.9 23.8 3.7 Higher Education 37.7 8.2 85.0 13.3 Adult education 0.003 0.0 0.2 0.0 Other services of education 371.7 81.2 474.8 74.5 Source: MINFIN database. 133. It is also important to note that in 1999, the US$ 49,600,000 spent on subsidiary services were all invested in scholarships, representing more than 32 percent of the total education expenditure. In 2000, the US$ 30,200,000 spent on subsidiary services were also all for scholarships, representing 13.6 percent of the total education expenditure; in 2001 the investment in scholarships was US$ 29,300,000 or 10 percent of total expenditure. In 2002 and 2003, the expenditures of the education function do not include any investment in scholarships, which seems to be very inconsistent. It could be assumed that there was the same level of expenditure from 1999 to 2001 but if so, those expenditures are included in a different function. In 2002, the US$ 11,800,000 spent on subsidiary services of education went for school feeding programs. While in 1999 and 2000 no investment was done in school feeding programs, in 2001 there was an investment of around US$ 4,500,000, in 2002 more than the double of 2001 and in 2003 only around US$ 600,000 were invested for the same objective. 134. Regarding Table 2.21, we are reluctant to make any substantive comment or establish comparisons, since it is evident that the classification of the expenditure was not properly done. It would be important to have the expenditure included in `other services of education' classified according to the classification used for the period 1999-2003. The only sub-system where some comparison can be established is higher education (assuming that the expenditure reported was properly classified). In this sub-system total expenditure had already increased from 2003 to 2004 but the increase from 2004 to 2005 but details were not available to understand the reasons behind this increase. 135. Tables 2.22 and 2.23 below show that most of the resources in all sub- systems are used for recurrent costs. In primary education, teacher training and higher education recurrent costs always represented more than 90 percent of total expenditure in that sub-system of education (except for higher education in 1999). The main exceptions to this rule were secondary education in 2002 given the investment of US$ 64,700,000 in the province of Cabinda for the reintegration of youngsters in difficult situation considered in the State Budget accounts as capital costs; and other services of culture in 2001 and 2002 because of the rehabilitation of the National Library (US$ 375,000 in 2001 and US$ 32,000 in 2002). - 56 - Table 2. 22: Total recurrent expenditure by functional classification from 1999 to 2003. Services of education 1999 2000 2001 2002 2003 Total 137.7 210.6 272.3 544.7 435.8 Pre-primary and primary 27.1 88.6 151.6 388.5 283.6 Secondary 2.7 10.5 24.8 55.3 91.8 Secondary ­ teacher training 0.7 3.2 7.9 20.0 18.0 Tertiary 13.0 23.1 29.0 41.9 25.5 Other not by levels 1.4 3.4 10.8 11.2 8.2 Subsidiary services 49.6 30.2 33.9 11.8 1.3 Other education 30.4 35.1 11.1 14.2 6.9 Other culture 1.5 3.7 0.0 0.0 0.0 Other 11.2 12.7 3.3 1.8 0.5 Source: MINFIN database. Table 2. 23: Total recurrent expenditure by functional classification from 2004 to 2005 and proportion of recurrent expenditure in total expenditure by functional classification Services of education 2004 % total 2005 % total Total 417.1 555.2 Primary 4.0 20.1 6.0 16.3 Secondary 1.6 15.1 2.9 17.4 Technical-professional training 11.7 65.7 10.9 45.8 Higher Education 36.1 95.8 61.3 72.1 Adult education 0.003 100.0 0.2 100.0 Other services of education 363.7 97.8 473.9 99.8 Source: MINFIN database. 136. Regarding the information included in Table 2.23, apparently most recurrent expenditures of primary education and secondary education were classified under other services of education. Otherwise, it would be very difficult to justify the very low proportions of primary recurrent expenditure in total recurrent expenditure from 2003 to 2004-2005. The extent of the misspecification, however, cannot be established. Furthermore, one can assume that given the level of destruction of the physical infrastructure during the civil war, the Government increased the capital expenditure in all sub-systems of education (from internal and external sources) which could imply a lower share of recurrent expenditure in total expenditure in 2004-2005 as compared to 1999-2003. 137. Total expenditures, in absolute terms, in salaries and remunerations and its share in total education expenditure has been growing from 1999 to 2003 (see Table 2.24). In terms of goods and services the trend has been of considerable growth in expenditure, on absolute terms, in particular in the sub-system of secondary education. In 2001 and 2002, the Angolan education system spent more than twice the resources in goods and services in secondary education than in primary education7. This is mainly explained by the sharp increase in expenditures in goods and services from various Secondary Institutes from 2000 to 2001. Anecdotal evidence of this is that the Secondary Institutes and Pre-University Colleges of Luanda (10) increased their expenditure in goods and services from around US$ 380,000 in 2000 to around US$ 5,000,000 in 2001. 7The primary education sub-system, with the highest number of schools and pupils had less than half of expenditure level for goods and services in relation to secondary education in 2001 and 2002 (1,7 and 1,0 percent of total expenditure) and only in 2003 the level of expenditures between the two sub-systems were equalized. - 57 - Table 2. 24: Expenditures by detailed economic classification from 1999 to 2003 (in million USD). Services of education 1999 2000 2001 2002 2003 Total expenditure 152.9 222.8 292.3 641.3 472.6 Recurrent costs 137.7 210.6 272.3 544.7 435.8 Salaries and remunerations 44.4 109.9 169.7 472.1 315.7 Goods and Services 37.2 41.7 41.4 56.5 59.1 Transfers and subsidies 55.5 59.0 61.2 16.1 61.0 Capital costs 15.2 12.2 20.0 96.6 36.8 Investments 15.2 12.2 20.0 96.6 36.8 Source: MINFIN database. 138. Regarding the expenditure in recurrent transfers and subsidies it is important to note the considerable amount of expenditure in 2003 in secondary education services, attributed to what is called Public Institutes and Autonomous Services, which represented 11.1 percent of total education expenditure in 2003; the expenditure in tertiary services in 2000 and 2001 which is explained by the fact that the recurrent expenditure of most faculties of the Universidade Agostinho Neto, including salaries and remunerations were classified as recurrent transfers and subsidies. However, the expenditure with goods and services was below 10 percent of total education expenditure in 2002 (see Figure 2.9), when there was the most significant increase in total expenditure in relation to the previous years. In addition, it was only in 2002 that capital expenditure as a proportion of total expenditure represented more than 10 percent. 139. Furthermore, the proportion of salaries and remunerations in total education expenditure grew substantially from 29 percent in 1999 to 66.8 percent in 2003, whereas the expenditure on goods and services decreased from 24.3 percent in 1999 to 12.5 percent in 2003 and also decreased on transfers and subsidies from 36.3 percent in 1999 to 12.9 percent in 2003. Figure 2. 8: Proportion of expenditure by economic classification in total education expenditure from 1999 to 2003. 80,0 70,0 60,0 50,0 Salaries and remunerations 40,0 Goods and Services 30,0 Transfers and subsidies 20,0 Investments 10,0 0,0 1999 2000 2001 2002 2003 Classification of expenditures per provinces 140. From 1999 onwards Provincial Governments started to be responsible for all recurrent expenditure in primary education from grades 1 to 6 and capital - 58 - expenditure projects below US$ 1,000,000. It is therefore, understandable that the share of the Ministry of Education in total education expenditure decreased from 1999 to 2005 (see Table 2.25). The expenditure of the Ministry of Education and Culture and those of the provinces of Luanda and Benguela when taken together represented more than 50 percent of total expenditures in education in each year. The other two provinces with higher shares in total education expenditure in the period were Huíla and Huambo. There was also an increasing trend in the expenditure in the provinces of Bié, Cabinda, Kuanza Sul and Uíge. 141. The criteria for allocating resources to the provinces are not based on the number of pupils that those resources need to benefit. In 1999, even though Huíla had 15.7 percent of the total of number of pupils enrolled in the country from grade 1 to 8 it only spent 8.72 percent of total primary education expenditure (see Table 2.26). The same applies to Kuanza Sul and Uíge. On the other hand, provinces like Malange, Zaire and Moxico, present a proportion in terms of primary education expenditure which is higher than their share in the total number of primary pupils enrolled in the country. 142. In 2003, the disparities seem to have widened. Provinces like Benguela, Luanda, Moxico, Kuanza Norte, Namibe and Zaire present a higher share in total primary education expenditure than their respective share in total primary education enrolment, whereas provinces like Bié, Huambo, Kuanza Sul and Uíge show the reversed trend. The difference in terms of percentage points can be as significant as 8 percentage points in the case of Luanda with a higher proportion of expenditures and Huíla with higher proportion in the number of primary education pupils. In terms of resources, 8 percentage points in total primary education expenditure in 2003 represented US$ 24,900,000. Table 2. 25: Share of total education expenditure per province from 1999 to 2005. 1999 2000 2001 2002 2003 2004 2005 MEC 78.1 53.1 35.9 14.7 21.8 16.4 20.6 Bengo 0.2 0.4 0.7 0.8 1.2 1.1 1.5 Benguela 2.4 6.7 10.2 11.2 10.1 10.7 13.1 Bié 1.0 1.8 2.6 2.9 3.6 3.0 4.7 Cabinda 1.5 2.4 1.7 13.6 3.4 3.2 4.2 Huambo 1.2 3.4 4.2 5.3 6.8 5.2 7.1 Huíla 2.2 5.9 6.9 8.1 9.1 8.8 8.6 Cunene 0.3 0.8 1.0 1.2 1.6 1.6 1.8 Kuando Kubango 0.3 0.5 1.0 1.4 1.4 0.3 0.2 Kuanza Norte 0.4 1.1 1.5 1.8 2.2 2.1 2.0 Kuanza Sul 0.7 1.6 3.3 4.1 3.5 4.6 4.8 Luanda 8.3 15.2 22.5 22.7 23.4 29.6 19.7 Lunda Norte 0.2 0.6 0.6 1.0 1.0 0.8 1.4 Lunda Sul 0.3 0.7 0.8 1.2 1.1 1.8 0.9 Malange 0.7 1.1 1.7 2.3 2.3 2.2 2.4 Moxico 0.5 1.2 1.5 1.5 1.9 2.0 0.6 Namibe 0.4 1.0 1.1 1.2 1.8 1.2 1.6 Uíge 0.9 1.8 1.7 3.8 2.6 4.1 4.6 Zaire 0.4 0.8 1.0 1.2 1.4 1.2 0.2 Source: MINFIN database. - 59 - Table 2. 26: Total and Per Pupil Primary Education Expenditure per Province in 1999, 2001 and 2003. 1999 2001 2003 Exp. Pupils Exp. Pupils Exp. Pupils Bengo 0.82 2.3 1.06 1.8 1.67 1.4 Benguela 11.08 12.1 16.32 11.5 12.78 8.6 Bié 3.13 4.8 4.44 3.9 4.68 8.1 Cabinda 5.53 4.5 2.87 6.6 3.74 2.9 Huambo 5.31 4.1 6.70 7.2 8.51 10.3 Huíla 8.72 15.7 10.20 13.4 11.13 19.9 Cunene 1.65 1.7 1.48 1.7 2.02 3.2 Kuando Kubango 1.38 1.0 1.81 1.5 1.94 2.1 Kuanza Norte 1.89 2.8 2.62 2.6 2.82 1.9 Kuanza Sul 3.30 7.7 4.79 6.2 4.48 6.8 Luanda 40.10 32.3 33.50 29.8 26.74 18.7 Lunda Norte 0.60 1.8 0.99 1.5 1.22 2.1 Lunda Sul 1.49 0.9 1.33 0.8 1.37 1.4 Malange 3.72 1.3 2.97 2.5 3.30 3.8 Moxico 2.48 1.7 2.57 1.8 2.61 1.6 Namibe 1.73 0.2 1.73 2.1 2.32 1.5 Uíge 1.73 3.6 1.73 4.2 2.32 4.6 Zaire 4.75 1.3 2.85 1.2 3.35 1.1 Source: MINFIN database and MEC statistics. 143. Per capita spending in the education sector per province from 1999 to 2003 in grades 1 to 8 increased significantly from US$ 22.6 to US$117 bringing Angola closer to the regional averages.8 For example, in 1997 the average Sub-Saharan per capita recurrent primary school expenditure was US$143, which is just US$ 26 higher than the Angola per pupil recurrent and capital expenditure for grades 1 to 8 (Table 2.28). Considering grades 7 and 8 as secondary education and comparing it with regional secondary education unit recurrent expenditure, the difference is much higher. Again, in SSA the average per capita recurrent secondary school expenditure was as much as US$ 378 in 1997. 8It should be noted that this information is only based on government expenditure, not external recurrent expenditure since there was almost no data available regarding that source of funding. - 60 - Table 2. 27: Education unit expenditure, grade 1-8, per province from 1999 to 2003. 1999 2000 2001 2002 2003 National 22.6 66.8 101.9 219.4 117.1 Bengo 8.2 32.7 67.0 121.2 136.7 Benguela 20.6 78.1 127.1 314.9 173.3 Bié 14.7 68.1 91.3 174.5 67.6 Cabinda 27.4 42.5 80.6 275.6 151.0 Huambo 29.0 57.1 82.7 103.7 96.5 Huíla 12.6 54.7 77.7 173.4 65.5 Cunene 22.0 65.1 62.7 173.5 74.9 Kuando Kubango 31.3 52.4 123.1 311.9 110.2 Kuanza Norte 15.3 65.8 103.9 275.6 176.5 Kuanza Sul 9.6 36.5 83.1 225.4 77.2 Luanda 28.0 73.0 126.5 252.5 167.7 Lunda Norte 7.7 52.3 58.1 139.6 68.9 Lunda Sul 37.8 98.8 94.4 267.6 116.1 Malange 63.5 66.2 107.4 219.0 100.4 Moxico 32.4 98.8 107.7 225.3 195.7 Namibe 182.6 61.3 80.7 186.7 176.0 Uíge 29.4 62.4 76.0 350.6 85.9 Zaire 30.9 98.7 124.6 317.7 187.6 Source: MINFIN database and MEC statistics. 144. In addition, the differences in total education per capita expenditure between provinces show the imbalances and lack of criteria in resource allocations as well as the inefficiency of the system. Moreover, the significant decrease in per capita expenditure from 2002 to 2003 is certainly explained by the growth in enrolment (higher PTRs discussed in the following section confirm that). D. INSTITUTIONAL PLANNING AND BUDGETING 145. In what concerns budgeting procedures there are two levels of consolidation and negotiation with the Ministries of Plan and Finance. On the one hand, the Ministry of Education consolidates and negotiates the budgets for the areas under its responsibility as seen above. The exception to this rule is Universidade Agostinho Neto, which as an Autonomous Institution negotiates directly its own budget. On the other hand, the Provincial Governments themselves negotiate separately with the Ministries of Planning and Finance the budgets of the Provincial Governments, including the resources for the education sector. 146. Internally in MEC, the recurrent budget is consolidated by the General Secretary through the Department of Budget Management and Administration (Departamento de Administração e Gestão do Orçamento ­ DAGO) and the capital budget by GEPE9. At provincial level the education budget preparation is led by the Provincial Directorates of Education and in some cases by some schools considered Management Units which present their recurrent budget proposals directly to the Provincial Government. 147. The payment of salaries is made in accordance with the payroll lists prepared by the respective structures and submitted directly to the National 9 DAGO and GEPE consolidate the proposals coming from their own departments but also from the National Directorates. After that, the Executive Board (Conselho de Direcção) of MEC is responsible for analysing and approving the budget proposal to be submitted to MINFIN. - 61 - Directorate of Accounting (Direcção Nacional de Contabilidade - DNC), in the case of the central structures and to the Provincial Directorates of Finance (Delegações Provinciais de Finanças ­ DPF) in the case of the provincial structures for control purposes which then compare the payroll lists submitted with the files from MINFIN. Salaries start to be paid on the 15th of each month. For goods and salaries, the Budget Units (Ministries and Provincial Governments) submit quarterly requests to MINFIN, including for the institutions under their control or jurisdiction. 148. The execution of the capital budget follows the same procedures. The only difference is the fact that the capital budget is presented and discussed with the Ministry of Planning which is responsible for the Public Investment Program (PIP). Weaknesses in the budget process 149. There is not enough coordination and consultation in the preparation of the budget. Bearing in mind the information presented in the previous section, the main weakness identified in terms of the budget preparation and execution relates to the absence of close collaboration and coordination between the Ministry of Education and Culture and the Provincial Directorates of Education, a problem that most likely other sectors also face. MEC has information about the budget proposals and budget execution reports from the Provincial Directorates of Education too late in the process in order to be able to provide any useful inputs. 150. In addition, there are no mechanisms established at the moment that make it a condition for central and provincial structures to discuss and harmonize points of view to be included in the budget proposals. This is particularly important because Provincial Governments manage the resources for the sub-system of education that should have the highest impact in the sector. 151. Another concern is that the current classification of expenditures does not allow the analysis of expenditure by specific levels of education (grades 1 to 4, 5 to 6 and 7 to 8 separately and from pre-primary services; and teacher training separately from technical and secondary education). In addition, the current budgeting and execution procedures make it very difficult to have separate figures for the payment of salaries for teachers and auxiliary staff. 152. Finally, there are substantive weaknesses in the process of classification of public expenditures in education. The budgeting and execution system of public resources seems to suffer from inadequate institutional capacity because even those expenditures that should be easy to classify are not done according to the rules and even some permanent expenditures of the education sector are included in some years but not in others. For the benefit of the performance of the sector, the authorities should look carefully into this problem. E. THE VALUE FOR MONEY CHAIN 153. The broad framework applied in this section is that of value-for-money, which looks at the efficiency and effectiveness of the Angolan education system. In the context of education, it is appropriate to consider the number of students - 62 - enrolled as the intermediate output of the system. The final outcome of the system is related to the objectives set down in the policy documents. Frequently, the final outcomes of an education system are related to the number of students who complete the full course of education or who are successful at final exams (providing that pass rates reflect mastery of the curriculum content and have not been artificially manipulated, for example, to ration places in the next cycle). Using these definitions, efficiency is measured by the ratio of expenditures, teachers, classrooms and instructional materials to students enrolled. Effectiveness is measured using completion rates, cycle costs (cost per graduate) and success rates. More details about these concepts are given in annex 3. 154. One of the reasons why this framework provides a useful way of analyzing education systems is that it clearly reveals trade-offs between access and quality. For example, the pupil teacher ratio (PTR) is one of the most useful gross measures of efficiency. Generally the higher this ratio, the more efficient the sub- sector and the more children can be enrolled for the same amount of resources. However this cannot be extended indefinitely, as beyond a certain level there are effectiveness costs. At very high pupil teacher ratios, the quality of teaching declines because the teacher spends more time on class management than teaching, and the opportunity for personal interaction lessens. Teachers 155. The pupil teacher ratio is an important element in terms of education costs. Table 2.28 below shows the PTR's in the 1st, 2nd and 3rd levels of primary education in 1999, 2000, 2001 and 2003. As can be seen, PTR's are a little bit higher in the 1st level of PE compared to the other levels of education. In addition, they were increasing in all three levels of primary education, except from 2001 to 2003 in the 3rd level of PE. However, the number of pupils per class was higher than the PTR in both the 2nd and 3rd levels of PE in 2003. The higher pupil per class ratio than the pupil per teacher ratio is explained by the fact that there was more than one teacher per class (the teacher class ration was 1.4 for grades 5-6 and 2.3 for grades 7-8), which usually means that there are no shifts and an unnecessary wastage of resources where no such degree of specialization is generally required. 156. It is in principle recommended that the PTRs, or more importantly, the pupil class ratio, should not be higher than 40-45 to enssure sufficient pupil teacher contact. Thus, the pupil per class ratio in Angola seems to be appropriate in all levels of primary education. - 63 - Table 2. 28: Pupils per teacher, teacher's per class and pupilsper class ratiosin the 1st, 2nd and 3rd levels of PE in 1999 to 2001 and 2003, where available. 1999 2000 2001 2003 1st level PE Grade 1-4 Pupils per teacher 22 26 29 44 Teacher per class 1.0 Pupils per class 44 Pupils per classroom 74 81 89 94 2nd level PE Grade 5-6 Pupils per teacher 19 20 26 28 Teacher per class 1.4 Pupils per class 44 Pupils per classroom 82 82 89 100 3rd level PE Grade 7-8 Pupils per teacher 15 15 20 17 Teacher per class 2.3 Pupils per class 41 Pupils per classroom 73 76 129 Source: MEC statistics. 157. An international comparison also shows that. On average in 2003, low income countries had primary pupil-teacher ratio of 43:1, middle income countries 21:1, high income countries 14:1 and Sub-Saharan Africa 45:1. Figure 2.10 below presents examples of specific countries as well. Figure 2. 9: Primary Pupil-Teacher Ratio in 2002 for groups of countries and selected Sub-Saharan countries. Primary Pupil-Teacher Ratio 2002 67 70 49 60 43 45 43 50 29 40 27 % 21 22 30 14 20 10 0 High income Middle Low income SSA Angola Botswana Gabon Mozambique Namibia Zambia countries income countries countries Source: World Development Indicators, 2005. 158. While it is true that the level of specialization is usually higher at grades 5- 6 and 7-8 compared to grades 1-4 and those levels of education probably require more than 1 teacher per class, it is questionable whether there is a need for 1.4 teachers per class in the 2nd level of PE and 2.3 in the 3rd level of PE. At this level, where the pupil per class and, especially the PTR is quite low, teacher efficiency could be improved by increasing student intakes. 159. In addition, it should also be noted that the pupils per classroom ratio has been increasing significantly in all three levels of primary education from 1999 to 2003 and in particular the 3rd level of PE seems to be suffering from a strong pressure in terms of new infrastructures. The pupil per classroom ratio increased - 64 - from 74:1 to 94:1 in the 1st level of PE and it increased from 82:1 to 100:1 in the 2nd level of PE from 1999 to 2003 but it increased from 73:1 in 2000 to 129:1 in 2003. 160. There are strong variations in PTRs among the 18 provinces and those variations increased from 2001 to 2003. In the 1st level of PE while the lowest PTR in 2001 was found in the province of Zaire (19.9), the highest were in the province of Lunda Norte (71.3) and Luanda (45.8). In 2003, the lowest PTR was 20.6 in Malange and the highest was 76.5 in Huíla and 59.7 in Cabinda. In the 2nd level of PE the variations ranged from 13.3 in Cunene to 66,3 in Lunda Norte in 2001 and 6,8 in Moxico to 96,2 in Lunda Norte in 2003 (Figures 2.11 and 2.12). 161. Finally, in the case of the 3rd level of PE the variations represented mainly an issue of efficiency as well as an issue of equity because no province had a higher PTR than 45 in 2001 and very few had so in 2003. In this level, more than in the others, teacher efficiency could be improved by increasing student intakes and therefore increasing access. The lowest PTR in 2001 was 7.8 in Cunene and the highest was 40.1 in Bengo. The lowest PTR in 2003 was 2.4 in Benguela and 52.2 in Luanda. As can be seen the variations in PTRs could be reduced and the PTRs could even increase in the provinces with the lowest ratios, This could potentially have positive cost implications without necessarily compromising the quality of education. Figure 2. 10: Variations in PTR's between provinces in 2001. Pupil-Teacher Ratios per province in 2001 80,0 70,0 60,0 50,0 Grades 1-4 40,0 Grades 5-6 30,0 Grades 7-8 20,0 10,0 0,0 o a é ob alí az az co er ng Bi eb nda eg onal Be nguele bia ma Hu neneu andou an an mi Uí Zai Lunda Lunda angela xio C Hu C K Ku Ku Luanda M M Na cia B N Note: in the case of Lunda Sul the number of teachers used was of 2001. Otherwise the PTR would be much higher. - 65 - Figure 2. 11: Variations in PTR's between provinces in 2003. Pupil-Teacher Ratios per province in 2003 120,0 100,0 80,0 Grades 1-4 60,0 Grades 5-6 40,0 Grades 7-8 20,0 0,0 é bo alí lu co eg ngoe Bi etro l Su etro S xi ebi eri Hu B nguelae abinda uam neneu N C ubango a Luanda Mo onalica B H C K azna N langea m Uí Za M Na N anzu Lunda Ku Lunda andou K K Note: in the case of Huíla and Kuando Kubango the number of teachers used was of 2001. Otherwise the PTR would be much higher. Effectiveness of the Angolan education system Cycle costs 162. An effective education system facilitates learning and produces graduates (pupils who complete the cycle). Low survival rates and completion rates may have important implications for the cost of producing a graduate. In a perfectly effective system of education there would be no dropouts or repeaters (or exam failures) and therefore the theoretical cost of completing 8 years of primary education would be the same as the actual cost of those completing grade 8. 163. Table 2.29 shows that in Angola, using 2000 and 2002 figures, it takes 26 and 30 years of resource inputs, respectively, to produce a primary education graduate instead of the prescribed 8. In addition, it takes 16 and 18 years of resource inputs to produce a pupil with grade 6, instead of the prescribed 6 years if the education system was perfectly effective and 9 years of resource inputs to produce a pupil with grade 4 instead of the prescribed 4. 164. The level of wastage, which is related to the high repetition and drop-out rates, is strikingly high. The comparison between Angola and Zambia, for example, which itself has a high level of wastage, highlights the differences. Whereas in Zambia it takes 9.9 years of resource inputs to produce a primary graduate instead of the prescribed 7, it takes 16 years of resource inputs to produce a grade 6 student in Angola. - 66 - Table 2. 29: Effectiveness indicators for school education, 2000 and 2002. 2000 2002 Grades 1-8 Input/output ratio 25.8 30.0 Wastage ratio 3.2 3.8 Share of total pupil years on drop-out/retention 69.0% 73.4% Theoretical spend per completer (US$) 534.4 1755.2 Actual spend per class 8 student (US$) 1710.0 6669.7 Grades 1-6 Input/output ratio 15.6 17.8 Wastage ratio 2.6 2.9 Share of total pupil years on drop-out/retention 61.7% 66.4% Theoretical spend per completer (US$) 400.8 1316.4 Actual spend per class 6 student (US$) 1042.0 3817.5 Grades 1-4 Input/output ratio 8.7 9.2 Wastage ratio 2.2 2.3 Share of total pupil years on drop-out/retention 54.0% 56.7% Theoretical spend per completer (US$) 267.2 877.6 Actual spend per class 4 student (US$) 587.8 2018.5 165. Adding to the points mentioned before, the Angola education system did not turn out to be a more effective system from 2000 to 2002 but rather a more ineffective system in the three levels of primary education. In 2000 it took 25.8 years of resource inputs to produce a graduate of the 3rd level of PE (grade 8), 16 years of resources to produce a graduate of the 2nd level of PE (grade 6) and 8,7 years of resources to produce a graduate of the 1st level of PE (grade 4). In 2002, however, these statistics increased to 30 years for grade 8 (up by 4 years), 18 years for grade 6 (up by 2 years) and 9.2 years for grade 4 (up by 0.5 years). 166. This means that the state spends much more on each graduate than it would in the absence of dropouts and repeaters. To translate these observations into costs: a hypothetical system with the same average costs as Angola and no dropouts or repeaters would spend US$ 1,755 producing a primary graduate compared with US$ 6,669 under the current cohort survival rate (using the unit expenditure for 2002 and the cohort analysis for the same year). In Zambia, for example, the state spends (using 2000 figures) 41 percent more on each graduate than it would in the absence of dropouts and repeaters, whereas in Angola the government spends (using 2002 figures) 280 percent more than it would had there been no dropouts and repeaters. F. POLICY RECOMMENDATIONS 167. Better and more detailed data will be necessary to carry out reforms in the education sector. The analysis in this Chapter has highlighted a serious weakness in the quality of the data available for the education sector in Angola. Without reliable and comprehensive data it will be very difficult to implement successful reforms in the educational sector. The existing data in terms of utilization rates and performance indicators for the sector is old and not representative as the latest MICS available was produced during the war and covered mostly urban areas in a few a provinces. A more recent and comprehensive set of indicators need to be produced to support and guide reform in the sector of education. - 67 - 168. Once new data becomes available, the authorities should introduce clear criteria for the allocation of resources among provinces and sub- systems of education. This should be established at the national level and not depend on the proposals coming from the Provincial Governments each year. One of the fundamental criteria should be total enrolment in the different sub- systems of education and provinces. 169. Student intakes in primary education and provinces where PTRs are much lower than the teacher-class ratio should increase. The same is true for the elimination of the variations of PTRs among provinces (one possible way to do it would be through multi-grade teaching which requires appropriate qualifications from the teachers). Moreover, efforts should be made to reduce the pupil per classroom ratios in all levels of PE. 170. The authorities should carefully consider options to reduce repetition rates. This is one of the main sources of wastage in the education system in Angola. One of these options could be the introduction of the automatic promotion between cycles of primary education. 171. The planning and budgeting process need to include steps or mechanisms for coordination (budget preparation) and information sharing (budget preparation and execution of resources) between MEC and the Provincial Governments. This is particularly important because Provincial Governments manage the resources for the sub-system of education (primary grades 1-6) that should have the highest impact in the sector, since it is the level where most of the students are enrolled and the basis for achieving the MDGs and EFA goals. In addition, it is the most cost-effective way to address equity issues, a role that can best be played by a central structure and in this case by MEC as the institution with more specialized knowledge regarding education, the national asymmetries and its national implications. 172. Dramatic improvements are needed in the sector's financial management system. Improvements are required in several instances ranging from budget preparation, disbursements, management of expenditures, expenditure classification and institutional capacity. An adequate financial management system would prevent payment of expenditures not budgeted for (except in special circumstances) which would in turn be an incentive to improve budget preparation. Regarding expenditure classification, it would be most relevant to separate in the functional classification the services of pre-primary and primary education and have a different function for teacher training which is different from secondary and technical education. In terms of institutional capacity, it should be also a priority to train those using the financial management system in inserting data according to the proper expenditure classification. 173. The capacity for monitoring and evaluation in the sector of education needs to be strengthened. Another policy recommendation relates to the establishment of a system to assess learning outcomes in all levels of the education system in Angola. A well functioning monitoring and evaluation system can provide information to the planning and decision-making levels as to whether those students who are actually graduating from each cycle of primary education have achieved the minimum standards of learning recognized - 68 - internationally and whether the main causes of inefficiency of the education system are related to endogenous or exogenous factors to the system. - 69 - ANNEXES - 70 - Annex 1 ­ Key Data Limitations The main limitations found in terms of data were: 1. Number of pupils in the medium level (from grade 9 to 12), per province, gender and age in the period from 1999 to 2003. From 1999 to 2001, there is some data on total number of pupils for this level of education in MEC. However, it is very difficult to use the data because in some years the numbers of certain provinces exist but they don't exist for others. The same applies to repetition and drop-out figures. 2. Enrolment and promotion data in higher education. Since there is another study dealing specifically with the sub-system of higher education, we will exclude the analysis of enrolment and promotion and present only expenditure trends. 3. Numbers of teachers per education sub-system. Regarding grades 1 to 8 (full primary education) data was collected for 1999 to 2001, 2003 to 2005. The data for 2002 was not available. However, the data is very inconsistent with impressive decreases and increases in the number of teachers from year to year in several provinces which had an overall impact in the total number of teachers. Regarding the medium level (grades 9 to 12) almost no data was available and certainly no data that could be used in a systematic analysis. 4. Teacher's recruitment process, deployment and pay schedule. Efforts were made to obtain this data from MEC but with no success. 5. Exams results. From the information gathered there is no systematic data available regarding exams results in the country. 6. Public expenditure information. In general, the classification of expenditure using the national functional classification was inconsistent and misleading. See section 4 for the details and methodology used to avoid the limitations for 1999 to 2003. For the years 2004 and 2005 the adjustment couldn't be made because the information obtained was at a much aggregated level. Therefore, the analysis of trends at the national and provincial aggregated level is possible and provides already some relevant insights but by functional classification should be made with consideration of these limitations or not done at all. 7. External pledged and actual expenditure in the education sector from 1999 to 2003. Very little information was obtained for that period regarding external investments in the sector. 8. Private costs to education. No information could be obtained regarding this topic and from the consultations done, it is also not available in the country. 9. National Population Projections. One of the main problems that the current study faced was related to the population projections. In the first place, the last national population census undertaken in Angola dates back from 1970 and since then only one adjustment was made to those projections which were published by - 71 - the National Institute of Statistics in July 1991. This last revision or adjustment to the census of 1970 took, as the basis, the total projected population for the year 1985 and real data obtained from partial census done in the provinces of Cabinda, Zaire, Luanda and Namibe. For the remaining provinces the basis were the structure of the population obtained in previous census (1970, 1960, 1950 and 1940) and the structure of population from those countries bordering with those provinces. This adjustment also considered the hypothesis that Angola was closed to internal and external migrations. The projections published in 1991 contain: total population per province from 1985 to 2010 for age groups 0-4, 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50-54, 55-59, 60-64 and above 65. It is not provided information by single age or gender. Apart from that information published by INE, the study based its analysis in: The Demographic Indicators for Angola, 2000-2025, published by the United Nations in 2003, which only contains total population estimates by gender in 2000, 2005, 2010, 2015, 2020 and 2025 and total population estimates per age group (0-4, 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50-54, 55-59, 60-64, 64-69, 70-74, 75-79, 80-84 and above 85) and gender in 2000, 2005, 2010 and 2015. The information is not provided per province. Population estimates by the Ministry of Health for 2005 based on vaccination campaigns and other health interventions which includes total population per province, irrespective of age and gender. Multiple Indicator Cluster Survey from 2001, which contains information on total population per single age per gender but not provinces. In addition, the MICS also provides information on the masculinity index (ratio male/female for age groups 0-4, 5-9, 10-14, 15-19, 20-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50 and above). According to the methods used there are estimates of total population for 2005 ranging from 14.533 million inhabitants (UN) to 15.556 million (INE) or to 17.681 million (Ministry of Health) and 29.811 million (MICS ­ estimate for 2000). In this context, it was necessary to estimate the population per province including gender and single specific ages (from 6 to 13 to estimate both gross and net enrolment rates as well as the gross admission rates and completion rates). In order to do this, the study considered the total population per age group provided by the INE publication from 1991 and estimated the distribution per province considering the proportions obtained by the Ministry of Health (see appendix 1 to illustrate the differences in the provincial distribution of population obtained by INE and the Ministry of Health). In order to estimate the gender distribution, the study considered the MCIS masculinity index (see appendix 2 for a comparison between the United Nations, the 1970 census and the MCIS masculinity indexes) per age group and estimated single ages using the Sprague program provided by MEC. - 72 - Appendix 1 - Total population estimates for 2005 and provincial distribution from INE and Ministry of Health. 2005 2005 2005 2005 Province INE 1991 Ministry of INE 1991 Ministry of Health Health Total 1.6E+07 17681601 100.0 100.0 Bengo 249000 226767 1.6 1.3 Benguela 939000 2432209 6.0 13.8 Bié 1674000 1176279 10.8 6.7 Cabinda 260000 478349 1.7 2.7 Huambo 2262000 1449837 14.5 8.2 Huíla 1248000 1595348 8.0 9.0 Cunene 333000 437209 2.1 2.5 Kuando Kubango 182000 438465 1.2 2.5 Kuanza Norte 575000 406372 3.7 2.3 Kuanza Sul 928000 1113581 6.0 6.3 Luanda 2644000 4487883 17.0 25.4 Lunda Norte 418000 581953 2.7 3.3 Lunda Sul 216000 267628 1.4 1.5 Malange 1334000 486163 8.6 2.7 Moxico 471000 496977 3.0 2.8 Namibe 201000 200535 1.3 1.1 Uíge 1288000 1157279 8.3 6.5 Zaire 343000 248767 2.2 1.4 Appendix 2 - Masculinity index per age group using the following sources: UN 2000, 1970 population census and MICS 2001. Age 1970 Population MICS Masculinity Group UN 2000 Census Index M F M F M F 0-4 50.0 50.0 50.3 49.7 49.5 50.5 5-9 49.8 50.2 52.7 47.3 50.0 50.0 10-14 49.8 50.2 54.3 45.7 49.5 50.5 15-19 49.8 50.2 53.5 46.5 47.0 53.0 20-24 49.6 50.4 52.0 48.0 39.0 61.0 - 73 - Annex 3 ­ Glossary of performance indicators in education Adult literacy rate (percentage of the population aged 15 + that is literate): The percentage of the population aged 15 years and over who can both read, that is understanding a short simple statement on his/her everyday life, and write. Gross admission rate: New entrants into the first grade of an education cycle, regardless of age, expressed as a percentage of the relevant population in a given school-year (e.g. 6 year old for grade 1). Gross enrolment rate: Total enrolment, regardless of age, at an education level, expressed as a percentage of the official school-age population in a given school- year. Net enrolment rate: Total enrolment, only of students of the official school age, at an education level, expressed as a percentage of the official school-age population in a given school-year. Repetition rate: Proportion of pupils enrolled in a given grade in a given school- year who studied in the same grade the following school-year. Drop-out rate: total number pupils, in a given grade, who have left school during a given school-year, but before the next school-year, divided by the total number of the pupils enrolled in the given grade in the beginning of the school-year. Transition rate: The number of students admitted to the first grade of a higher level of education (e.g. 6th grade in EP2) in a given year, expressed as a percentage of the number of students enrolled in the final grade of the lower level of education (e.g. 5th grade in EP1) in the previous year. Measures for education effectiveness Various measures of education effectiveness are defined below: Completion rate: the total number of students graduating from the final year of a cycle of education e.g. primary, regardless of age, expressed as a percentage of the population at the official graduation age. Cycle costs: the cycle cost is the average cost to produce one graduate of the cycle. In a system where no students repeated years or dropped out, the cycle costs would be the same as the unit costs multiplied by the number of years in the cycle. Success rates: in a system where the cycle is completed by examination then effectiveness can be judged by the results, provided that the pass rates on the examination reflect mastery of the curriculum. Educationists use a variety of indictors to analyse the factors driving up cycle costs. Some of these are pupil flow indicators derived from the Reconstructed Cohort Method. These can be calculated as follows: - 74 - Survival rate: percentage of a cohort of pupils enrolled in the first grade of a given cycle of education, in a given school year, who are expected to reach each successive grade. Average duration of study by graduates: total number of pupil- years graduates spend in the school system divided by the total number of graduates. One school- year spent in a grade by a pupil is defined as one pupil-year. Average duration of study by dropouts: total number of pupil-years dropouts spend in the school system divided by the total number of dropouts. Average study time for cohort: total number of pupil-years spent in the education system by graduates and dropouts divided by total number of graduates and dropouts Input: Output ratio: total number of pupil-years used by the cohort divided by total number of graduates. Wastage Ratio: this is another way of expressing the input-output ratio in terms of wasted pupil years (on repetition or drop-out) relative to the optimal number of years. This is calculated as the ratio between the actual input output ratio and the ideal input output ratio Share of total pupil-years on drop-out/repetition: total pupil-years used by drop- outs and repeaters expressed as a percentage of total pupil- years spent by the cohort - 75 - Annex 4 ­ Functional classification of expenditure In terms of functional classification of expenditure the study considered the following sub-functions in that period: a. pre-primary and primary education services which also includes the entire administration costs of the provincial directorates of education and municipal delegations of education when these are included; b. secondary education services divided between secondary services and teacher training services; c. tertiary education services which only includes information related to the sub-system of higher education; d. other education services non identified by levels which includes the adult literacy sub-system and special education; e. subsidiary services to education which contains data on expenditure in sports, school feeding and scholarships; f. other services of education which includes the expenditures of the central Ministry of Education for the education function; g. other services of culture which includes the expenditures of the central Ministry of Education for the function of culture; h. other services which includes the expenditures from other Ministries and government structures in the education function. The only main difference between the classification used in the study and the official classification is that f), g) and h) are grouped into one single function in the official classification whereas in the study it seemed important to establish a difference between what MEC spent for the education function separately from the culture function and also to include the expenditures of other Ministries and decentralised government structures in a different category. In addition, and most importantly, since there was a lack of consistency in terms of attribution of expenditure in accordance with the sub-functional classification of expenditure, each single line of expenditure was reclassified sub-functionally for each single year, whenever possible. If to consider the classification as it was provided, most of the education expenditure would fall into other services of education and it would be very difficult to make any assessment in terms of levels of education, for example. Since it was not possible to obtain as detailed expenditure information for the years 2004 and 2005, the study uses simply the functional classification provided in the Ministry of Finance publications, namely: - 76 - (b) Primary education; (c) Secondary education; (d) Higher education; (e) Adult education; (f) Other services of education. However, we did not find an official source of information which would explain how is the expenditure classified among those sub-systems. For example, we do not know whether expenditure in sports, school feeding programs or scholarships is classified in each sub-system it refers to or whether it is included in other services of education. Another example would be the expenditure by the provincial directorates of education (whether they are classified under primary education or other services of education), among many others. - 77 - CHAPTER III PUBLIC SPENDING IN THE HEALTH SECTOR A. THE CONTEXT: MULTIDIMENSIONAL CHALLENGES 174. In Angola, the challenges facing the health sector are multi-dimensional and across all levels. When evaluating a country's health status, it is important to take into account three levels--health outcomes, health service coverage, and health activities--as well as its broad political, economic and social context. This is because the health system operates within a given context. It is in that context that health activities are organized and managed in order to increase service coverage, and in turn improve health outcomes. This section analyses the challenges that affect the health sector both in a broad context and across the three levels. Challenges at the health outcome level 175. Angola has not gone through the demographic and epidemiological transitions yet. Its growing population is estimated at 14.5 million, of whom about 60 percent are less than 18 years old. The epidemiological profile in Angola features a high prevalence of communicable diseases and high child and maternal mortality. The infant mortality rate is 154 per 1000 live births; the under-5 mortality rate is 260 per 1000 live births (MICS 2001); the total fertility rate is estimated to be 7.2 births per women (MICS 2001); and the average life expectancy is only 40 years. The maternal mortality ratio is reported by WHO at 1,700 per 100,000 (2003), one of the highest in the world. This compares unfavorably with other Sub-Saharan African countries, which themselves have significantly higher rates compared to the rest of the world (Table 3.1). One in every seven pregnant women dies from avoidable reasons. Table 3. 1: Key health outcome indicators Sub-Saharan Indicator Angola Average Life expectancy at birth (years - 2003) 40 49 Fertility rate (2002) 7.0 5.0 Infant mortality rate (per 1000 live births - 2000) 154 92 Under-five mortality rate (per 1000 live births-2000) 260 171 Maternal mortality ratio (estimates) 1,700 914 Contraceptive prevalence/100,000 (2003) 6.0 22.9 GDP/Capita US$ 975 1,073 Source: MICS 2001 and World Development Indicators 2006 - 78 - 176. Health status has not improved over time, even after the peace agreement. For example, as shown in Figures 3.1 and 3.2, although the average infant mortality rate and under-five mortality rate of the Sub-Saharan Africa region and the world as a whole have been declining consistently, child mortality in Angola has stagnated since 1980, showing no signs of improvement. Figure 3. 2: Under-five mortality rates in Angola and Figure 3. 1: IMR in Angola and Sub-Saharan Sub-Saharan Africa Africa 250 400 350 200 shtrib 300 250 150 Angola IMR shtrib Angola U-5 MR veil Sub-Saharan IMR veli 200 Sub-Saharan U5MR 000 100 World IMR 150 World U5MR 1,rep 1,000reP100 50 50 0 0 1960 1970 1980 1990 1995 2000 2004 1960 1970 1980 1990 1995 2000 2004 Source: World Bank: World Development Indicators 2006 177. Child mortality. Child mortality is mainly caused by malaria, acute diarrhoeal diseases, acute respiratory infections, measles and neonatal tetanus, which account for about 60 percent of child deaths (Figure 3.3). These can be easily prevented or treated at the primary health care level, and through healthy practices and care at the household level. Malnutrition is the main associated cause of child mortality. Figure 3. 3: Causes of under five mortality, Luanda 2001 Others , 16% M alaria, 23% M easles, 2% Tetanus, 2% Premature birth, 7% Sepsis, 8% Diarrhea , 18% Asphyxiation Neonatal , 9% ARI, 15% Source: Ministry of Health -Angola. Studies, Planning and Statistics Office. Deaths in Luanda cemeteries.2002-2003 - 79 - 178. Regional differences in child mortality are salient in Angola. The west, central and capital regions have the highest under-five mortality rates (Figure 3.4). Noticeably, these regions also possess the highest population concentration, which indicates that a large share of children deaths occur in these regions. Figure 3. 4: Under-five mortality by region Under-five mortality by region 350 sh 300 rtib 250 veil 200 150 000 100 1,reP 50 0 Capital: West: Centre-South North: Zaire, South: Huila, East: Lunda Average Luanda, Bengo, Benguela and Huambo, Bié Uige and Namibe, Norte, Lunda Kuanza Norte Kuanza Sul and Kuando M alange. Cunene Sul and and Cabinda. Kubango M oxico. Region Source: Ministry of Health -Angola. 179. Unlike other countries where child mortality rates in rural areas are considerably higher than in urban ones, in Angola they are almost the same. This may be partly explained by the major displacement of the rural population into the cities due to the war, the lack of access of the peripheral urban population to health care, the low quality of services, and the added environmental risks. 180. Maternal mortality. The main causes of death for women are malaria, hemorrhage, eclampsia, abortion complications and prolonged labor. Among them, complications in pregnancy, birth and after birth occur in about 15 percent of all pregnancies. Many of these causes are directly associated to poverty. Yet, with emergency obstetric care in health facilities, skilled staff, proper surgical equipment, a safe blood supply, and sufficient drugs, maternal mortality can be greatly reduced. Experience in Angola itself shows that this is possible. In the Malanje province, increased access and limited training provided in the provincial capital, helped to reduce the maternal mortality ratio (MMR) from 3007 per 100,000 in 2000 to 1085 per 100,000 in 2003.10 181. Malnutrition. Angolan children suffer from a poor nutritional status. According to the 2001 MICS, about 45 percent of children were malnourished. Angola has a high level of stunting and wasting which will make children vulnerable to diseases and health problems, and this could have enormous social and economic implications in the future. 10 Malanje DPS - 2003 Annual Report (draft) - 80 - 182. Environmental health. Environmental health problems are significant. Only 56 percent of the population of Luanda had access to piped water at home or from public pipes (1998 data), and the situation has not greatly improved since 2002. In the crowded large cities, people live near mounds of uncollected rubbish and stagnant water. Less than a quarter of the population of Luanda is served by a sewage system (Hodges 2004). 183. Communicable diseases. Angola suffers from a heavy weight of communicable diseases, which have affected the economic recovery and quality of life. Malaria, tuberculosis, diarrhea, HIV/AIDS are among the most serious problems. 184. Malaria. Malaria is regarded as the principal cause of mortality and morbidity in the country with a total of 3.25 million cases and 38,000 deaths reported in 2003 (USAID 2005). In 2005, malaria was the most common cause of visits to health facilities, accounting for more than 60 percent of all outpatient visits and deaths by transmissible diseases. According to the National Malaria Control Program of the MOH, every Angolan typically experiences 3-5 malaria episodes per year. 185. Diarrhea. The prevalence is 25 percent among the under-five year olds (MICS 2001), but only 7 percent of these cases were treated with rehydration fluids and continued feeding. The use of oral rehydration salt (ORS) was only 40 percent in these cases. 186. Tuberculosis. In total, about 21,000 cases of tuberculosis (TB) have been reported in the country, and it is estimated that around 7,000 new cases are diagnosed every year. TB is one of the common reasons for visits to health facilities. Recent estimates indicate that tuberculosis prevalence is increasing. Since 1996, the country has adopted the DOTS strategy for treatment. However, it only covers 9 percent of health facilities (138 out of 1,465 facilities) in the country and about 30 percent of the population. 187. HIV/AIDS. The overall HIV prevalence rate in Angola is estimated at 2.5 percent (MOH), which is not very high when compared with Southern Africa neighbors. There is much variation between provinces with a minimum of 0.8 percent in the central province of Bié and a maximum of 11 percent in southern province of Cunene that borders Namibia, a country with a HIV prevalence of 20 percent. In a survey carried out in 2001 in pregnant women, HIV prevalence was shown at 8.6 per cent against 3.4 per cent in 1999. Also in sex workers, surveys demonstrated a rapid increase in prevalence from 19 percent in 1999 to 32.8 percent in 2001. 188. Outbreaks: During the last five years there have been outbreaks of Marburg disease in Uige, and meningitis and cholera in Luanda and some southern and central provinces that put an added burden on the already weakened health system. - 81 - Challenges at the health service coverage level. 189. The coverage of basic and effective health services is low, indicating that the majority of the population is not protected by basic health services. As shown in Table 3.2, except in a few areas, Angola is worse than most of Sub-Saharan Africa countries in coverage of key services. Table 3. 2: Health service indicators Angola Sub-Saharan Africa Indicator Measure Source Year Data Source Year Data Doctors in the public Doctors per 100,000 MOH 2000a 5 HDR 1991 32 sector inhabitants WHO 1997 7.7 WHO 1995- 16 SIS SIS 2003 Access to drugs % of the population with WHO 2001 20 access to essential drugs n/a Vaccination % of children 12-23 MICS 1996 24 SOWC d 1999 46 coverage for DPT months vaccinated 2001b 34 (3rd dosage) Vaccination % of children 12-23 MICS 1996 28 SOWC d 1999 48 coverage for polio months vaccinated 2001b 63 Deliveries in health % of deliveries attended MICS 1996 22 SOWC d 1995- 39 facilities by trained health 2001b 45 2000 personnel Pre-natal % of pregnant women who MICS 1996 64 SOWC d 1995- 64 consultations attend one or more pre- 2001b 66 2000 natal consultations Contraceptive use % of women of MICS 1996 8 reproductive age who use 2001b 6 1995- SOWC d 22 any method of 2000 contraception MSH 2002c 17 Source: adapted from Vinyals 2002 Notes: a) Estimate based only on doctors in the National Health Service, as published in the MOH's "Anuário Estatístico do MINSA" for 2000. b) MICS 2001 survey only covered areas accessible during the war. c) The Management Sciences for Health (MSH) survey covered only 3 municipalities in Luanda province. d) State of the World's Children, UNICEF 2002. 190. The health service delivery system was badly damaged by the war, with close to 65 percent of the health facilities destroyed, and most of the equipment either stolen or deteriorated by years of lack of maintenance. The lower level facilities suffered the most damage. Even after peace, a high percentage of health facilities are still not functional, mainly due to a damaged infrastructure that needs to be rebuilt or repaired, and the lack of staff and key inputs. The result of all this is that more than 60 percent of the population does not have reasonable access to health care. The country's response to the challenges 191. The government has formulated a number of strategic documents such as the Sector Development Plan for 2000-2005, the 2005-2006 Health Sector Program, and the Accelerated Program to Reduce Child and Maternal Mortality. These documents provide a general direction to which the health sector should be heading to. However, they usually do not link expected outcomes with effective and efficient spending and do not include detailed implementation arrangements. As a result, resources are not - 82 - being used optimally and better health status has not been achieved. Specific issues around health spending include: · Health spending is not well aligned with the country's main health problems and burden of diseases; · Resources are not planned to target the cost-effective interventions; · Investment strongly favors hospital care, as 40 percent of public spending on health goes to tertiary care, and only 27 percent goes to primary and secondary care; · The facility-heavy spending is only benefiting a small proportion of the population (mostly urban, and a large proportion in Luanda) as more than 60 percent of citizens do not have access to health facilities; · Spending on the critical inputs (human resources, drugs, supplies) is not aligned with the need to tackle the main health problems; · Despite the urgent need, budget execution is low, only 65 percent on average in 2004 and 2005 · While success was achieved in some vertical programs, such as the expanded program of immunization (EPI), the effort has not been integrated across the health system, causing duplication of efforts and a waste of resources; and · Following decentralization, the coordination between the MOH and provinces has not been optimal, causing inconsistent and ineffective spending across provinces, and also away from the national goals. 192. It is a critical time now for Angola to ensure that the rebuilding efforts are effective, and resources are well used. International experience can be useful for Angola to learn from other countries' experience in ineffective spending. In this context, it is imperative for Angola to ensure that: · Money is spent toward solving the main health problems, such as high child and maternal mortality and the high level of infectious diseases; · Money is spent on cost-effective services, which have an impact on the main health problems; · Money is spent on workable and integrated service delivery arrangements; · Money is spent on the necessary inputs of the service delivery system; and · Money is spent on the frontline of services and the removal of bottlenecks of expending service coverage. B. THE STRUCTURE OF THE HEALTH SECTOR 193. The National Health System (NHS) in Angola consists of the Ministry of Health (MOH) and the primary health care network. The MOH is responsible for: (i) the definition of sector policies: (ii) the preparation, monitoring and evaluation of strategic plans for the sector; and (iii) regulation. Provincial governments are in charge of the provinces' primary health network, ensuring the functioning of all units - 83 - except provincial hospitals, which are directly financed by the Ministry of Finance (MOF). 194. Under the MOH, the main body for delivering the essential package of health services is the National Public Health Directorate. Its organization is reflected in Chart 1. Chart 1: National Public Health Directorate DEPUTY MINISTER OF HEALTH NATIONAL PUBLIC HEALTH DIRECTORATE Logistics Health Promotion Office National Program of Essential Drugs Statistics DEPARTMENT OF DEPARTMENT OF DEPARTMENT OF DEPARTMENT OF REPRODUCTIVE HYGIENE AND DISEASE CONTROL SPECIAL PROGRAMS HEALTH EPIDEMIOLOGY Women's National Malaria Health coverage Immunization health department Control Program department department Nutrition department Epidemiological surveillance Infant and adolescent health department 195. The organization structure reflects a vertical approach, and the system is heavily organized by diseases and health problems. This organization encourages vertical arrangement of service delivery and the creation of sub-service delivery systems. The national strategy to reduce child and maternal mortality requires a horizontal arrangement around service delivery modes. In order to integrate efforts and avoid duplication, the government should coordinate the vertical programs and existing system functions (such as planning, human resource, procurement, etc.). 196. Access to health care is still very limited, especially in the provinces. The public health care provision model is divided into three levels: (i) the primary level consisting of 228 municipal hospitals and referral health centers, 1,450 health posts, and community health workers; (ii) the secondary level, comprising 32 general - 84 - hospitals, including provincial hospitals; and (iii) the tertiary level, consisting of 8 central and specialized hospitals. At the peripheral level (primary network and municipal hospitals, which form the basis of the NHS) one finds health units in the municipal capitals and some additional ones in the communal capitals. 197. The primary network does not function properly, and the population tends to go straight to provincial hospitals. The normal logic of a health system is that patients access first the lower level of care that should serve as a first filter for more simple cases, before going to the higher levels. This does not apply in Angola. As a result, hospitals are overloaded with patients that could be treated at a lower level. In such a setup, the referral and counter-referral system between levels is not operative. Provincial hospitals lack a systematic relation with municipalities or provincial health centers, and the latter with municipal health centers. C. THE BUDGET CYCLE IN THE HEALTH SECTOR 198. The budget preparation process is led by the MOF. It begins in June and ends in September. The first step is the updating of the country's macroeconomic framework and the preparation of government revenue projections. This exercise is carried out by the MOF and by the Ministry of Planning (MOP). They establish budget ceilings for provinces and sectors, an exercise that determines the overall allocation of resources for the sector. 199. The steps for determining current expenditures, including personnel and goods and services, are as follows: i. The process begins when budgetary units (BUs) receive guidelines from the MOF, placed on its website (www.minfin.gv.ao). The guidelines include application forms to be filled in by BU managers and submitted to the National Budget Directorate. ii. The BUs and their dependent units (DUs) have 3-4 weeks to prepare budget proposals, using the program budgeting methodology. The government does not take full advantage of this approach because budget preparation is not usually preceded by an explicit planning exercise (with definition of priority activities and costing). Programs are defined by the MOF, without prior work with the sectors, taking as a basis the programs included in the government Plan of Action. Monitoring indicators are not required and there are conceptual errors in the proposals. iii. Budget proposals are adjusted to the ceilings defined for each BU and DU. Ceilings for the DUs are prepared by the responsible BUs. Provincial governments decide the budget ceilings for all sectors in the province, except for provincial hospitals, which are BUs and have their ceiling determined directly by the MOF. During the last - 85 - two years, the process was made easier in the provinces with access to the internet and where the SIGFE functions on-line. iv. Once the proposals are received by the National Budget Directorate (NBD), via the Provincial Finance Directorate, the technical staff consolidates them and corrects possible errors. There is no real negotiation between BUs and the NBD, as shown by the fact that budget ceilings are rarely altered. 200. Investment planning and budgeting are made in parallel. This process is led by the MOP which deals directly with the ministries' planning departments, and provinces. The investment budget preparation process lacks technical instruments. In the health sector, there is no map of health services to guide investment requirements as a function of the population and the epidemiological profile. Nor is there a national investment plan that would define medium term interventions for the national, regional and provincial level units. Such a plan would normally be followed by provincial plans which would design interventions at municipal and communal level. These plans would include current expenditure projections, principally human resources and drugs required for the new infrastructure. The absence of these instruments puts at risk the coherence and sustainability of the government's ongoing investments. 201. The institutions taking part in the allocation process for goods and services are the MOH, the NBD, and the MOF which set the budgetary ceilings jointly. In the provinces, the ceilings for provincial hospitals are provided by the NBD and the MOF, so far without the participation of the MOH. Provincial governments define the ceilings for their dependent units, namely the Provincial Health Directorate (PHD) and municipal hospitals. Regarding personnel, resource allocation in the short term is semi-automatic, and resources are put where the employees are based. In the case of goods and services, there is much more flexibility. As for investments, given that commitments are made for multi-annual investments, flexibility in allocation is not as high as for goods and services. The allocation of resources for investment is made by the MOP, the MOF, the MOH and the provincial governments. 202. The micro-planning process is led by the MOF. Over the last decade the MOH did not lead the strategic planning process, as it should as sector leader. There have been attempts since 1995 to prepare a medium term strategic plan and the MOH is currently reviewing the National Health Policy. However, to date the sector remains characterized by the absence of planning documents and regulation for the 1992 law, which substitutes for the National Health Policy in many aspects. 203. With respect to macro planning, the budget preparation is the only time apart from budget execution when health institutions undertake a planning exercise for their activities. Officially, program budgeting is used, which links resources to specific programs, with objectives, targets, indicators, and detailed costs. However, as this methodology is not well implemented, there is no link between inputs, outputs, - 86 - and outcomes. A major concern is that budgeting is made by administrative staff, with little involvement of key managers. 204. Another problem is the absence of connection between the planning at the central and provincial levels. Generally, health programs at the national level plan their activities without knowing the needs at the provincial level. Since provinces need the central level to finance the most expensive resources such as drugs or information, education and communication materials, the lack of coordination results in a deficient implementation of program activities. In some provinces the only public health activities are those receiving support from international partners. 205. The National Assembly exerts limited oversight on the budget. When the 2006 budget proposal was submitted by the government, deputies only scrutinized some visible aspects such as the proportion of health spending in the total budget or expenditures for endemic disease control programs. To reinforce its role as controller of government finances, the National Assembly should train deputies in budgetary analytical skills and use comparative data from countries in the region. 206. State financial management is characterized by a single account system with payments made through bank transfers between sub-accounts of the Single Treasury Account and suppliers. First, each BU submits a cash plan for the following quarter to the MOF and to the Provincial Directorate of Finance. Budget execution starts when BUs and DUs receive financial quotas (FQs), i.e. the financial limit they must execute during a certain period, generally a month. At that point, units verify whether the expenditure to be made is covered in the budget and available at that moment. This allows them to start issuing cash drawing orders as payment documents to suppliers. Since 2004, in some provinces with access to the internet11 state budget execution is made on-line. Purchases using state funds can only be made from licensed suppliers with a taxpayer card and, in some provinces, following public tenders at the provincial level. All informal suppliers are excluded from the process. 207. In practice, managers do not consider that the budget is a rigorous exercise that serves to plan activities for the year. They perceive that another planning exercise must be undertaken when FQs are received. At the beginning of the execution process, a micro-planning exercise is carried out to decide the activities to be financed with the financial quota for each period, generally on a monthly basis. 11 The provinces with an online execution of the budget are: Luanda, Cabinda, Benguela, Kwanza Sul, Bengo, Huíla, Namibe and Cunene. In 2006, the online execution was expected to be expanded to Huambo, Kwanza Norte, Kuando Kubango, Lunda Sul and Malange. - 87 - Table 3. 3: Budget execution rate for goods and services in municipal, provincial and national hospitals 2002 2003 2004 2005 Municipal (DU)a 78% 77% 33% 82% Provincial (BU) 88% 76% 59% 90% National (BU) 93% 91% 83% 95% Source: Ministry of Health. 208. There are significant differences between budget execution rates at the central, provincial, and municipal levels (see Table 3.3). One of the determining factors seems to be the financial decentralization statute. The execution rate for goods and services is significantly higher for national and provincial hospitals than for municipal hospitals. This can result from the financial limits, which are lower in municipal hospitals and from the execution periods, which are substantially longer at this level. This is because municipal hospitals depend on the budgetary unit "provincial government". For example, the dependent units of the municipality of Andulo, in Bié, may need up to one month to make a payment because it must be authorized by four institutions: the health unit, the municipal administration, the PHD, and the provincial government. According to the technical staff of the MOH, the same procedure takes only 2-3 days at the central level. Table 3. 4: Budget execution rate by economic category 2000 2001 2002 2003 2004 2005 Personnel 79 % 80 % 88 % 73 % 62 % 69 % Goods and services 55 % 70 % 86 % 81 % 71 % 88 % Transfers 79 % 89 % 92 % 90 % Investment 74 % 68 % 87 % 59 % 46 % 43 % Total budget 70 % 76 % 87 % 75 % 62 % 67 % Source: Ministry of Finance. 209. Table 3.4 presents the execution rate for different economic categories between 2002 and 2005. Until 2002, the execution rate for personnel was higher than for goods and services, but since 2003, it has been systematically lower. This may be the result of generous provisions made by the state budget for personnel, foreseeing an immediate introduction of new staff or a generalized salary increase. Another possibility is inadequate planning. Although the MOF provides a large volume of resources for personnel, the determinants of payment are the wage sheets controlled by the Ministry of Public Administration, Employment and Social Security (MPA). This problem can be easily solved through a better coordination of data between the MPA and the MOF. 210. The budget execution rate for investment is lower than for other categories, and it has decreased over the last three years. A factor that can explain this trend is the allocation in just one year of the whole amount for some investments that need to be executed over several years. - 88 - 211. Managers do not have the appropriate academic level and profile. Many health units are led by paramedical staff with no basic accounting skills. Weaknesses at both hospital and peripheral levels are shown by the difficulty in gathering basic statistical data. There is a lack of management instruments such as accounting programs, budget classifiers, and guidelines. Materials for the training of managers are imported and not adapted to the day-to-day situation in Angola. 212. The responsibility for the financial management of investments rests with the MOF, which deals with the BUs, i.e. the MOH and the provincial governments. The MOF negotiates and manages external funding for investments. Given the enormous volume of resources from the new credit lines of the government, the MOH has created a Technical Support Office in charge of supervising investments made with external funding. The establishment of this office, which works closely with the MOH's Planning Department, has improved government capacity, but the technical ability to manage investments remains weak. 213. Financial management procedures and instruments have been developed, but less has been done regarding tools for accounting and monitoring and evaluation. Much attention is given to the publication of budgets approved by the National Assembly. However, information about budget revision and execution is not taken as seriously and it is not well publicized. The main accounting instruments are produced by SIGFE, but they do not allow for an adequate monitoring of the health sector. This results from the fact that the economic category classifier is common for all sectors. Categories are too broad and do not attend to the particular needs of a sector. 214. The Health Information System (HIS) was not designed to interface with SIGFE or SINGERH (the MPA data base for the management of public employees), and there is not much rigor in expenditure classification. It is common practice to use the "other services" category for the purchase of all kinds of goods. This makes it impossible to monitor the performance of the sector on a regular basis, and as a result, when an analysis of the sector is necessary, special studies need to be conducted. Although the Tribunal of Accounts has begun to play a disciplinary role in the management of public funds, irregularities are common, notably in public tenders. Proposed planning calendar for the health sector 215. Chart 2 below shows a planning calendar which combines the actions to be taken by the MOH with the MOF (upper boxes), and the actions to be undertaken within the sector's institutions (lower boxes). During the last two years, the MOH has been promoting an explicit planning calendar, but it has not been implemented. Among the reasons behind the failure are: (i) the lack of concrete instruments which managers can use, with an added value for their work; (ii) the poor leadership by the MOH; and (iii) the limited technical capacity of the MOH Planning Department, - 89 - either qualitative or quantitative12. The key to success will rely upon: (i) the creation of concrete instruments; and (ii) a stronger effort by the MOH to increase its technical capacity, by recruiting new personnel or subcontracting the development of the instruments. Chart 2: Proposed planning calendar for the health sector 7 March 15 June Constitution of Joint setting of budget THE MOH working 1-15 April ceilings per level, province group Public expenditure analysis 15 July and review of allocation of the previous year indicators Defined ceilings approved by the sectors 1 May Balance meeting MINFIN- THE MOH on the 31 May execution in the previous Definition of priorities of year the sector, consistent with the PESS March April May June July 1 April Adjustment of the budget 31 of July Training support proposals o the attributed teams for the provinces ceilings 1 May Provincial diagnosis : evaluation of 15 June - 15 July activities of the previous Estimate of the needs of year each institution of the sector 216. Two aspects are important for the budget preparation process. First, the MOF and the MOH must work in coordination, establishing joint commissions or routine meetings during the year to evaluate sector performance and negotiate an adequate resource allocation to achieve the objectives. Second, the MOH should lead the preparation of the planning instruments for: (i) budgeting purposes; and (ii) other management aspects of health units, such as accounting (per cost center in future), human resources management, procurement, and logistics. 217. The figure suggests a 5-month calendar for planning and budgeting within the health sector, in line with the timetable of the MOF. The process would start on March 7. An MOH working group would lead the process until its completion. The first mission of this working group would be to analyze the spending of the previous year, and review production and impact indicators. This analysis would be presented at a review meeting at the beginning of May. A month later (May 31), the working group would define the priorities of the sector for the following year. These priorities should be in line with the Health Sector's Strategic Plan (to be developed). On June 15, the working group would make an internal allocation of resources (budget ceiling 12 The Planning Department of the MOH (GEPE) has made a considerable effort to increase the number of economists in the Planning Department. However, the largest shortages are in public health. The entire GEPE has only one doctor. - 90 - scenarios) consistent with the priorities and the analyses of the previous year exercise. Lastly, on July 15, the MOH would defend this resource allocation proposal before the MOF, in a joint meeting. 218. Parallel to this, at the beginning of April, the MOH would need to provide training to the teams that would support the PHDs and provincial level planning institutions. At the beginning of May, each province, with support from the MOH and partners, would carry out an evaluation of the province's performance in the previous year. Between June 15 and July 15, the needs of the health institutions of each province would be estimated, using the methodology introduced by the MOH in 2004. At the end of July, coinciding with the beginning of budgeting by the MOF, the estimated needs would be adjusted to the budget ceilings attributed. 219. For this process to work, it will need instruments for each of the defined stages. Together with the partners providing institutional support, the MOH should develop these instruments and establish its planning function. D. THE STRUCTURE OF EXPENDITURES IN HEALTH 220. Table 3.5 shows the expenditures by the economic categories defined under the State Financial Management Integrated System (SIGFE): personnel, goods and services, transfers, and investments. Except for 2003, personnel has been the largest category, varying from 44 percent of total expenditures in 2000 to 47 percent in 2005. Personnel expenditures have remained stable because they represent a long-term commitment. Spending on goods and services has fluctuated from 26 percent in 2000, increasing to 49 percent in 2003, and then falling to 36 percent in 2005. After reaching a low point of 6 percent in 2002, investment increased until it reached 17 percent of the total in 2005. The substantial increase in investments in 2006 make it the largest category of expenditures for that year. Table 3. 5: Expenditures (2000-05) and budget (2006) per economic category (US$ mi and percentages) Category 2000 2001 2002 2003 2004 2005 2006 Personnel 79 44 % 124 47 % 105 49 % 120 40 % 135 43 % 208 47 % 475 37 % Goods and 67 37 % 110 42 % 97 45 % 148 49 % 129 41 % 163 36 % 299 23 % Services Transfers 0 0 % 0 0 % 0 0 % 1 0 % 0 0 % 0 0 % 0 0 % Investments 33 19 % 29 11 % 11 5 % 30 10 % 50 16 % 77 17 % 523 40 % Total 179 100 % 264 100 % 213 100 % 299 100 % 314 100 % 447 100 % 1.296 100 % Source: Ministry of Finance. 221. In 2005, expenditures for personnel increased in absolute terms to reach US$208 million. The main reasons were wage increases and employment of new staff, estimated at 12,500 new employees. Normally, the entry of new staff and the motivation from the salary increase should help improve the sector's performance, but the impact is likely to be limited since a large proportion of the new staff is administrative and not technical. The number of administrative staff grew from about - 91 - 15,000 in 1999 to 27,000 in 2005. Figure 3.5 shows that the proportion of administrative staff increased from 42 percent in 1999 to 53 percent in 2005. Figure 3. 5: Breakdown of the health sector labor force: 1999 and 2005 1999 2005 Medical Medical 3 percent 2 percent Administrative 42 percent Administrative 53 percent Technical 45 percent Technical 55 percent Source: Ministry of Health. 222. Investments increased substantially and reached US$76 million in 2005. The main reason was the reconstruction and rehabilitation of hospitals, including municipal hospitals, partially financed by credit lines. The 2006 health sector budget included US$400 million for investment. Efforts towards rebuilding the sector are likely to continue in the coming years. 223. Investment in the health sector faces severe difficulties that may limit the government's capacity to effectively use the resources. The absence of an investment plan for the sector, and the lack of coordination between government institutions is just one example of such difficulties. Public investment is managed by the MOH at national and provincial levels, and by provincial governments responsible for the primary network. Various investment initiatives including those of provincial governments and the Social Action Fund (FAS), among others, are not controlled or coordinated by the MOH, leading to a potential duplication of efforts and inefficiencies. 224. It seems that there is no clear policy according to which the government determines how much resources should be assigned to each level. The allocation is influenced by long-term commitments for personnel and investment that tend to be stable over time, and short term commitments for goods and services for which spending can fluctuate. Expenditures for hospitals and administration represent a large proportion of sector spending (Table 3.6). At the primary level, the proportion of human resources in total expenditures is higher compared to the second and third level. - 92 - Table 3. 6: Analysis of spending by level of care Total expenditures Goods and services Level 2001 2002 2003 2004 2005 2001 2002 2003 2004 2005 Primary network and 26 % 29 % 30 % 33 % 40 % 30 % 27 % 20 % 16 % 29 % Municipal Hospitals 2nd and 3rd Hospital Level 51 % 55 % 49 % 51 % 44 % 42 % 62 % 42 % 46 % 44 % Administration and other 20 % 13 % 16 % 12 % 13 % 27 % 10 % 34 % 36 % 24 % Training 3 % 3 % 4 % 3 % 3 % 1 % 2 % 4 % 3 % 2 % Total 100 % 100 % 100 % 100 % 100 % 100 % 100 % 100 % 100 % 100 % Source: Ministry of Health. 225. Although, there is no ideal allocation for each level of health care, two conclusions can be drawn. First, it seems clear that the relative weight of administration in relation to goods and services should be reviewed, since this level is not directly responsible for the functioning of NHS units. Second, if the national health policy objective is to rapidly improve the health status of the population, it is more efficient to spend one additional dollar on primary health care than at the hospital level. Table 3. 7: Provincial expenditures (US$ million) Province 2000 2001 2002 2003 2004 2005 Bengo 0.79 2.15 2.75 4.01 2.95 6.51 Benguela 9.86 15.88 14.41 20.29 17.68 32.98 Bié 2.42 5.19 6.83 10.51 5.87 14.85 Cabinda 3.87 5.69 5.56 5.93 5.97 24.11 Huambo 3.90 8.67 11.42 11.30 14.86 24.96 Huíla 5.25 8.80 8.61 10.74 14.39 18.89 Kuando 1.11 2.65 4.03 6.26 2.67 5.47 Kubango Kunene 1.13 3.89 3.19 4.04 2.83 8.85 Kwanza Norte 1.04 2.60 3.00 4.63 4.78 6.78 Kwanza Sul 1.86 4.69 6.59 7.26 8.99 17.14 Luanda 52.07 28.20 26.60 34.26 34.50 52.06 Lunda Norte 1.79 3.67 4.78 5.77 3.37 7.78 Lunda Sul 1.30 2.59 2.22 3.35 5.24 5.71 Malanje 1.38 3.70 4.47 4.40 3.97 6.29 Moxico 1.61 3.72 3.39 5.32 4.23 5.75 Namibe 1.99 5.28 4.80 4.14 5.46 8.23 Uíge 1.61 2.72 3.58 3.74 5.30 8.03 Zaire 1.66 2.66 2.67 4.30 4.73 4.80 Total 94.63 112.75 118.90 150.25 147.78 259.20 Provinces Central 84.38 150.82 94.21 148.33 166.02 188.05 Structure General Total 179.01 263.57 213.11 298.58 313.80 447.25 Source: Ministry of Health. 226. In the years 2001 and 2004 central spending exceeded that of the provinces, and in the other years they represented 82 percent of these, on average (Table 3.7). - 93 - The proportion of central level spending is large because the central level is responsible for the management of investments in national and provincial hospitals, the control of the National Health Board, and the national hospitals, which have a high running cost. Like in other developing countries, national reference hospitals tend to be better financed, because of the pressure on the government from urban elites for modern hospital care. 227. This geographic inequality would not have happened if the allocation per province had followed criteria related to the size of the population. Although provinces have been managing increasing resources, especially for investments in primary health care, the distribution of resources by the MOF is still made on an ad hoc basis. For example, when the price of oil increased in 2005, each province received an additional amount of US$20 million for investment, regardless of their needs or relative size. This lack of objective criteria favors provinces with a smaller population, and could perpetuate an uneven geographic growth. 228. The unequal distribution of financial resources is also partially caused by the imbalanced distribution of health workers throughout the country, because personnel expenditures represent a high portion of the total resources for the sector. As an example, more than half the doctors working for the NHS are in Luanda. For nurses, the number varies between 9 per every 10,000 inhabitants in Uige and 50 per 10,000 in Namibe (Table 3.8). This is partly the result of the civil war that caused a large migration of health personnel to Luanda and partly because of the different priority that provinces assign to health. - 94 - Table 3. 8: Number of doctors and nurses in the NHS per province, 2005 Doctors/ Nurses/ Provinces Doctors Nurses 10,000 10,000 inhab. inhab. Bengo 30 850 1.50 43 Benguela 65 2.663 0.30 12 Bié 37 1.350 0.36 13 Cabinda 42 1.173 1.00 28 Huambo 48 1.436 0.38 11 Huíla 63 1.688 0.45 12 Kuando Kubango 14 544 0.36 14 Kunene 34 785 0.88 20 Kwanza Norte 28 977 0.78 27 Kwanza Sul 62 992 0.63 10 Luanda 802 7.968 2.03 20 Lunda Norte 38 789 0.74 15 Lunda Sul 40 696 1.70 30 Malanje 32 939 0.75 22 Moxico 27 1.135 0.62 26 Namibe 31 883 1.76 50 Uíge 33 932 0.32 9 Zaire 32 686 1.46 31 Angola 1,458 26,486 0.94 17 Source: DNRH, Ministry of Health 229. Angola has 0.9 doctor per 10,000 inhabitants compared to 2.4 in Botswana (2002), 1.3 in Zimbabwe (2003), and 0.2 in Mozambique (2000). The concentration of health service provision in the provincial capitals and coastal provinces is understandable because it results from the war, but this situation needs to be corrected. As part of the next review of the Human Resources Development Plan, the MOH should draw up a strategy to reverse this inequitable distribution, notably by creating incentives for the settlement of cadres in the peripheral and disadvantaged areas. 230. Weak classification of expenditures blurs monitoring of spending per program. First, it is striking to note the discontinuity of some important programs during the period 2000-05 (Table 3.9). The malaria program, started in 2001, appears to stop in 2003. This corresponds to a change of definition in the registry of programs. A second problem is the change of names. The general administration program was later renamed health policy management. It includes, among other things, the salaries of all permanent employees in the sector, regardless of their level. The name of the program is not adequate given that the function of the staff, largely technicians, is not health policy management, but direct provision of services. The third problem is the unreliability of the spending classification for each program. For example, the public health program corresponds only to resources managed at the central level, leaving out the spending made at provincial and municipal levels. - 95 - 231. Table 3.9 also shows that the overwhelming majority of spending is made on two programs: general administration/health policy management, and medical assistance/health coverage. Included are expenditures related to the provision of services, staff salaries and current spending of health units. These programs represent 63-79 percent of the sector's total spending. Table 3. 9: Health sector spending per program (US$ `000) Programs 2000 2001 2002 2003 2004 2005 General administration/ Health 43,966 117,454 125,146 145,793 206,913 policy management Medical assistance/ health coverage 77,725 68,704 167,798 78,038 55,897 75,560 Other public health programs 5,325 7,018 5,881 16,044 16,418 18,069 Fight against major endemic 14,415 11,720 8,981 diseases Malaria 3,449 1,169 TB and Leprosy 1,677 427 HIV/AIDS 2,599 287 2,171 1,937 11,227 Immunization 1,106 703 1,499 2,076 2,928 Drugs 12,556 4,091 8,972 15,275 18,163 Epidemiological surveillance 3,558 1,228 2,645 1,490 1,027 Medical assistance abroad 16,647 12,339 12,067 14,173 13,540 Public investment 17,904 10,754 24,065 44,208 77,480 Ministry of Health 6,878 3,175 2,966 Higher education 3,063 4,294 2,846 5,423 2,173 5,490 Sanitation 39,273 Other 2,785 3,429 2,621 8,094 2,640 7,867 TOTAL 179,014 263,570 213,110 298,579 313,799 447,245 Source: MINFIN/MOH. Note:Tthe programs defined by SIGFE do not coincide with the existing programs of the National Directorate of Public Health 232. During the preparation of the 2006 budget, the MOF attempted to link the programs that appear in the budget with the government's overall plan. This initiative, however, did not improve the situation as the majority of resources were allocated to routine program activities, with changes only in programs that could be monitored, e.g. those related to the fight against endemic diseases, managed at the central level. The government will need to undertake further work on program classifiers to allow for a proper monitoring of activities, in line with MOF instruments. E. HEALTH FINANCING 233. Health expenditures have remained stable as a proportion of total expenditures and GDP. Since 2001, health expenditures have remained at 2-3 percent of GDP, which is considered low by international standards. As a proportion of total expenditures, health expenditures have been stable at around 4-5 percent or the equivalent to some US$213 million in 2002 and US$447 million in 2005 (Table 3.10). This could be an indication that the health sector is not capable of absorbing more resources because it is still recovering from the damages of the war. - 96 - 234. In 2005 per capita expenditures rose to US$28.8, close to the region's average. In 2006 the health budget represented US$71 per capita, substantially above the majority of the countries in the region, and above the basic health package calculated at US$37 by the Commission of Macroeconomics and Health. Although other sources13 estimate at US$ 75-120 the cost of a universal package of basic services, it is clear that Angola has the means to provide the resources necessary to finance a package of basic services for the entire population, without having to resort to financing from donors, or user fees. Table 3. 10: GDP, Total public expenditures and health expenditures (in US$ million and as a proportion). 2001 2002 2003 2004 2005 2006a Nominal GDP 9,474 11,204 13,826.0 19,915 23,226 30,051 Total Public Expenditures 4,386 5,401 6,141 7,095 10,159 23,110 Health Expenditures 263.6 213.1 298.6 313.8 447.2 1,020.9 % Health in GDP 2.78 % 1.90 % 2.16 % 1.58 % 1.93 % 3.40 % % Health in Total Public Expenditures 6.01 % 3.95 % 4.86 % 4.42 % 4.40 % 4.42 % % Total of Public Expenditures in GDP 46.3 % 48.21 % 44.42 % 35.63 % 43.74 % 76.9 % Source: ECP and SIGFE; a. numbers for 2006 are budgeted numbers 235. Budget execution has remained low at about 70 percent in the last few years. In the past, this situation could have been explained by the higher priority given to other sectors such as defense, but today this level of execution is probably due to: (i) the weak absorption capacity of the health sector; (ii) the lack of transfer of financial resources to the first level; and (iii) frequent revisions of the budget during the fiscal year. There are substantial differences in the execution rates for expenditures made at the different health care levels and for different economic categories (Table 3.11). Table 3. 11: Budget and total public health expenditures per capita (US$ `000)a, 2000 2001 2002 2003 2004 2005 Budgeted 254,127 337,950 244,297 398,586 509,383 663,598 Executed 179,014 26,.570 213,110 298,579 313,799 447,245 Budget per capita 19.0 24.5 17.2 27.2 33.8 42.7 Executed per capita 13.4 19.1 15.0 20.4 20.8 28.8 Execution rate 70 % 78 % 87 % 75 % 62 % 67 % Source: MINFIN/MOH. Note: a. The exchange rates used for budgeted and executed figures: Average official and informal rate during the execution year. 236. Angola continues to have poor health outcomes compared to other Southern Africa Development Community (SADC) countries that spend less per capita. Zimbabwe spends less than Angola on health (US$14 per capita), but has a lower under-five mortality rate with 129 deaths per 1,000 live births. South Africa spends US$114 per capita and has a maternal mortality ratio of 67 per every 100,000 live births, compared with 1,700 deaths per 100,000 live births in Angola (Table 3.12). 13 Hay, 2003 (extracted from Pavignani 2005) - 97 - 237. The health sector is the social sector that has benefited the most from donor contributions, accounting for 16 percent of total external assistance. During the armed conflict, the donor community played an important role of providing emergency health services in many provinces of the country. After the end of the war, their support diminished. According to a recent study carried out by the European Union (EU), the contributions of donors reached US$59 million in 2005. In addition, UN agencies and the Global Fund are estimated to provide US$15-20 million annually, totalling some US$75-80 million. This represents about 14 percent of total public health expenditures. Over the coming years, donor contributions are expected to remain stable in absolute terms, and to diminish in relative terms, as government spending increases. The main donors in 2005 (excluding UN agencies) were the EU (10.41%), the World Bank (8.71%), the United States Agency for International Development - USAID (7.88%), and Japan (6.37%). Individual EU states accounted for most of the remaining funds. Table 3. 12: Public health expenditures and health indicators in some SADC countries percent of Deliveries public Child Maternal Public health assisted by expenditure in mortality mortality rate expenditure qualified Country health / total rate (per (per 100,000 per capita technical public 1,000) live births) (2003) personnel expenditure (2004) (2000) ( %) (2003) Angola 20 4.9 260 1700 47 (2000) Botswana 135 7.5 116 100 40.4 (2000) Mozambique 7 10.9 152 1000 48 (2003) Namibia 101 12.4 63 300 76 (2000) South Africa 114 10.2 67 230 84 (1998) Zambia 11 11.8 182 750 43 (2001/2) Zimbabwe 14 9.2 129 1100 73 (1999) Source: MOH ­ PASS (2007), p. 21. 238. Families contribute to health financing through direct payments or pre- payments (insurance), but the amounts are unknown. A study of the population's participation in health financing14 was carried out in 2005, covering a sample of 22 units. It concluded that cost recovery represents 4.1 percent of the total costs and 22.7 percent of goods and services (non-salary current expenditures). In the case of health centers, user fees represent 53.2 percent of the cost of goods and services. The law does not define how fees should be used (e.g. to improve the quality of services or working conditions). As per Table 3.13, the average expenditures incurred by households during a complete episode of a common illness such as malaria or diarrhea were about US$40. Most of the spending took place outside public health units, with drugs representing the largest share. 14 Study on user fees in health facilities of the National Health Service, MOH 2005, Luanda - 98 - Table 3. 13: Average expenditures incurred by individuals for a full episode of a common illness (in kwanzas) Category of expenditures Luanda Benguela Bié Huambo Huíla Average % Direct expenditures Drugs 1,762 1,853 715 1,536 2,202 1,709 51 % Services by private suppliers 393 64 169 327 232 275 8 % In-patient services 214 62 115 73 136 142 4 % Indirect expenditures Transport 366 710 200 329 200 382 11 % Food 1.023 661 291 270 1.483 850 25 % Average cost 3,757 3,350 1,491 2,535 4,253 3,359 100 % Source: MOH ­ GEPE ­ PASS "Study on user fees in health units of the Angolan National Health Service F. DECENTRALIZATION AND SERVICE DELIVERY 239. There are several legal instruments regulating the process of administrative decentralization, but they often contradict each other and are not implemented in practice. The MOH is responsible for national hospitals, the National Health Board and the public institutes. Provincial governments control PHDs and municipal administrations, which themselves are responsible for the Municipal Departments of Social Services, created under Decree No. 27/00, and the Municipal Departments of Social Services are accountable for health centers and posts. PHDs are responsible for municipal hospitals. Administrative decentralization, for example, is defined under decree 2/00, Decree 17/99 and Decree 27/00 related to the MOH and sponsored by the MTA. The latter decree defines the organic structure of provincial governments, municipalities, and communes. It stipulates that the Provincial Health Directorate only comprises the Department of Public Health and the Medical Assistance Department. This contradicts MOH decree 2/00, which states that PHDs cover the Department of Public Health, Human Resources, Drugs and Equipment, and municipality health departments. However, point 6 of article 20 of the organic statute of the MOH states that the organic structure of the health sector in the provinces will adjust itself to the relevant legislation on local administration. 240. Much improvement is needed to clarify the functional relations of public administration, including that of the health sector. Article 52 of Decree 27/00 establishes that the "Municipal Department of Social Services" is responsible for the health sector within the municipality, de facto eliminating municipal health departments or delegations defined in article 21 of Decree 2/00. Yet, a 2003 UNDP study shows that, three years after the approval of decree 27/00, one third of PHDs had not created Medical Assistance Departments. Municipal health delegations or department in general have not been eliminated. They live alongside the municipal departments of social services and the municipal health delegations, creating institutional and operative confusion. Furthermore, there are municipal delegations such as in Huambo, which receive direct financial allocations. - 99 - 241. Financial decentralization is determined by the SIGFE legislation and by decisions taken by the MOF in the context of the implementation of SIGFE. There are three stages associated with the process of financial decentralization: (i) The first stage took place from 1995 to 1999, during the war. Management was centralized in Luanda and within the provincial governments. The budget was incremental and not linked to outcomes. For example, the Cubal Central Hospital was not a BU and did not even have its own line item in the budget. In 1997, the total of goods and services for the Benguela PHD, which was responsible for municipal hospitals and health centers and posts of the province, was only US$247,000. (ii) The war continued until 2002, but the decentralization process started in 2000 with the transformation of national and provincial reference hospitals and gradually of some municipal health units. In 2000, the Cubal Municipal Hospital, for example, a dependent unit from the Benguela Provincial Hospital, for the first time received its own budget, equivalent to US$272,000, for goods and services. (iii) Since 2004 the decentralization process intensified, with the creation of new dependent units. Financial decentralization has gone as far as municipal hospitals or municipal health centers of reference. However, decentralization deteriorated in terms of quality following the withdrawal of the BU status from 80 municipal hospitals, which had received it in 2002. Table 3. 14: Proportion of spending in the provinces, Luanda, and central units 2002 2003 2004 2005 BUDGET ­ Total budget Provinces 43.3 38.8 36.1 46.3 Luanda 12.5 11.5 11.0 11.6 Central Units 44.2 49.7 52.9 42.0 Total 100.0 100.0 100.0 100.0 Goods and services Provinces 36.9 23.3 18.4 34.4 Luanda 6.5 6.5 4.3 5.7 Central Units 56.6 70.2 77.4 59.9 Total 100.0 100.0 100.0 100.0 Source: SIGFE ­ MINFIN. 242. Table 3.14 presents the percentage of resources managed by the provinces, in Luanda province and in central units, compared to: (i) the total budget; and (ii) spending for goods and services. The data show the significant weight of the central units, as decentralization of resources outside Luanda is still in its early stages. The majority of resources for the sector are still managed in the budgetary units of the central units (the MOH, public institutes and central hospitals), except in 2005 when the provinces, with the exception of Luanda, managed more resources. This could be due to a substantial increase of spending on salaries following the employment of - 100 - thousands of new staff into the sector. When Luanda and the central units are taken together, it becomes evident that most resources are managed in the capital. 243. The largest share of goods and services is controlled by the central units, although provinces significantly increased their share in 2005. Some of this is explained by the voluminous budgets received by the central hospitals and the MOH, and the lower budget execution of peripheral units. The weight of goods and services for Luanda province decreased over the period. This may be due to the reduced degree of decentralization in Luanda, where municipalities such as Cacuaco or Sambizanga do not have any health unit with its own budget for goods and services. During the 2002-2005 period, provinces represented 46.2 percent of health spending, but only 13.7 percent of spending for all sectors. 244. Trough the decentralization model, the government has given the regulatory role to the MOH, but has de facto removed its responsibility for the direct provision of services. Yet, there is no legal document stating that the MOH must abandon the responsibility of delivering health services. The weakness of the model results from the absence of an explicit framework defining the role of each actor. This is shown by the inconsistency between resource allocation and responsibilities attributed. Resources for goods and services at the PHD are an example of this incoherence. In theory, PHDs are responsible for operating health centers and posts that are not DUs, but the resources allocated to them are so limited that they can only cover their administrative costs. For any additional spending (e.g. fuel for supervision), health centers and posts depend entirely on provincial governments. 245. As a result of the lack of guidelines, different approaches prevail. In some provinces, municipal hospitals take care of primary-level health units, sharing the goods and services that they have acquired, but in others they do not. This shows the urgency for the government to put more coherence in the decentralization framework. G. EFFECTIVENESS AND EFFICIENCY OF SERVICE DELIVERY 246. There are a number of interventions that can make public service delivery more effective and that can help improve the efficiency of health authorities in Angola. This section discusses a strategic approach and assesses scenarios for scaling up service coverage using the marginal budgeting for bottlenecks approach (MBB). 247. Focus on an effective package of interventions. Health spending can only achieve good results when a country knows its objectives, i.e., what are the major health problems it wants to tackle. This is not easy in the health field because there are multiple outcomes, and these are the result of multiple determinants and multiple inputs. First, health is not a single product: it produces multiple outcomes (or multiple products). Good health not only means high life expectancy, freedom from disease and disability, sufficient nutrition and normal growth, but it also refers to psychological, behavioral and productive aspects of human life. - 101 - 248. To make it more complicated, different health outcome indicators may be subject to different sets of determinants. For example, child mortality is influenced notably by maternal education, household hygiene practice, food safety, immunization and micronutrient interventions, and timely care seeking. On the other hand, maternal mortality has a somewhat different set of determinants, which are associated with institutional obstetric care, delivery facilities, and transport conditions to emergency care, notably. Even within the health sector, various health interventions contribute to different health outcomes. For example, immunization aims at improving child health, and emergency obstetric care aims at reducing maternal mortality. An illustrative example of how to link inputs to health outcomes is shown in Chart 3. Chart 3: Linking inputs to health outcomes Pathway Example Phases of Planning Health outcomes U5MR Setting up goals ? Immunization Open-heart surgery Health services Prioritizing ? Build urban Health inputs Organize rural tertiary hospitals Allocating & targeting outreach teams to provide door to door services 249. Money can be easily wasted by investing in the wrong services and less cost- effective delivery mechanisms. In the Angola context, it is obvious that high child mortality, maternal mortality, the high burden of infectious diseases, and poor environmental health need primary attention. What is difficult is to prioritize a package of health interventions which are effective. 250. In its Strategic Plan for the Accelerated Reduction of Maternal and Child Mortality in Angola (2004-2008), the MOH prioritized an essential package of interventions as its means to reduce child and maternal mortality (Boxes 3.1 to 3.3). The package is organized by service delivery mode: (i) primary health services; (ii) mobile and advance health teams; and (iii) community health agents). 251. The interventions included in the package are proven to be cost-effective and with high impact on child and maternal mortality and the package is in line with the international best-buy list of interventions. Money will be well spent to finance these interventions, if they are delivered properly. However, as discussed earlier, there is no - 102 - clear link currently between resources and objectives, and the primary care network is severely under-funded. Box 3. 1: Essential package of mother and child care and services at the primary level. Child Health Maternal Health Preventive care Ante-natal care Growth monitoring Administration of iron + folic acid Vaccination Presumptive intermittent treatment against malaria Administration of vitamin A Anti-parasite treatment Administration of iron syrup Vaccination against Tetanus Appropriate treatment of cases: Care at birth and of the new-born child Malaria Antibiotics in case of premature rupture of membranes Diarrhea Clean birth Acute Respiratory Infections Resuscitation of the new-born child Removal of parasites Administration of Nevirapine to HIV+ mothers (Maternity Counseling: wards) Identification of signs of danger to seek attention. Administration of vitamin A after the birth. Compliance of treatment of diseases. Counseling: Feeding during common illnesses. Identification of signs of danger of complications Compliance with monitoring of child growth Prevention and voluntary testing for HIV/AIDS Compliance with the vaccination scheme Use of treated mosquito nets Breast feeding Rest and feeding during pregnancy Appropriate complementary feeding Compliance with the treatments and vaccination Early stimulation Preparation for the birth Use of treated mosquito nets Breast feeding Spacing between births T f ê i d ê i b é i Box 3. 2: Essential package of mother and child care and services to be applied by mobile and advance health teams. Child Health Maternal Health Regular preventive care Ante-natal care Vaccination Identification of pregnant women at risk Administration of vitamin A Administration of iron + folic acid Removal of parasites Presumptive treatment against malaria Distribution of treated mosquito nets Removal of parasites Re-treatment of mosquito nets Vaccination against Tetanus Appropriate treatment of cases: Counseling Malaria Identification of signs of danger of complications Diarrhea Prevention of HIV/AIDS Acute Respiratory Infections Use of treated mosquito nets Rest and feeding during pregnancy Counseling: Preparation for the birth Identification of signs of danger to seek care. Breastfeeding Compliance of treatment of diseases. Spacing between births Feeding during common illnesses. Compliance with the vaccination scheme Identification of means of transport in order to refer Breastfeeding obstetric emergencies Use of treated mosquito nets - 103 - Box 3. 3: Essential package of mother and child care for community and neighbourhood voluntary health agents. Community Health Agents Community birth assistants (Midwives) Appropriate handling of cases: Ante-natal care Malaria at all ages Diarrhea in children (oral rehydration) Administration of iron + folic acid Acute Respiratory Infections in children Care at birth and of the new-born child Regular removal of parasites in children and pregnant Clean birth (clean birth kit) women Administration of vitamin A immediately after birth Regular vitamin A supplement for children Counseling: Transfer of severe cases to the health services Identification and timely organization of means of transport Identification of signs of danger of obstetric complications. Use of treated mosquito nets Counseling and promotional activities: Prevention of HIV/AIDS Identification of signs of danger to seek care. Breast feeding Compliance with treatment. Spacing between births Exclusive breastfeeding Feeding during common illnesses. Transfer of obstetric emergencies to reference hospitals Use of treated mosquito nets Use of clean water, use of latrines and personal hygiene Identification and timely organization of means of transport. Prevention of HIV/AIDS Those in charge of Mosquito Net Treatment Activists from grass roots social organizations Units (UTIs). Promotion of demand for health services: Distribution/sale of treated mosquito nets: Routine vaccination for children and pregnant women Current nets treated Care for common diseases Long lasting pre-treated nets. Ante-natal control Care at birth Regular treatment with insecticide Current nets Democratization of appropriate health knowledge and Screens practices in households Counseling and promotional activities: Exclusive breastfeeding Promotion of use of mosquito nets Feeding of small children Adequate use of treated mosquito nets Identification of signs of danger in children and pregnant Identification of signs of danger of cases of malaria to seek women care. Prevention of HIV/AIDS Promotion of malaria prophylaxis for pregnant women Use of treated mosquito nets by children and pregnant women Use of clean water, use of latrines and personal hygiene 252. Investing in effective service delivery modes. Health services can be delivered not only within health facilities, but also through outreach or mobile arrangements, and by communities themselves. Based on international experience, health services can be classified according to three types of service delivery mode: · Family or community-based care. This consists of interventions that include preventive measures and the management of maternal and childhood illnesses. They can be delivered by households or communities themselves under guidance from health professionals. Insecticide treated bed nets (ITNs) for pregnant women and children under five, condom use, breastfeeding, and oral re-hydration therapy (ORT) are some examples of interventions that are family or community based. · Population-oriented outreach services. These services are delivered to all the population, regardless of whether or not people are currently sick. They are - 104 - usually delivered through periodic outreach or scheduled clinical services. This delivery mode includes preventive care interventions, such as immunizations, antenatal care, family planning, TB treatment. · Clinical-based individual care. These activities include all types of individual curative care interventions that need to be delivered at a health facility and by a trained health care professional. They are offered in a continuous manner so that they can respond to unpredictable situations, such as a sudden illness, or the delivery of a baby. 253. Each service delivery mode requires a different set of inputs. For outreach services, the availability of vehicles, essential drugs and supplies, qualified nurses, and the participation of communities are the basic requirements to improve service coverage. The success of community-based services is based on the dissemination of health knowledge, the supply of essential materials, and the follow-up by community health workers. Meanwhile, for facility-based services, qualified health professionals, infrastructure, equipment, and drug supply are essential to ensure quality of care. 254. International experience has shown that delivering a set of essential services through community-based service and outreach arrangement are very cost-effective in reducing mortality (particularly child mortality) and controlling infectious diseases. A study of 12 countries (Benin, Burkina Faso, Ethiopia, Ghana [north], Madagascar, Mali, Mauritania, Mozambique, Niger, Nigeria, Rwanda, Senegal [west]) shows that, at a low cost, community-based and outreach services can potentially reduce child mortality by 43 percent and 16 percent, respectively. 255. While the rebuilding of health facilities will take some time and requires a large investment, Angola needs to focus on building a comprehensive service delivery system instead of just building health facilities. The development of community- based and population-based outreach service delivery should go in parallel with facility-based service delivery. And even facilities should not just be points of service provision, but serve as bases to support outreach teams and community-based services. 256. Investing in an integrated service delivery approach. While the health system is weak and being rebuilt, the delivery of some key services heavily relies on vertical arrangements in Angola. In a vertical arrangement, a department at the central level has a mandate to carry on a specific health program vertically, often on a disease base (e.g. HIV/AIDS, TB, and malaria). To that effect, it plans and organizes service delivery for that program only, vertically finances the activities, and vertically supervises and monitors the program. In Angola, the MOH national departments, especially the ones linked to disease control, are organized by diseases and are supported vertically by donors. This has created various parallel management, logistics and information systems that are generally not integrated with the overall national system. As the supervision of these programs is not done jointly, there is no benefit from economies of scale. - 105 - 257. It is important to recognize that successes have been achieved through the vertical arrangements. For example, as shown in Figure 3.7, the immunization program reached a relatively high coverage level in a short time. The HIV/AIDS, TB, and malaria programs have been able to raise significant funding from donors and have started to expand the coverage of prevention and treatment services. Figure 3. 6: Immunization rates 80 70 Immunization, DPT 60 (% of children ages 50 12-23 months) % 40 Immunization, 30 measles (% of children ages 12-23 20 months) 10 0 2000 2001 2002 2003 2004 Year Source: World Development Indicators, World Bank 2006 258. However, in the long run, vertical programs will unavoidably cause duplication, inefficiency, and competition for limited resources. Also, uncoordinated vertical programs will add an unnecessary burden on the lower levels where resources are even scarcer. Angola needs to aim at strengthening a service delivery system that is able to integrate the vertical programs. In doing so, Angola will not only deliver a package of services effectively and in a coordinated way, but also maximize the impact of health spending 259. One way to integrate the vertical programs is to link them with the three delivery modes, i.e. to divide the essential package of services into specific service delivery arrangements (see Table 3.14). Doing so helps to facilitate how the key services are integrated into the three delivery modes. Some inputs can be shared by different programs. Other are program-specific, such as malaria drugs. It is important to identify common inputs to improve efficiency. For example, a nurse, if properly trained, can possess skills to carry out both malaria and immunization tasks. Even for program-specific inputs such as drugs, integration can be sought through a common distribution system. - 106 - Table 3. 15: Key high impact health interventions by service delivery level Service Child Health Maternal Health Malaria Environmental HIV / AIDS TB Delivery Health Arrangements Breastfeeding Clean Delivery ITNs promotion Latrines Awareness - 1. Family promotion Condom general Safe Water raising through Community Safe Water Promotion population Storage peer based based health Handling and Oral Contraceptive Handling education interventions Storage promotion Promotion Anti-malarials promotion Safe Sex (including Chlorine) Supplementary (chloroquine) for (including Promotion Hand-Washing feeding for children less than Chlorine) Condoms promotion malnourished 5 and adults Hand-Washing marketing ITNs use promotion pregnant women promotion Condom children less than 5 Solid and Liquid Promotion ORT Waste Mass media Zinc treatment in management campaigns association with promotion Support to ORT orphans Advise on Complementary and supplementary Feeding Clean Delivery Temperature Management and Kangaroo care Supervision of Supervision of Supervision of Healthy homes Supervision of TB awareness 2. Population Health Promoters Health Promoters Health Promoters environment Health raising based outreach Family Planning Family Planning promotion Promoters Case services Iron and Foliate (Depo-Provera, ) Indoor Insecticide Identification Management identification supplementation Iron and Foliate spraying Management of Support and TB DOTS Tetanus Toxoid acid Surveys/HMIS mosquito care follow-up BCG, Measles, supplementation breeding places First Aid DPT3 to pregnant Indoor Insecticide Universal Vitamin A women spraying. Precautions supplementation Births planning Control of HIB vaccine and complications insects, rodents ACT anti-malarials readiness etc for children less ITNs pregnant Food safety than 5 women measures Supervised ORS Prenatal care, Education on Surveys/HMIS postnatal care prevention of Surveys/HMIS accidents and illnesses Surveys/HMIS 3. Clinical Assisted deliveries Assisted deliveries ACT UP TT of STI TB identification services Antibiotics for Antibiotics for HAART follow and DOTS pneumonia (ARI tt) premature rapture up initiation a. primary Antibiotic treatment of membrane PMTCT clinical care for dysentery (PRM) Treatment of Resuscitation Tt of STI Opportunistic Treatment of Basic to infections Severe Anemia comprehensive UP Vitamin A treatment EOC Treatment of Post-abortion care neonatal sepsis Norplant PMTCT IUD insertion Severe malaria b. referral Management of CEOC Management of Management of Management of clinical care severe Blood Safety complicated resistant AIDS Multi drug prematurity/LBW Malaria Resistant TB and neonatal sepsis Management of complicated Malaria - 107 - 260. Aligning the system function for service delivery. Delivering the package of essential services relies on a functional health system. Efforts need to be made and resources need to be allocated to align the key functions of the heath system to support and facilitate the service delivery. 261. Planning and budgeting in the context of decentralization. The planning and budgeting process is the first step to ensure that health resources are effectively allocated and then used. To deliver the defined services and achieve the intended results, the health system needs to carefully plan the activities and allocate sufficient budget to ensure its operation. This needs to happen from the central level to the service frontline in the provinces and municipalities. As discussed earlier, decentralization and the strong role played by the MOF, have left the MOH with little involvement in the planning and budgeting process. As a result, the MOH has not been engaging in a meaningful way in sector-wide planning and programming, to implement the national health policies and strategies, and to influence primary health care and local service delivery. With decreasing policy and technical guidance from the MOH, health budgets across the provinces have shown a large variation, and priority for allocating resources has not been consistent. 262. At the same time, although budget management units have been created in the provinces to facilitate fiscal decentralization, the capacity for result-oriented planning and budgeting and budget management is weak at the provincial level and even weaker at the municipal level. The weak capacity has resulted in insufficient funding for key services. With support from the EU and other partners, the MOH is trying to improve its capacity for budget preparation and management, and link the budget to a plan of action prepared by the provinces. 263. Human resources. The delivery of the defined package of services calls for sufficient and qualified health professionals and volunteers to provide care under the three service delivery arrangements: facility-based services, population-based outreach services, and community-based services. The government needs to make efforts to align resources, and to re-examine the staffing policy at each level of service delivery to establish necessary posts to meet the requirement, which should also take into consideration local conditions in the provinces. 264. Monitoring and evaluation. One of the weaknesses of the Angolan health system is the lack of timely health information and health data to guide policy- making and service delivery. Data are lacking across three levels: health outcomes, health services and health inputs. Data on service coverage date back to the 2001 MICS and are not comprehensive. Health input data from the service frontline are not readily available. To reach the country's goals to reduce child and maternal mortality and infectious disease prevalence, the country needs to undertake a thorough transformation of the M&E system. A well-designed monitoring and evaluation system should include key health indicators, facilitate timely data flow, and encourage data use to guide strategy and service delivery. All this needs efforts and resources. International experience shows that money spent on evidence-based decision-making is worthwhile. - 108 - 265. Spending to overcome bottlenecks in the service delivery chain. To deliver good-quality health services to its users, five dimensions must be considered along the service delivery chain. These are: availability, accessibility, utilization, continuity, and quality. · Availability: Availability of critical health system inputs such as drugs, vaccines, supplies, and human resources. · Accessibility: Physical access of health services to the clients. This includes the presence of trained human resources at the community level, the number of villages reached frequently by outreach services, and the time taken to reach a facility providing basic and emergency obstetric and neonatal care services. · Utilization: First use of multi-contact services, such as the first antenatal contact or BCG immunization. · Continuity of coverage (or adequate coverage). Represents the actual contacts of the target population with health services compared to optimal contacts. Applies to services that require multiple visits, such as DPT3 immunization that require three shots, or antenatal care where women should have at least three visits to a health facility during their pregnancy. · Quality of coverage (or effective coverage). Measures the quality of care by assessing the skills of health workers. Skills are assessed in terms of workers' ability to examine beneficiaries, diagnose, provide the requisite interventions, use the equipment appropriately, and advise properly. Chart 5: Sequential Relationships of Health Services 5. Quality 4. Continuity 3. Utilization 2. Accessibility 1. Availability Population eligible for services 266. As shown in Chart 5, the five determinants are sequential, following the logic of service delivery flow. First, services have to be available in a given area. Second, the service locations have to be physically accessible to users. Third, to receive health - 109 - benefits, potential users have to actually use the services. Fourth, potential beneficiaries should utilize the services with continuity, e.g. follow the complete schedule of care such as three doses of DPT. Finally, potential users should utilize the services in a correct and effective matter. 267. In a perfect world, services are delivered smoothly and effectively when key inputs are available, services are widely accessible, the quality of care is good, and the utilization rate by patients is high. In this case, the investment will achieve a high return and users will obtain maximal benefits from the services. In reality, bottlenecks can occur in any of the five dimensions to hinder service delivery. Consequently, this not only stops services from reaching their intended target, but also makes resources invested in other dimensions ineffective. So, it is imperative to understand the nature of the bottlenecks, which is not an easy task, as different service delivery modes may face different bottlenecks. 268. At present in Angola, the major bottlenecks in the community-based services include: (i) The poor availability of services. For example, as shown in Figure 3.8, it is estimated that only 20 percent of villages or communities have community health workers who could distribute mosquito nets, and the availability of bed nets is low; (ii) The low utilization of services: even for households who have bed nets, only a small percentage of pregnant women are using them; and (iii) The poor quality of the service coverage. For example, as shown in Figure 3.9, although more than 90 percent of Angolan mothers breastfeed their babies, only 15 percent of them are breastfeeding exclusively. Figure 3. 7 : ITN Use Figure 3. 8: Breastfeeding ITN Use 100% 90% 35% 80% 30% 70% 25% 60% 20% % 50% % 15% 40% 10% 30% 5% 20% 0% 10% ITN in stock %villages w ith % HH w th net % pregnant % pregnant 0% CHW using net using treated Children w ith Child ever Prolonged early Exclusive net mother breastfed breastfeeding breastfeeding breastfeeding Source: MICS (2001). 269. For population-based outreach services, bottlenecks exist in the continued use of the services. For example, as shown in Figure 3.10, more than 60 percent of - 110 - pregnant women had antenatal visits, but less than 50 percent completed the full course of antenatal care. The same evidence can be found in immunization. As shown in Figure 3.11, although more than 70 percent of children receive at least one dose of DPT, less than 50 percent are fully immunized. Figure 3. 9: Antenatal care Figure 3. 10: Immunization 90% 100% 80% 90% 80% 70% 70% 60% 60% 50% % 50% % 40% 40% 30% 30% 20% 20% 10% 10% 0% PHC stock access to DPT1 Fully immunized Fully immunized 0% vaccines immunization 12-23 months 12-23 months & SHC IFA stock access to ANC ANC1+ TT2+ BPcorrect safe inj. Source: MICS (2001). 270. For facility-based care, as shown in Figure 3.12, availability and accessibility are the major bottlenecks. Only 30 percent of households have access to institutional delivery services (i.e. deliveries in hospitals) and only 29 percent of deliveries are assisted by health professionals. At the same time, quality is also a concern in facility- based care, as only 20 percent of assisted deliveries were backed up by life-saving capacity. Figure 3. 11: Delivery services 31% 30% 29% % 28% 27% 26% 25% 24% Availability Access to daily assisted PNC < 6 days ass.del. with delivery delivery deliveries life saving skills equipment services Source: MICS (2001). 271. Different bottlenecks require different solutions. For example, solving the problem of low availability and accessibility requires an expansion of health services. However, low utilization and poor quality need to be addressed differently. Carefully - 111 - studying the nature of bottlenecks can help target resources effectively and ensure the smooth operation of service delivery. 272. The bottleneck analysis reflects the complex situation in Angola, where the lack of services is a major problem in rural settings. In Luanda and the provincial capitals, where services are concentrated, the quality of services is a challenge. Health spending can only be effective if it targets specific problems. Scenarios for Scaling Up Service Coverage 273. The government plans to significantly scale up the coverage of the services included in the defined essential package. Figure 3.14 shows the planned marginal increase in service coverage for the next three years according to the government's strategy for reducing child and maternal mortality. However, the government's plan does not specify the cost implied, nor the impact to be achieved on child and maternal mortality. Figure 3. 12: Planned Marginal Increase of Service Coverage for the Next Three Years Number of health services that undertake prevention of mother-to-child transmission of HIV. Number of health services that provide HIV and syphilis counselling and voluntary testing. Percentage of use of modern contraceptives among women of child bearing age Percentage of births attended by trained staff Percentage of births in health units Number of provinces with a reference system established for obstetric emergencies. Percentage of households consuming iodised salt Percentage of children aged 6-24 months with appropriate complementary feeding Percentage of children under six months old who are exclusively breast fed Percentage of municipalities with DTP-3 coverage of 80%or more Percentage of pregnant women vaccinated with TT- two or more doses Percentage of children under one vaccinated (DTP-3, Polio-3, measles, BCG and yellow fever). Percentage of pneumonia cases in under fives receiving standardised treatment. Percentage of diarrhoea cases in under fives receiving ORT and continued feeding. Percentage of malaria cases in under fives receiving standardised treatment. 0 20 40 60 80 100 120 % of marginal increase Source: Strategic Plan for the Accelerated Reduction of Maternal and Child Mortality in Angola (2004-2008) 274. The World Bank and UNICEF have conducted a simulation exercise to generate evidence on the steps needed and resources required to reach these targets. The exercise recommends the adoption of five steps to scale up service coverage: - 112 - · Step 1: Undertake social mobilization and behavioral interventions as well as supply essential materials to households through community-based interventions; · Step 2: Organize outreach and mobile teams to provide a set of standardized services to populations without access to health facilities; · Step 3: Expand the primary health care network to provide preventive and basic curative care; · Step 4: Strengthen the first level referral care that can provide comprehensive and emergency health care; · Step 5: Improve the second-level referral care that can provide specialized care. 275. Step 1 aims at reducing child mortality, controlling communicable diseases, and improving environment health. To that effect, it is essential to scale up community-based services (such as breast feeding, bed net use, etc.). The implementation of this step requires a sufficient number of community health volunteers and health promoters to actively disseminate health knowledge and promote healthy behaviors. It is estimated that a ratio of 1:1000 (volunteer to number of people served) is necessary in order to carry out the required workload. Other inputs required include training, incentives, and essential materials (such as mosquito nets). Table 3. 16: Cost and Impact of Scaling Up Service Coverage Reduction in Reduction in Reduction in Cost (US$ per IMR U5MR MMR capita per year) Step 1: Undertake community-based social 29% 39% 1% 2.51 mobilization and behavioral interventions Step 2: Scale up population-based outreach 9% 8% 9% 1.05 services Step 3: Expand primary health care 17% 23% 1% 3.05 Step 4: Strengthen the first level referral care 2% 2% 3% 0.97 Step 5: Improve the second level referral care 1% 1% 3% 0.89 Total Cost of All Five Steps 51% 62% 17% 8.48 Source: World Bank staff calculations. 276. Although its health infrastructure was damaged during the war, Angola still has an important community volunteer base and many NGOs are present in communities. Therefore, the development of community health interventions can be started relatively quickly as it is not conditional on the rebuilding of health infrastructure. The challenge will be how to establish a sustainable incentive mechanism to motivate community health workers. Many countries have issues on the proper form and amount of the incentives, and in the case of Angola this will need to be decided as a function of the local context. The simulation result indicates that the additional average annual cost for this step would be $2.51 per capita, and if properly implemented, it would potentially reduce infant mortality by 29 percent and under-five child mortality by 39 percent. - 113 - 277. Step 2 includes efforts to take a package of highly effective and standardized services (such as immunization and antenatal care) outside health facilities and bring it to the households and communities. This is particularly important in Angola, as 60 percent of the population does not have a reasonable access to health services. The country has achieved a significant success in delivering immunization services in communities, and this effort should be expanded to integrate other interventions and deliver a broader package. This step requires putting together an outreach team with at least two qualified nurses or other types of health professionals to visit communities totaling about 5,000 people in a timely manner. To ensure quality and efficiency, the outreach team needs to work together with community health workers and its support teams in health centers or health posts. Key inputs to make the outreach team functional include vehicles, essential drugs and supplies, and training. The simulation exercise indicates that this step costs an additional US $1.05 per capita per year, and it can potentially reduce infant mortality by 9 percent, under-five mortality by 8 percent, and maternal mortality by 9 percent. 278. Step 3 aims at expanding the health network so that the majority of the population has reasonable access (less than 2 hours walking distance) to basic primary health care. This step requires a major investment effort and sufficient recurrent budget to ensure that the required human resources and other inputs are in place. At this level, health provided in facilities includes both individual curative treatment and preventive services. The cost will be higher for this step as it involves construction, equipment and recurrent costs (salary, supplies). It requires additional US$ 3.05 per capita per year, and it could potentially reduce infant mortality by 17 percent, under-five mortality by 23 percent, and maternal mortality by 1 percent. 279. Step 4 aims at improving the first referral care, which is particularly crucial for complicated cases and maternal complications. Individual treatment and emergency obstetric care needs to be provided at this level, and at least one general doctor or qualified clinical technician needs to be installed. To implement this step, Angola will need to produce more physicians and deploy more physicians in municipalities. This would cost US$ 0.97 per capita per year, and could reduce infant mortality by 2 percent, under-five mortality by 2 percent, and maternal mortality by 3 percent. 280. Step 5 focuses on the delivery of comprehensive emergency obstetric care and specialized care. Specialists are needed at this level. It would cost US$ 0.89 per capita per year and could reduce infant mortality by 1 percent, under-five mortality by 1 percent, and maternal mortality by 3 percent. 281. In total, the five steps, if implemented successfully altogether, would have a significant impact on child and maternal mortality. They could altogether reduce infant mortality by 51 percent, under-five mortality by 62 percent, and maternal mortality by 17 percent. This achievement would cost Angola, on average, an additional US$ 8.48 per capita. Under these assumptions and considering a horizon of three years (as in the country's strategic plan), Angola would need to increase its budget for health by US$8.5 per capita per year to reach US$85.5 per capita in year 1, - 114 - US$94 per capita in year 2, and US$102.5 per capita in year 3. This is well within Angola's economic possibilities. 282. The simulations are by no means a prescription for the country's future health strategy to reach the MDGs. Rather they provide examples of how this evidence- based simulation exercise could help the country's policy development. Other policy scenarios, if requested by the country, can also be incorporated into future simulation exercises. In any case, Angola needs to take concrete steps to scale up the coverage of essential health services and to improve the population's health status. H. PROPOSED RECOMMENDATIONS Inter-sectoral allocation 283. In the medium term, Angola should increase public spending on health. This increase should be gradual so as to ensure an adequate absorption of resources by the sector. The government could increase the health sector share by an extra 0.5 percent annually until the regional average of 8-9 percent is achieved. It is important that this additional expenditure be well channelled, with special attention to scaling up the coverage of services included in the essential package. To maximize the impact of the additional expenditure, this PER recommends that the country consider the following steps to scale up health service coverage: Step 1: Undertake social mobilization and behavioral interventions as well as supply essential materials to households through community-based interventions. This step aims at reducing child mortality, controlling communicable diseases, and improving environment health by implementing a set of cost-effective community-based services. It takes advantage of the fact that, although the country's health infrastructure was damaged during the war, Angola still has an important community volunteer base and many NGOs are present in communities, which allows the interventions to start relatively quickly. If properly implemented, simulation exercises using the marginal budgeting for bottlenecks (MBB) instrument show that it would potentially reduce infant mortality by 29 percent and under-five child mortality by 39 percent with a relatively low cost of $2.51 per capita per year. Step 2: Organize outreach and mobile teams to provide a set of standardized services to populations without access to health facilities. This step builds on the success of the country's immunization program. It intends to ease the country's lack of access to health facilities for the majority of the population particularly the poor by taking a package of highly effective and standardized services outside health facilities and bringing it to the households and communities. This step is highly cost effective, which could potentially reduce infant mortality by 9 percent, under-five mortality by 8 percent, and maternal mortality by 9 percent, at a low cost of $1.05 per capita per year. - 115 - Step 3: Expand the primary health care network to provide preventive and basic curative care. This step aims at expanding the health network so that the majority of the population has reasonable access (less than 2 hours walking distance) to basic primary health care. It requires a major investment effort and sufficient recurrent budget to ensure that the required human resources and other inputs are in place. Although expensive, it is necessary to help the health system to recover from the damage of the war. It could potentially reduce infant mortality by 17 percent, under-five mortality by 23 percent, and maternal mortality by 1 percent, with a cost of $3.05 per capita per year. Step 4: Strengthen the first level referral care that can provide comprehensive and emergency health care. This step aims at improving the quality and accessibility of the first-level referral care particularly in dealing with complicated maternal cases and complications. It could reduce infant mortality by 2 percent, under-five mortality by 2 percent, and maternal mortality by 3 percent at an average annual cost of $0.97 per capita. Step 5: Improve the second-level referral care that can provide specialized care. This step focuses on strengthening the second-level referral hospitals in providing quality and comprehensive emergency obstetric and specialized care. It could reduce infant mortality by 1 percent, under-five mortality by 1 percent, and maternal mortality by 3 percent at a low cost of $0.89 per capita per year. Intra-sectoral allocation 284. Technical criteria for resource distribution should be introduced. Health sector allocation and expenditures are made on ad-hoc basis. This results in an inequitable and inefficient distribution of resources. The absence of technical criteria for resource allocation is recognized as a problem both at provincial and national level. For a better geographical distribution, population and poverty indicators could be used, especially if information systems and national statistics are improved in the medium term. Health personnel distribution can be another criterion. The government should also consider doing a "positive discrimination" to help the provinces that suffered the most during the war catch up with the others. 285. The following alternatives could be considered: (i) Allocate resources to the health sector, maintaining the subordination to the provincial governments. In other words, resources would be allocated for the sector at municipal and provincial level, on the basis of decisions taken by the MOF and the MOH. The provincial governments would not have the power to reallocate these resources to other sectors, although they would still control their utilization. - 116 - (ii) Substantially increase the resources to the PHDs and internally allocate them to the health units under their responsibility. This approach has been implemented by the Provincial Education Directorate of the province of Luanda, which distributes a minimum of financial resources to each school. To that effect, it would be necessary to create a specific "activity" in the budgets of PHDs with the resources earmarked for this internal redistribution. (iii) Convert PHDs into budgetary units, and the health units with in- patient capacity and maternity wards into dependent units. However, this solution clashes with the current legislation on state administration at provincial and municipal level. It could be reviewed by the MTA in consultation with the provincial governments, and other ministries involved. . The PIP should be careful not to use only population criteria, but also epidemiological data and utilization patterns. 286. The health sector needs a Strategic Investment Plan. Investment spending has increased rapidly in recent years but this rehabilitation effort has not been preceded by a needs analysis and a proper planning of the desired health network. Also, there is insufficient coordination between the central and those provincial governments and other partners. The current approach may undermine the future of the health network, making it unsustainable and inadequate for the country's needs. For that reason, investment spending should be based on a map of the existing health network and projections of different scenarios with their costs. Such a map is currently being prepared in five provinces, and this exercise should be expanded to the whole country, as quickly as possible in order to reorient the investments already planned. The map would be the backbone of the Public Investment Plan (PIP) that would be defined jointly between the MOH and the MOP, and provincial governments. It would facilitate the involvement of the MOF in the formulation of the PIP and would allow calculating recurrent costs for new investments. The PIP should be careful not to use only population criteria, but also epidemiological data and utilization patterns. The budget preparation process 287. The MOH should strengthen its capacity in strategic planning. The budget preparation process lacks strategy. The MOH does not have an explicit National Health Policy, a Medium Term Strategic Plan, nor other medium term planning instruments such as a National Investment Plan. To date fundamental aspects such as the definition of medium-term activities are undertaken by the MOF, with little involvement of the MOH. With the help of international partners, the MOH has started to redefine the National Health Policy. This policy would establish the main lines for the management of the sector, its funding, the role of the MOH, the eventual integration of vertical programs, and should set the way for a better quality of spending and improvements in the health status of the population. The policy would then be transformed into a medium-term plan, in line with the government's - 117 - Economic and Social Plan. Finally, medium-term plans would be prepared in the area of investments and human resources. 288. The MOH should build its capacity and create instruments allowing the government to monitor Angola's progress towards the Millennium Development Goals. The MOH should work with INE to ensure that surveys are periodically conducted to evaluate the health status of the population. The most important is to undertake a Demographic and Health Survey (DHS). As population data are essential for decision making in the sector, the MOH should also promote the undertaking of a population census. Financial management 289. The government should increase the provision of financial quotas to municipal hospitals and to the Provincial Health Directorates. Budget execution rates in the sector have been low, and vary with the economic category and the level of care. For example, execution rates for investments have been below average, and the execution rates for goods and services of municipal hospitals are systematically below those of hospitals. To increase execution rates, it will be necessary that the MOF effectively channel financial resources to the budgetary units, and that managers in turn make a timely transfer to the DUs. To that effect, the MOH could undertake advocacy actions towards the MOF and provincial governments or request the allocation of resources for the health sector at a level below the provincial level. The latter would clash with the current decentralization model of state power. Monitoring of budget execution 290. The efficiency of public expenditure should be increased at all levels. The cost-effectiveness of some public spending is below the optimal level. For example, the unit costs paid by different health units are very different for the same type of subcontracted services (cleaning, food, etc.). In the area of investments, the costs per square metre for the units of the same level do not follow the market logic. Efficiency improvements are possible through compliance with the legislation in force on public tendering and the establishment of indicative unit prices for key services and for infrastructure. These should be defined by the MOH, the MOP and the MOF. 291. The MOH should work with the MOF to review budget categories, within the limits established by the State General Plan of Accounts. The instruments available for financial and technical performance monitoring do not allow evaluating progress in the health sector. For example, it would be interesting for the MOH to know how much is spent on drugs and materials, which are now under the category "specialized current consumption material". It is also necessary to evaluate the usefulness of the current program classifier. A classifier per level of care (primary network, maternities, and general hospitals, specialized and central hospitals) would provide the information relevant for macro decision-making. Finally, when redesigning the Health Information System (HIS), it is essential to develop an - 118 - interface with SIGFE and SINGERH. This will allow SIGFE to provide information for the analysis of cost-effectiveness or unit costs of the facilities. Further decentralization 292. The new legislation on the state structure at provincial and municipal level should consider the specific needs of the health sector, especially basic health care provision. The inconsistency between the administrative and financial hierarchy directly results in the lack of coordination and leadership of the sector. The contradiction in the legislation, and the "disorderly abandonment" of the MOH functions without an adequate transfer of resources, have resulted in the current fragmentation of the system. The near elimination of the municipal health level, together with the integration of social sectors at this level, have created a vacuum at the level responsible for primary health care. The MOH should have a proactive role in the development of a legal framework that would determine the organization of the sector at the municipal level. It will be necessary for the government to resolve the dilemma posed by the decentralization of authority over the heath sector to the provincial governments, which clashes with the need for the central government to implement a policy of national dimension. 293. As the government reviews the National Health Policy and redefines the decentralization of local administrations, it should clearly define the role of each level. The change from a deconcentration model to the current decentralized one was not accompanied by measures to transfer responsibilities for the provision of services that used to be centralized in the MOH. For example, the public health programs of the central level still implement activities at municipal level, while the same programs at provincial level have practically no resources available. The definition of responsibilities should be followed by the transfer of corresponding financial resources, and a major effort to create capacity. This should result in the channelling of much more resources to the provincial level, especially for goods and services. The implementation of this change would take several years and should be designed accordingly. 294. It will be important to clarify the mandate of the vertical programs. The following reforms are suggested: (i) that vertical programs stop taking responsibility for procuring and distributing resources; (ii) there should be a centralized procurement for drugs and materials, as justified by the economies of scale; (iii) public health programs at provincial level should receive adequate financial resources, including for the training of the technical staff at municipal level and for supervision; and (iv) the presence of vertical programs should cease at the provincial level, and all public health activities should be integrated at the municipal and communal levels. Primary network financing 295. The government should channel adequate resources to ensure the good functioning of health centers and posts. Health centers and posts do not receive - 119 - direct financing for goods and services. This is because the PHDs, with their own budgets, are responsible for the functioning of these units. Yet, resources allocated to the PHDs are not sufficient to guarantee the functioning of the health centers and posts since: (i) 54 percent of the resources are earmarked for administrative expenditures and operation of the PHDs themselves; and (ii) the PHDs do not have capacity in procurement and logistics to attend to the needs of the network of health centers and posts. The government needs to decide on a "recipient" who is better equipped to manage the financial resources at that level, and the decision will need to fit within the government's decentralization framework. Two alternatives may be considered: (i) In the current legal framework, where the PHDs and municipal hospitals are DUs of provincial governments, the government would create an "activity" in the budget of PHDs or municipal hospitals, with resources earmarked to finance current non-salary expenditures of health centers and posts. (ii) Assuming that municipal administrations achieve the status of budgetary units with administrative and financial autonomy, municipal hospitals and "Municipal Health Directorates" could become DUs of municipal administrations. These directorates would be responsible for the functioning of the health centers and posts within the municipality. Under the current law, municipal directorates have disappeared and have been merged with directorates of other social services. Therefore they would need to recover some autonomy, taking into account the specificity of the health sector. Functions such as the procurement of essential drugs should remain at least at the provincial level, given the advantages of economies of scale. 296. The increase of resources at the primary level should be phased, in line with the increase in management capacity at this level. The increase in resources will not result in improved service delivery unless the local absorption capacity is improved. This means introducing management instruments and training managers. The success of this initiative will also depend on the presence of banks, suppliers with a tax payer number, an efficient drug market, and good logistics. Agenda for Reform 297. Annex 1 summarizes the recommendations and proposes a schedule of activity implementation for the next four years. The calendar must be consistent and part of the medium term strategic plan, which would follow the new national health policy. - 120 - Annex 1: Proposed Calendar for Implementation of Proposals Recommendations In charge 2007 2008 2009 2010 Increase the allocation for the sector by 0.5 percent Government in annually general Implementation of allocation criteria in the Budgetary Units MOH and MOF Implementation of allocation criteria for the entire sector MOH and MOF Work together with MAT to improve the financing of the primary service network, in the framework of the review of MOH and MAT the legislation on decentralization Preparation of the National Investment Plan MOH MOH and Preparation of provincial plans and investments Provincial governments Reinforcement of the management capacity of investments through the creation of a technical office that defines and MOH monitors the norms that shall govern the new health constructions Develop policy planning instruments and strategies, MOH namely PNS and PESS Implementation of the planning calendar, with the MOH utilization of the developed instruments Consign resources to the primary network through the MOH, MOF, creation of an activity managed from the PHD or municipal and PGs hospitals (both DUs of the Provincial governments) Develop management instruments for the health units MOH adequate to the Country's context Define the recommended unit costs and publication of a manual of good practices on the management of public MOH resources Review of the cadastre of natures and substitution of the MOF and MOH program classifier per level of assistance Recovery of the SIS, attempting to express the potential MOH, MOF, synergies with SIGFE and SINGERH in order to improve and MAPESS the systematic analysis of the sector Work together with MAT in the decentralization aiming to ensure a uniform implementation of the national health MOH and MAT policy all over the territory Provide response in the framework of PNS to the establishment of responsibilities at each level, with special MOH attention on the role of public health programs and the regulatory role of the MOH Define (at medium term) the most adequate "recipient" to manage current resources to finance the primary network, MOH e MOF observing the legal framework in force at each moment To immediately implement the financing of the primary THE MOH, MOF network through the creation of an activity with resources and Provincial for health centers and posts governments - 121 - CHAPTER IV PUBLIC SPENDING IN AGRICULTURE A. SECTOR BACKGROUND 298. This Chapter assesses the pattern of public spending in the agriculture sector. The analysis suggests that although agriculture is perhaps the most promising sector in Angola in terms of employment and incomes, its participation in the budget has been relatively timid as compared to other countries. Furthermore, since the majority of the poorest people are in rural areas, poverty alleviation can be more effective with rural development efforts. Given the potential of the sector and the large disparities that exist in the country, the Chapter suggests that there is space for increasing spending in agriculture in a way that is also consistent with a broader objective of exploring the country's natural wealth in a more strategic fashion. Productive potential 299. Angola has a huge potential to become a major player in agriculture and cattle breeding. This is confirmed by recent comparative studies.15 Table 4.1 below shows that Angola ranks seventh among all developing countries (excluding China) regarding food production potential. Only Brazil, Zaire, India, Indonesia, Sudan, and Argentina have more potential. As it can be seen, the capacity to produce food in Angola, taking into consideration its climate restrictions, is remarkably high. The maximum production that is possible to achieve applying better technology would allow feeding 929 million people, which shows the extremely high potential to be exploited in Angola. The contribution from irrigation accounts for only 5 percent of potential production, which is the equivalent to say that over 95 percent of potential production is to be found in dry crops. In fact, Table 4.2 suggests that Angola has 92.6 million hectares of dry cropping potential, which would rank it in the fourth place among 91 developing countries (excluding China). Finally, Table 4.3 suggests that Angola has an availability of 60.6 million of hectares available for crops 15 Data from the following studies and publications could be used to evidence such statement: 1) "CapacADIdes Potenciales de Carga Demográfica de las Tierras del Mundo en Desarrollo", FAO, United Nations Fund Population Fund (UNFPA) and International Institute for Applied Systems Analysis, Roma, 1984; 2) "Agricultura Mundial Hacia el Año 2010", A study undertaken by FAO, and lead by Nikos Alexandratos, Roma, 1995 and 3) "Agro-MAPS a Global Spatial Database of Sub- national Agricultural Land-Use Statistics", CDs N° 32 of the series Land and Water Digital Media from FAO, International Food Policy Research Institute (IFPRI), Center for Sustainability and the Global Environment (SAGE) and the International Center for Tropical Agriculture (CIAT), Roma, 2006. - 122 - (excluding pastures), and that the country is ranked in the tenth place in the world in this regard. Table 4. 1: Productive Potential ­ Angola and Selected Countries Country Potential Population Potential Population Potential Population with low inputs with medium inputs with high inputs Millions of people Millions of people Millions of people Brazil 649 2,995 7,119 Zaire 291 1,282 2,874 India 1,298 1,800 2,620 Indonesia 350 1,027 1,471 Sudan 81 259 1,055 Argentina 169 451 938 Angola 53 277 929 Colombia 142 411 902 Zambia 49 215 763 Venezuela 90 253 746 Other 107 countries 2,436 5,962 13,778 Total 117 countries 5,608 14,932 33,195 Source: "Capacidades Potenciais de Carga Demográfica de las Tierras del Mundo en Desarrollo", FAO, United Nations Population Fund (UNFPA) and International Institute for Applied Systems Analysis, Roma, 1984 Table 4. 2: Arable Land for Dry Crops ­ Angola and Selected Countries Country Arable Land for Dry Cropping Millions of hectares Brazil 582,2 India 191,5 Zaire 172,5 Angola 92,6 Argentina 86,5 Sudan 81,3 Other 85 countries 1.330,0 Total 91 countries 2.536,6 Source: "Agricultura Mundial Hacia el Año 2010", Study undertaken by FAO, and lead by Nikos Alexandratos, Roma, 1995. Table 4. 3: Arable Land ­ Angola and Selected Countries Country Arable Land* Millions of hectares Brazil 245.7 United States of America 219.4 Federation of Russian Republics 174.5 India 171.6 China 137.5 Australia 112.6 Argentina 89.6 Sudan 86.4 Zaire 66.7 Angola 60.6 Other 148 countries 1,321.8 Total 158 countries 2,686.4 Source: "Agro-MAPS a Global Spatial Database of Sub national Agricultural Land-Use Statistics", CDs N° 32 of the series Land and Water Digital Media by FAO, International Food Policy Research Institute (IFPRI), Center for Sustainability and the Global Environment (SAGE) and the International Center for Tropical Agriculture (CIAT), Roma, 2006. Note: * Area without restriction for agriculture (excluding pastures) or with moderate restrictions. Includes dry areas and irrigation, and it excludes areas with forestry potential. - 123 - Socio-geographic characteristics 300. Angola's municipalities present different conditions regarding spatial occupation, human settlement, and public agricultural attention that can be consolidated into three roughly homogeneous groups. The first one includes the majority of urban municipalities of the country and a few others that due to their reduced surface and high population levels have little land availability per capita. This group of 23 municipalities represents an area of around 3.5 million hectares, where 53 percent of the population of Angola are settled and its agricultural land availability is much reduced being estimated at around 3,000 square meters per capita (see Table No. 1 in the Appendix). Yet, and because they are close to urban centers, 13 out of the 23 municipalities have agriculture as the major activity of their population, and 8 out those 13 have commercial or market oriented agriculture. In addition, 10 among them have irrigation systems (almost 80 percent of those working with agriculture), and 10 also receive assistance from the Rural Extension and Development Program (PEDR) of the Agrarian Development Institute (ADI). The total population benefited by ADI services reaches almost 1.5 million people who represent around 16 percent of the total population, and around 37 percent of the population dedicated to agricultural work in these municipalities. 301. The second classification represented in Table 1 in the Appendix comprises 99 municipalities where over 7 million people are settled, having a theoretical average land availability of 7.9 hectares per capita. This average area coincides with the traditions, technological and working capacities of rural families that usually cultivate areas of less than 2 or 3 hectares per family group. They also have herds that occupy the same area. Of these 99 municipalities, 95 percent of them work with subsistence agriculture. In as much as 22 of the 89 municipalities have agriculture as their major activity have their own irrigation systems (25 percent of the total), and 60 receive assistance from ADI. The beneficiary population of ADI services reaches almost 2.7 million people, representing approximately 37 percent of the total population and around 54 percent of the population working in agriculture in these municipalities. 302. The third set is made up of 42 municipalities where land availability per capita averages 30 hectares. These 42 municipalities occupy just over 50 percent of Angola's total area, while their population is slightly over 900,000 people. They are the municipalities that show the largest potential to expand the agriculture frontier of the country. In 70 percent of them (30 municipalities), agriculture is the major activity - 9 have cattle-breeding vocation, and only 3 of them dedicate preferentially to fisheries. However, none of these 30 municipalities exclusively dedicated to agriculture are inclined to market production. Only 3 of them benefit from irrigation, and ADI presence is limited to 9 of the 42 municipalities that encompass this third set. The population that benefits from ADI services is very reduced and only reaches 96,000 people; they represent approximately 11 percent of the total population and encompass around 43 percent of the population dedicated to agriculture in these municipalities. - 124 - 303. In the absence of official censuses, authorities of the Ministry of Agriculture and Rural Development (MINADER) along with international and domestic technicians who are interested in agriculture activities, consensually accept that the rural population of Angola would be constituted of approximately 1.8 million families and 9 million people. According to data provided by ADI, it assists 800,000 families and 4.2 million people living in approximately 6,000 villages out of a total of 12,000 to 13,000 villages of the country. This means that ADI coverage only reaches 45 percent of families working with agriculture. 304. There are no detailed data on the entrepreneurial or commercial agriculture sub sector. Only an estimate is known about the land area granted through concessions of the State, which owns all land in the country. In the few cases of rural communities or villages having concessions from the State, land is a community right and its distribution is regulated by the traditional authorities of the group. According to official data provided, the National Rural Ordaining Directorate and the Rural Development Directorate of MINADER have granted, since 1992 and until the promulgation of the current Land Law of October, 2004, a total of 1,151 lots with an approximate area of 4 million hectares. This total corresponds to the use of concessions of lots with more than 1,000 hectares (including areas cultivating coffee), and those of irrigation perimeters. They had also granted concessions of lots with less than 1,000 hectares to 7,735 medium and small rural enterprises as well as to almost 2.8 million rural families. The total area of land granted could reach 8.6 million hectares. Cultivated area, production and yielding 305. According to the Special Reports prepared by the Food Security Cabinet of MINADER in cooperation with FAO and PMA, rural production in Angola grew by 109 percent between 1999 and 2005 (Table 4.4). Basically, this growth is due to an increase in cultivated area ­ of the 1.7 million of hectares in the 1999/0 harvest to 3.1 million of hectares for the period 2005/06 while the general average yielding reached a high of 18 percent. - 125 - Table 4. 4: Evolution of Rural Agriculture Production of Angola % of 2005/6 Product 1999/ 2000/ 2001/ 2002/ 2003/ 2004/ 2005/ Increment 00 01 02 03 04 05 06 over 1999/0 Grains Production 000 mt 500 577 546 670 721 881 745 49 Area in 000 ha 884 995 1,114 1,267 1,385 1,455 1,491 69 Yielding mt/ha 0.57 0.58 0.49 0.53 0.52 0.61 0.50 -12 Manioc, Vegetables Production 000 mt 88 87 88 128 131 175 167 90 Area in 000 ha 234 310 326 454 505 534 551 135 Yielding mt/ha 0.38 0.28 0.27 0.28 0.26 0.33 0.30 -19 Roots, Tubers Production 000 mt 4,686 5,670 6,043 6,439 7,495 9,788 10,102 116 Area in 000 ha 607 667 693 835 936 1,017 1,014 67 Yielding mt/ha 7.7 8.5 8.7 7.7 8.0 9.6 10.0 29 Total Production 000 mt 5,274 6,334 6,677 7,237 8,347 10,844 11,014 109 Area in 000 ha 1,725 1,972 2,133 2,556 2,826 3,006 3,056 77 Yielding mt/ha 3.1 3.2 3.1 2.8 3.0 3.6 3.6 18 Source: FAO/PMA Reports to the Food Security Cabinet, (www.fao.org). 306. The increment in cultivated areas was a direct consequence of the reinsertion in rural areas of families that were displaced by the civil war. The average increase in general yielding is a result of algebraic addition of two different types of results: a drop of 19 percent and of 12 percent, respectively, in yielding of grains, manioc and vegetables per hectare, and an increase of 29 percent in the roots and tubers group. According to these data, Angola could consolidate its supremacy regarding manioc yielding per hectare surpassing in more than 30 percent the average rate in the continent. On the other hand, the country lags behind in grains, beans and manioc, group for which average yielding per hectare is 50 percent to 60 percent lower than the African average. 16 Supporting services for production, trading and competitiveness 307. Subsistence agriculture depends on supporting services rendered by several public agencies and enterprises. The most important institution in this regard is the Agrarian Development Institute (ADI). The support provided by ADI which includes supplies of seeds, fertilizers, equipment and mechanized services only covers a minimal part of what is required by rural families. Besides, it is important to stress that since 2006, the supply of these inputs, tools, equipment and services are not for free and rural families have to pay back the resulting amounts with a portion of the harvested production. The only free service provided is Technical Assistance. Unfortunately, low availability of technical staff and mobility elements significantly reduce the quality and coverage of this work. 16See "Strategy Targeted to Agrarian Development: From Rehabilitation to Growth", World Bank, July 2005. - 126 - 308. The Agrarian Development Stations (EDA) and PEDR staff are responsible for ADI Extension and technical assistance. EDAs are currently located only in 84 municipalities of the country. EDA and PEDR staff comprises approximately 400 people among technicians, clerks and handworkers, what adds up to an average of 4 to 5 people per assisted municipality. As it can be seen in Table 1 in the Appendix, ADI intends to assist in 30 municipalities a rural population of more than 10,000 families in each one of them. Given the limited technical capacity and the time that it takes to train new personnel, it would be more strategic to build up EDAs in a few areas first rather than trying to do all of them at once. 309. Commercialization of the little self-consumption surpluses that rural families have can only be traded in nearby rural markets because of transportation problems. According to the "Dynamic Atlas of Angolan Municipalities", there are 550 of those markets which operate in 164 municipalities. There are no networks, either formal or informal, for stocking, conservation, packing and processing systems of small landowners' production to be supplied to distant urban centers or for exports. 310. The bad conditions of highways and, consequently, high transportation costs, is one of the main problems that slows down agriculture sector development. The World Bank report of July 2005, entitled "Strategy Targeted to Agrarian Development: from Rehabilitation to Growth", estimates that the domestic transportation cost of grains in bags varies between US$0.12 and US$0.50 per ton/km, depending on the conditions and constraints of the road in which they are supposed to go through. Another element to consider is the value of imported wheat and corn after customs in Angolan ports (including CIF plus import taxes and customs procedures expenses) that varies from US$190 to US$200 per ton. Taking these reference values into consideration, it becomes clear that the competitiveness of national wheat and corn production to fulfill the demand from large coastal urban centers is quite reduced. B. THE BUDGET FOR AGRICULTURE Importance of the State in agriculture 311. Table 3 in the Appendix shows the evolution that major variables of the agriculture sector budget in current and constant values had during 2003-2005, and their comparison with the sectoral GDP. The importance of the state in the sector of Agriculture, Silviculture and Fisheries as measured by budgetary expenditures and investments as a share of the sector's GDP is low but it has shown a strong growth during the last few years. This is because budget allocations during the exercises of 2003, 2004 and 2005 have been equivalent to 1.72 percent, 4.51 percent and 6,3 percent of the sectoral GDP, respectively. Public investments in the sector increased from 0.66 percent of GDP in 2003 to 0.79 percent in 2004, to 3 percent in 2005. 312. Despite the fact that public investments targeted to Agriculture, Silviculture and Fisheries increased in relation to sectoral GDP by almost fourfold from 2003 to 2005, this does not mean that its current level is suitable. Public investment in - 127 - Agriculture, Silviculture and Fisheries is the equivalent to 88 percent of total investment in the sector, while private investment takes the remaining 12 percent. 313. Total expenditures targeted to the agriculture sector expressed in constant values of 2005 have increased by 506.6 percent between 2003 and 2005 (Table 3 in the Appendix). This is due to growth in the `Public Investment' item that increased 658.6 percent in the period, while the `Personnel Expenditure' item increased over the same period by 415.2 percent, and the `Goods and Current Services' item increased by 411.9 percent. The share of total expenditures for the sector of Agriculture, Silviculture and Fisheries in the Global National Budget increased by 241 percent between 2003 and 2005, increasing from 0.39 percent of the public sector's total resources in 2003 to 1.33 percent in 2005. The increase in the share of the sector in the Public Investment Program has been even more significant (1,602 percent) jumping from 0.30 percent of total investments in 2003 up to 5.12 percent in 2005. 314. As noted in Table 4.5 below, expenditures with personnel (wages, social contributions and subsidies) represent 18 percent to 29 percent of total expenditures of the State targeted to Agriculture, Silviculture and Fisheries while expenditures with goods and services account for 39 percent to 53 percent, and investments represent between 18 percent and 48 percent. The period average is of 23 percent for personnel, 44 percent for goods and services, and 33 percent for investment. Table 4. 5: Economic Classification of Expenditures in Agriculture ­ 2003 - 2005 Nature of expenditure 2003 2004 2005 Total Expenditures Amount (in billions of current kwanzas) 1.536 7.245 14.467 % over total expenditures 100 100 100 Personnel Expenditures Amount (in billions of current kwanzas) 318 2.128 2.544 % over total expenditures 21 29 18 Goods and Services Expenditures Amount (in billions of current kwanzas) 633 3.855 5.572 % over total expenditures 41 53 39 Public Investment Amount (in billions of current kwanzas) 585 1.262 6.891 % over total expenditures 38 18 48 Source: Ministry of Finance (see: www.minfin.gv.ao). Concentration of budget execution 315. Table No. 4 in the Appendix shows that executed expenditures and investments in Agriculture, Silviculture and Fisheries in 2005, the only year for which such details exist, totaled 16,005.3 million of kwanzas. Six Budgetary Units that coordinate and control 31 Depending Agencies implemented this total amount. Those six Budgetary Units were: Ministry of Agriculture and Rural Development, Ministry of Fisheries, Ministry of Urbanism and Environment, Agrarian Development Support Fund, Coffee Development Fund, and Fund to Support Fishing Industry. 316. The Ministry of Agriculture and Rural Development is the most important Budgetary Unit of the Agriculture, Silviculture and Fisheries Function as it implements 74.5 percent of the sector's total budget, and coordinates the operation of - 128 - 22 Depending Agencies. The Ministry of Fisheries is the second in importance with 10.8 percent of the total budget along with its 5 Depending Agencies, followed by the Coffee Development Fund with 7.5 percent. Consequently, these three Budgetary Units implemented 93 percent of total public expenditures and investments targeted to Agriculture, Silviculture and Fisheries Function, and also control and coordinate 27 of the 31 Depending Agencies within this Function. 317. As it can be noted in Table No. 5 in the Appendix 85 percent of the budget is implemented by 6 central agencies, 6 percent by 25 deconcentrated delegations of the national ministries, and 9 percent by 6 deconcentrated institutes. Consequently, it is correct to affirm that the execution of the budget for the sector Agriculture, Silviculture and Fisheries is very centralized. Programs 318. Table 4.6 below shows the Programs that were implemented in 2005 in the Agriculture, Silviculture and Fisheries Sectors. The Program for the Relaunching of Agricultural Production included 5,457 million of kwanzas targeted to investments undertaken by the Ministry of Agriculture and Rural Development. The Agriculture Campaign Program, on its turn, is composed by funds managed by ADI to support 800,000 rural families with technical assistance, and provision of inputs, tools and equipment. Average investment per Program has been reduced since it reached only the sum of 4,900 current kwanzas, approximately US$ 56 per family or US$31 per hectare in current terms. 319. The Program for Relaunching the Coffee Production, supports the rehabilitation of this crop. The National Coffee Institute-(INCA) is in charge of its re- habilitation. Assuming that this program will encompass all the land that once was occupied by this crop ­ it is estimated that such area is in excess of 300,000 hectares ­ public support to rehabilitate the coffee sector is now currently equivalent to only US$ 45 per hectare. - 129 - Table 4. 6: Implemented Programs in Agriculture, Fisheries and Environment Function - 2005 Program Millions of kwanzas Re-launching of Agricultural Production 6,827 Agricultural Campaign 3,951 Fisheries Development 2,070 Re-launching of Coffee Production 1,169 Development of Productive Activities 489 General Administration 416 Infrastructure Rehabilitation 354 Re-launching of Agricultural Production 192 Other Several Programs 186 Intervention in Rural Ambiance 132 Forest Preservation and Exploitation 123 Re-launching of Agricultural Production 86 Environmental Preservation 9 Development of Seawaters Research and Inspection 1 Extension and Rural Development 0 Agriculture Policy Management 0 Total 16,005 Source: "Budgetary, Financial and Patrimonial Report of the State - 2005", Ministry of Finance. 320. Despite the previous estimate of 4,900 kwanzas as support for each beneficiary family of the Agriculture Campaign Program implemented by ADI, actual support seems to be much lower than that. The detailed data of tables named as "Simplified Summary of the Execution by Expenditure and per Activity", part of the "Budgetary, Financial and Patrimonial Report of the State ­ 2005 financial exercise" of the Ministry of Finance, indicate that major Agriculture Campaign Program Activities directly targeted to rural families included the following: · EDAs technical Assistance (wages and transportation expenses), 1,278 million of kwanzas; · Procuring of working tools, 818 million de kwanzas; · Procuring of animals for traction, 408 million of kwanzas; · Procuring of fertilizers, 65 million of kwanzas; · Procuring of plows, 63 million of kwanzas; · Mechanic agriculture, 25 million of kwanzas. 321. The amount of funds targeted to these six items, which are the major aiming at the 800,000 rural families assisted by ADI, was of 2,657 million of kwanzas, meaning a real support of 3,300 kwanzas per family (US$ 38); this is equivalent to 1,840 kwanzas per cultivated hectare (US$ 21). Approximately, 50 percent of the support consisted in EDA and PDER staff technical support, while the remaining 50 percent were targeted for providing inputs, services, working tools, and equipment. 322. The implemented amount for EDAs Technical Assistance item (1,278 million of kwanzas), represents an annual average cost per technician and/or worker of 3,200,000 kwanzas (US$ 36,700). This average value seems to be somewhat - 130 - overstated, as its utilization should only cover wages and transportation expenses for each of the approximately EDA and PDER 400 technicians and workers. Also, the amount of 408 million of kwanzas targeted to procure 250 oxen pairs to be distributed among some of the rural families seems disproportionate. The average resulting value for each pair would be of 1,630,000 kwanzas, the equivalent to approximately US$ 18,700. Public Investment Projects 323. There is not enough detailed information to analyze the evolution of Public Investment Projects by the three ministries that intervene in the Agriculture, Silviculture and Fisheries sector. Unfortunately, this due to the fact that detailed tables, which until 2003 were part of the "Balance of Government Program" of the Ministry of Planning, were left out in 2004. Only recently in 2005 they were again included in the "Budgetary, Financial and Patrimonial Report of State ­ 2005 financial exercise" of the Ministry of Finance. 324. In Table No.6 in the Appendix, 68 agriculture projects are detailed by importance. They are part of the current portfolio of investment projects of the Ministry of Agriculture and Rural Development, plus an agriculture development project ("Aldea Nova") financed by the Government of Israel, while the Ministry of Public Works is in charge of its implementation. These 69 projects amount to a total of 47.5 billion of kwanzas, or the equivalent to US$546.1 million. Until 2005, 16.2 billion of kwanzas were executed, representing an average increase of 34 percent over 2004. If this implementation pace is kept, completion of the 59 projects will require 14 more years. 325. Data presented in Table No. 5 of the Appendix show a large concentration and a high dependence of these projects on foreign financing. In effect, the major 5 projects (all with foreign financing) out of the 57 under implementation represent 83 percent of the total executed from 1999 until 2005, while the total for the 13 projects with foreign financing represent 91 percent of the total executed. As it can be seen in Table 4.7, 57 of the 69 projects in the portfolio have their source of funds assigned, while the other 12 remain unfunded. Table 4. 7: Source of Financing for Agricultural Projects Type of Project Quantity Total Cost ­ millions of kz. % Implemented ­ millions % kz. Foreign Source 13 24,628 52 14,759 91 Domestic Source 44 9,275 20 1.,422 9 No Source 12 13,607 28 0 0 Total 69 47,510 100 16.181 100 Source: Table No. 6 in the Appendix. 326. A high dependence on foreign financing is also evident when the level of project execution is analyzed. The 13 projects with foreign financing have an execution average that represents 60 percent of the total cost, while the 44 projects that are financed with domestic funds have an execution rate of just 15 percent. - 131 - Spatial Distribution of Agriculture Investment Projects 327. Out of the 69 projects comprising the agriculture PIP, 51 are being executed at the provincial level, while 18 have national coverage. These 51 provincial projects are located in 14 provinces. The provinces of Benguela, Bie, Lunda Norte and Zaire do not have any projects in the agriculture PIP. Table No. 6 in Appendix and Table 4.8 below also highlight the high spatial concentration observed in PIP projects. The provinces of Kwanza Sul, Huila and Bengo are responsible for 60 percent of the cost of all provincial level projects with an execution rate of 93 percent. The province of Kwanza Sul is the most privileged one as it is responsible alone for almost 70 percent of the total that has been executed at the provincial level. Table 4. 8: Distribution of Investment Projects per Province Province Quantity Total Cost ­ millions of kz. % Executed ­ millions of kz. % Bengo 8 5,000 14. 1,500 13. 2 1 Bengo y Uige 1 143 0.4 73 0.6 Cabinda 2 254 0.7 10 0.1 Cunene 3 842 2.4 46 0.4 Huambo 1 17 0.0 13 0.1 Huila 10 5,380 15. 1,198 10. 3 5 Kuando Kubango 1 1,033 2.9 29 0.3 Kwanza Norte 1 925 2.6 58 0.5 Kwanza Sul 5 10,873 30. 7,902 69. 9 2 Luanda 8 9,057 25. 253 2.2 8 Lunda Sul 1 79 0.2 1 0.0 Malange 1 104 0.3 1 0.0 Moxico 2 780 2.2 197 1.7 National 18 12,352 4,760 Namibe 5 666 1.9 140 1.2 Uige 2 5 0.0 2 0.0 Total 69 47.510 100 16.181 100 Source: Table No. 6, in the Appendix 328. The major provinces with the possibility of experiencing an expansion in their agriculture frontier are Kuando Kubango, Moxico, Lunda Norte, Lunda Sul, Namibe, Malange, Cunene, Bie and Zaire, but the level of project implementation in these provinces from 1999 until now is of just 4 percent. On the other hand, provinces that have low land availability per capita such as Kwanza Sul, Bengo and Huila are those that have received more investments during the period of 1999/2005. Thematic Distribution of Agriculture Investment Projects 329. The Agriculture PIP is also characterized by high thematic concentration. The 33 irrigation projects represent 69 percent of the total costs and have absorbed 70 percent of the funds that have been executed. Adding up the execution rates of these 33 projects with those from the 14 projects in the Mechanization and Food Security groups accounts for almost 97 percent of the funds executed so far (Table 4.9). - 132 - Table 4. 9: Thematic Distribution of Investment Projects Topic Group Quantity Total Cost­ % Executed ­ millions % millions of kz. of kz. Irrigation 33 32.755 68.9 11.343 70.1 Mechanization 1 4.889 10.3 3.418 21.1 Research 14 4.557 9.6 168.9 1.0 Food Security 13 3.644 7.7 885 5.5 Extension 1 93 0.2 61 0.4 Others 7 1.572 3.3 305 1.9 Total 69 47.510 100 16.181 100 Source: Table No. 6, in annex. 330. The notorious bias that the agriculture PIP has in favor of irrigation projects is mainly due to interest manifested by foreign financers and does not bear any correlation with the productive potential of these areas compared to dry cultivation areas. The projects in the Irrigation thematic group and those under Food Security are basically oriented to support small farmers. 331. Investments projects in the Mechanization group deserve special attention. These projects are assigned to the National Agriculture Mechanization Enterprise - MECANAGRO. This is a public corporation that provides services to ADI and to Provincial Governments on agriculture land preparation in rural units, rehabilitation of tertiary roads, irrigation and drainage systems rehabilitation, and rural edification. MECANAGRO is, from a budgetary standpoint, a Depending Agency of MINADER that also has legal responsibility. The corporation has a restructuring plan to increase the stock of its machinery and tractors to 1,500 units in comparison with the existing 200 units. Investment Projects in Fisheries 332. The fisheries PIP comprises a portfolio of 69 projects that only recently, in 2003, started to be executed. Total original cost of these 69 projects was 33 billion of kwanzas in 2003 values; its current cost is estimated to be around 45.3 billion of kwanzas of 2005 (approximately, US$520 million). The total executed until 2005 reaches 4.4 billion of kwanzas, which represents an average level of progress of 9.7 percent. If this execution pace is kept, total completion of the 69 projects will require 27 more years. 333. In 2003, only 25 out of the 69 projects were executed totaling 2,097 million of kwanzas. The most important projects executed in that exercise included: · Revitalization of EDIPESCAS, 546 million of kwanzas; · Procuring of semi-industrial boats, 417 million of kwanzas; · Salt production, 201 millions of kwanzas; · General technical assistance to the Ministry, 162 million of kwanzas; · Support to continental fishing, 131 million of kwanzas; · Improvement of the dried and smoked technology, 120 million of kwanzas; · FADI PESNORTE Project, 101 millions of kwanzas. - 133 - 334. In 2004, the execution of the fisheries PIP came to a complete halt as the records show that the total executed remaining at just 11 million of kwanzas. In 2005, investments in the sector reached a total of 1.308 billion of kwanzas that have almost exclusively been concentrated in the following three projects: · Acquisition of 150 boats of 7 to 12 meters, 372 million of kwanzas; · Acquisition of 5 shrimp boats, 348 million of kwanzas; · Construction and rehabilitation of cold storages rooms and aquariums, 585 million of kwanzas. C. ALIGNMENT OF BUDGET AND SECTORAL PLANS 335. The legal framework for interventions in the agriculture sector is described in Law No. 15 ("Basic Law for Agrarian Development") enacted on December 7, 2005. In its Article 3, the law sets the following objectives for the sector: a) Rational use of natural resources and conservation of its regeneration capacity standards. b) Increase in agriculture production, productivity and competitiveness, and improvement of the economic and social status of the rural population. c) Preservation of the socioeconomic balances of rural environment in acknowledgement of the importance of agriculture activity bestowing the important that it deserves for the development of the country. 336. Article 4 of this same law establishes that to achieve these objectives, the State should promote the following activities: a) Human resource development through professional capacity building of farmers and workers. b) Establishment of incentives for direct land exploration and maintenance of young farmers in rural areas. c) Development of irrigated, forestry and silviculture, and cattle breeding areas. d) Organization of agriculture and silviculture markets, and improvement of commercial efficiency. e) Encouragement of cooperatives and associations, and participation of producers in the agriculture policy definition, and in production transformation and commercialization. f) Support to develop activities associated to agriculture exploitation. - 134 - g) Gradual reduction of State attributions in the agriculture sector, and their progressive transfer to professional organizations. h) Development of rural research and experiments and participative streamlining. 337. The Ministry of Planning established a Global Objective for the agriculture sector after sector consultations undertaken in July 2006. The Ministry's directives aim at promoting socioeconomic development of rural and farming communities, and increasing their welfare levels to contribute to simultaneously achieve food security for the whole population, and alleviation of absolute poverty. 338. In general terms, budgetary activities of the State are consistent with the objective and the promotion of activities detailed in Law No. 15, but there are some aspects where more coherence is needed. The striking bias that sector expenditures and investments have towards small farmers and irrigated areas is not the best way to promote the rational exploitation of natural resources. Differences between potential and actual production are a clear signal of the under exploitation that characterizes dry cropping areas. For example, budgetary expenditures and investments targeted to the sector do not allow for the thorough accomplishment of objective b) of Law No. 15. Efforts by the State in the last years have preferably aimed at assuring reinsertion of families that have been displaced during wartime, but they did not have success in increasing yields of grain and vegetable, products whose production do not meet the demand. 339. Distortions between expenditures and investment and the basis set by Law No.15 are most noticeable in the last three activities established in Article 4 of this norm. This is so because budgetary activity does not show a marked decision to expand the agriculture frontier of the country but rather it seems just to support small farm agriculture already settled, and with special emphasis in assisting irrigated areas, which are a minimal portion of the territory available for agricultural activities. 340. Promotion of agriculture research and extension is where lack of concordance between objectives and activities set by the Law and budgetary expenditures and investment is to be found. Given the characteristics of the Angolan economy, highly dependent on oil and with an inadequate road infrastructure, research and extension is the type of activity that should be performed by the State if agriculture is to play a role in the development of the non-oil economy. In this sense, it is necessary to scale up the allocation in the budget for these two basic activities that from 1999 to 2005 received only 1.4 percent of the total executed, whose amount was of 16,181 million of kwanzas. D. POLICY OPTIONS FOR THE FUTURE 341. The following represents a set of measures derived from the analysis above and that are recommended for the sector of agriculture: - 135 - 1) Sector objectives should be revised to make them more focused and less global. Objectives should be more specific and with yearly quantitative goals, and the priority should be the development of market agriculture in dry areas. A specific objective should be spelled out to set deadlines to achieve self- sufficiency in grains. 2) The process for the concession of land should be reviewed to make it more agile and transparent. 3) The authorities should consider implementing a specific training program to educate technicians who could be decentralized to work on extension projects in the interior of the country. 4) A gradual policy to decentralize the provision of extension services in agriculture to provinces and municipalities should be adopted in a concerted way with a training program as described above. 5) As expenditures on agriculture are scaled up, extension services are decentralized, and the rehabilitation of transport infrastructure progresses and entails an increased circulation of goods and people to the interior of the country it will be necessary to launch HIV/AIDS education and prevention programs. It would be extremely negligent to wait until higher infection rates manifest themselves before putting in place a program to create a minimum of basic awareness of HIV facts and prevention measures. 6) Public investment should be preferably guided towards municipalities with the largest potential for the expansion of the agricultural frontier. 7) Finally, the authorities should consider adopting measures that could help the sector of agriculture cope with the overvaluation of the kwanza against the dollar. The recently published CEM presents some options in this regard. - 136 - CHAPTER V THE DECENTRALIZATION AGENDA 342. This Chapter assesses the major institutional, economic and financial factors that may influence a gradual process of decentralization in Angola. It starts with the presentation of the existing legal and institutional framework associated with the transfer of responsibilities to sub-national governments, followed by the presentation of theoretical arguments in favor of moving towards decentralization illustrating the discussion with some international experiences, and concludes with an analysis of the different phases of a possible decentralization process that could be gradually implemented by the authorities. A. THE EXISTING LEGAL AND INSTITUTIONAL FRAMEWORK The Political and Institutional Environment 343. Administratively, Angola is divided into 18 provinces, 163 municipalities, and 532 communes. The communes represent a subdivision of municipalities consisting of administrative bodies that have been decentralized from the central government in their respective areas. As there have been no elections since 1992, all sub-national governments are appointed by the central government. Some 85 percent of the total tax revenues are collected by the central government; the remainder is collected only by provincial governments since municipalities and communes are not empowered to collect their own taxes. 344. The authorities have taken the initiative to start a program of deconcentration and administrative decentralization. A new law (No. 02/07) has recently been passed and addresses some important issues including a reduction of the degree of dependence of the provinces from the central government and a larger division of responsibilities. The existing administrative division of the sate is reinforced with "neighborhoods or villages gathering into communes, the communes into municipalities, and municipalities into provinces."17 The provinces remain as major arms of the central government. Law No. 02/07 also defines the principles for the organization and operation of the government structure, foresees a gradual process of administrative deconcentration and transfer of resources to sub-national governments, and acknowledges the need for participation and collegiality. 18 17Law 02/07, chapter III, article 7 18The law contains 102 articles that regulate the responsibilities, competence and legal system for the organization and operation of sub-national governments, including provincial, municipal and communal governments. - 137 - 345. Under the new law, the organizational structure of sub-national governments encompasses collegiate and singular organizations as well as Provincial and Municipal Council for Auscultation and Social Concertation (CACS). Tthese are comprehensive advisory bodies, including provincial and municipal authorities, and also union and CSO representatives. Traditional community leaders participate in both Councils. The Municipal Council is far larger than the Provincial one, including public and private businesses, peasant associations, and churches. The idea of a decentralization process, with the participation of fundamental social agents is highlighted, and a major innovation in the Angolan legislation. 346. Additionally, municipalities were given complex responsibilities, such as the distribution, management and maintenance of water and electricity services, which are essential. The law allows for the establishment of local businesses and for a public-private partnership. The allocation of human, financial and material resources for these responsibilities, however, is an area that deserves further scrutiny. The law also shares similarities between the organizational structure of provinces and municipalities, namely: advisory support services, technical support services, instrumental support services and deconcentrated services. 347. Although remarkable improvements have been made in comparison to previous laws19, several central units deal with local issues, and the distribution of responsibilities among provinces, municipalities, and communes is not well defined and still involves the risk of concurrent competencies. For instance: · In social and cultural development at the provincial level, and in economic and social development at the municipal level, the responsibilities for social, educational and health services are of both governments; · In public safety, the protection of national and foreign citizens is a common practice, as well as the protection of public and private property; · Although there is no specific reference to the rural area at the provincial level, the vice-governor's responsibilities for economic and social sectors include agriculture and fishing. At the municipal level, there are agricultural and industrial activities linked to agricultural development stations, acquisition and distribution of farming inputs and support to producers, and licensing of agricultural and industrial units; · With regard to sanitation and rural equipment, municipalities and communes share the responsibilities for the maintenance of markets, management, cleaning and maintenance of beaches and bathing resorts, management of graveyards, and self-directed construction. 348. Concurrent competencies pose a great risk. Although this is still an incipient trend, some provinces, such as Luanda and Benguela, have already passed some responsibilities on to municipalities. Sooner or later, with the advance of the process, 19A recent UNDP/UNCDF report provides a background of previous legislation on administrative deconcentration and decentralization in Angola (see UNDP/UNCDF, 2006). - 138 - inconveniences may arise, similarly to what occurred in other countries with the same problem (e.g., Brazil). The consequences are well-known: waste of efforts and resources; difficulty in determining accountability for inadequate provision of services or lack thereof; institutional conflicts provoked by the absence of institutional boundaries between different powers. 349. In terms of dealing with the potential problem of concurrent competencies, it is important to acknowledge that the problem is not so much allocating a responsibility to more than one level (since there are usually good arguments for doing that), but that these responsibilities are defined at a very general level of aggregation, and so it becomes unclear who does what within each sector. What is needed is to unbundled the specific activities that are critical for each service and get more clarity as to which level is responsible for what specific activity. The Organization of Civil Society 350. The way in which the Angolan civil society is organized resembles other countries, regardless of them being more or less industrialized: multiple types of organizations and associations; mixed degree of institutional organization; significant heterogeneity in terms of origin, scope of action, and capacity to operate and economic sustainability. According to the Technical Unit for the Coordination of Humanitarian Assistance (UTCAH), there are nearly 300 Angolan CSOs, of which 125 are legal entities and 114 are active. Moreover, more than 100 international CSOs operate in Angola. 351. Health, education and community development are the three most common areas of operation of CSOs. These areas are followed by demining, agriculture, social services, human rights, institutional qualification, and emergencies. Although incomplete, this information provides clues to the supply of civil society services. Regionally, civil society is active mainly along the coastal stretch and in the plateau areas. There is an increasing trend towards the setup and operation of CSOs. B. CONCEPTUAL ASPECTS AND INTERNATIONAL EXPERIENCES 352. This section compiles conceptual aspects of public finances and federalism, and briefly analyses international experiences that bear resemblance to the Angolan case aimed at informing alternative paths to Angola's decentralization process. As fiscal decentralization is seen as the optimal arrival point in the reform agenda associated with the transfer of responsibilities to sub-national governments this section focuses on the underlying reasons and likely outcomes of reaching that stage. However, it is equally necessary to stress that other aspects of decentralization are important and relevant too. For example, decentralization can contribute to improve stability and conflict mitigation, improve democratic participation, and harness economic development through improved service delivery. In order to be successful, the drive towards further decentralization cannot neglect the building blocks of a - 139 - decentralized approach, i.e. moving towards more autonomous sub-national governments, creating and enforcing accountability mechanisms, and building capacity to assume increased responsibilities. Why Transfer Resources 353. Many industrialized countries are organized into and/or operate as decentralized governments with diverse institutional arrangements and different degrees of taxation and tax collection, intergovernmental relationships, processing of expenditures, and ability to contract debt. The most emblematic example is that of the European Union (EU). Such cluster of countries turned out to be a special and effective form of federation. The Maastricht Treaty, which led to the creation of the Euro, devised the principle of subsidiarity, according to which taxation, public spending, and regulatory functions must be under the responsibility of the lowest government level, unless there is a justifiable reason indicating otherwise. 354. Theoretically, several arguments may be raised in favor of decentralization and of federalist arrangements.20 In general, there is a consensus that sub-national governments are closer to citizens, allowing governments to be more aware of citizen's needs and demands, and at the same time citizens to hold the government accountable for its actions. 355. In practice, there are six reasons for fiscal transfers,21 as briefly explained below: i. To reduce or eliminate the fiscal gap ii. To correct fiscal inequity and fiscal inefficiency iii. To compensate for benefit spillovers iv. To set minimum national standards for the provision of public services v. To influence the prioritization of some activities by the local governments vi. To correct and reduce regional disparities. a) Reduction of the fiscal gap 356. The fiscal gap is defined as an imbalance between the sub-national government's capacity to raise funds and its public spending allocations, as defined by law. This imbalance may stem from three factors. First, it may occur due to an 20 For a good starting point for these arguments, see Oates (1998). 21 The six reasons described here were reproduced from a typology developed in Shah (2006). OECD (2006) reproduces a classification of Oates (1990), which identifies only three objectives for transfers: financing of services and investments of sub-national governments, subsidies and equalization. - 140 - inappropriate allocation of own fund-raising possibilities and public spending accountability. It could also be that some taxes, especially those levied on movable bases, such as wage and capital earnings, are more efficiently collected by the central government, being later allocated to sub-national governments. The fiscal gap can also result from the central government's necessity to collect more to make up for the transfers, cope with macroeconomic cycles and manage public indebtedness. For Angola, this reason is inescapable. Given the state of of urban development after years of conflict, transfers from the central government to sub-national governments will remain necessary for many years to come, and so will the redistribution of resources collected from richer regions or through dominant economic activities (e.g.: oil). b) Correction of fiscal inequity and fiscal inefficiency 357. Decentralization of expenditures may produce fiscal inequity and fiscal inefficiency. The former occurs when citizens with the same levels of income receive different levels of public goods, as they live in different cities and/or belong to different sub-national governments. This means that the provision of public services varies among different sub-national governments within the same jurisdiction for individuals who meet the same eligibility criteria. Fiscal inefficiency results from inequity, since this net fiscal benefit22 may encourage migration of people, in societies with greater mobility, to areas where this benefit is more largely available. 358. If the Constitution of a country guarantees equality for all its citizens and if the government of this country has justice and fiscal equality as objectives, then it has to transfer resources to sub-national governments so as to ensure the uniform allocation of the net fiscal benefit to the whole country. This principle of equity may be particularly appealing to Angola owing to the centralized planning orientation that prevailed after the end of the civil war. However, when the population of a country is large and income inequality is high, it is relatively difficult to organize mechanisms for the distribution of resources and to control spending in an attempt to correct inequalities either partially or entirely. c) Compensation for benefit spillovers 359. Public goods are those in which it is not possible to prevent individuals who did not pay for them from using the benefits of a service or product, either directly as consumers or indirectly through taxes. Consequently, the individual or the local government will buy or offer a less than socially desirable amount of this service. This is a classic case in the economic literature that occurs in the context of fiscal decentralization when the provision of a good or service by the municipal or provincial government produces benefits to the inhabitants of other jurisdictions. A way to address that is by transferring resources from the central government to complete the investment of sub-national governments up to a socially optimal 22 The net fiscal benefit is the difference between input benefits from public spending and the paid taxes. - 141 - spending level. However, in these cases, national taxpayers will be helping to pay for a regional service. d) Establishment of minimum national standards for the provision of public services 360. The establishment of minimum standards of quality, universality, and other parameters, by the central government for the provision of public services by the sub- national government serves a dual purpose. The first one is to promote equality. The second one is to contribute to the flow of goods, services and factors of production and not to allow sub-national governments to engage in tax competition. In effect, such transfers aim to increase trade gains in the domestic market. 361. The establishment of minimum national standards and their implementation through intergovernmental transfers is preferable to the direct provision of these services by the central government, due to matters of efficiency, preferences and accountability, as mentioned at the beginning of this section. Furthermore, this prevents sub-national governments from using the provision of low levels of public goods as a competitive edge in order to reduce their tax burden and lure a larger number of companies to their jurisdiction. e) Influence over the prioritization of some activities by sub-national governments 362. In a decentralized system, the priority for policies and expenditures usually does not coincide among the different government levels. This may cause a problem to the central government, as it is elected on the basis of a political platform with certain priorities, but whose execution, in some cases, is under the responsibility of sub-national governments whose priorities are not necessarily the same. Transfers can therefore be used to influence the prioritization of certain expenditures by sub- national governments, as more resources are made available to certain activities than to some others, and the demand for major or minor counterpart funds from recipient governments. 363. When one departs from an extremely centralized context as in Angola, not relying on a past history of organization and good administration of sub-national governments, it is recommendable to transfer the responsibility for very specific public services to local governments, where their provision is more closely related to local communities and results in reasonably more limited expenditures. By extension, it would be appropriate to set up schemes specifically targeted at the transfer of resources, by initially mobilizing marginal sources of resources and, more importantly, by applying the resources in well-specified activities without giving sub- national authorities too much leeway. - 142 - f) Reduction of Regional Disparities 364. Intergovernmental transfers can be used for policies targeted at the correction of regional disparities. These can be structural as in the case of regions of a country which have developed less than others; or they can be temporary caused by successive adverse effects, such as harvest loss, drop in export prices, etc. Regional disparities, and not only economic, or even social, religious and ethnic ones, are a predominant factor in leading a country to adopt a federalist system. In a developing economy, especially when dependence on the oil sector is large and when its economic activities are concentrated on few regions, reducing and correcting regional disparities is a constant necessity, so that fiscal policies can be developed and implemented. A Brief Account of Other Experiences 365. This section focuses on the decentralization experiences of other countries. The selection of two Latin American countries (Brazil and Colombia) and of two African countries (Malawi and Tanzania) was based on their relevance to the case of Angola. The main features of the decentralization processes in these countries are: (i) emphasis on fiscal issues, in the Brazilian case, with transfer of sectoral responsibilities (i.e., education, health) to sub-national governments; (ii) a unitarian government framework in Colombia, a "quasi-Federalist" system of government, with large initial emphasis on major laws, revised over one decade ago; (iii) decentralization implementation in Malawi and in Tanzania through a social fund mechanism that is quite similar to FAS in Angola. The common features between the cases are noteworthy: universal elections for sub-national governments; organized civil service framework; spending decisions taken by sub-national governments. a) Brazil 366. Brazil is a case of relatively high advanced fiscal, financial, and administrative decentralization, with large autonomy to sub-national governments. Angola shares many characteristics with Brazil, which include Portuguese colonization and the Portuguese language to begin with. The Brazilian federation is deeply rooted in the colonial period, in which the government used to be organized into autonomous units. Interestingly enough, the commercial, social, cultural and political relationships with Portugal were narrower than between the colony units. Angola followed a different path, of a unitarian state, despite regional and individual disparities that place it at the same level as Brazil. 367. Brazil's 1891 Constitution established a federative system in response to profound regional disparities and administrative needs, and also due its continental territorial dimension. Unlike the U.S., the Brazilian Federation did not originate from a "bottom-up" coalition: it was imposed as a unitarian State. The federative arrangement in Brazil comprises three levels of government: central, state, which includes 26 states plus the Federal District (equivalent to Angolan provinces); and local, which includes 5,564 municipalities. - 143 - 368. The 1988 Constitution promoted a vast political, administrative and fiscal decentralization. This resulted more from a political moment of re-democratization than from a technical plan. A tax reform was implemented, transferring the responsibility for the collection of some taxes to sub-national governments (e.g.: indirect taxes on fuels, energy, and communications) and increasing their participation in federal taxes23, with detailed rules provided for in the Constitution. Thus, the central government played a minor role in the decentralization of taxes, actions, and services ­ as opposed to most emerging economies. 369. The participation of Brazilian sub-national governments in public accounts, directly responsible for one third of the tax collection on a national basis, for two fifths of overall expenditures and for 35 percentage of public debts, has attested to the advanced stage of decentralization in Brazil. In the political sphere, the chief of the Executive and the members of the Legislative are directly elected in each state and each municipality for a four-year term. Furthermore, local governments enjoy the status of members of the Federal government according to the Constitution, with the same prerogatives as state governments ­ and that the central government does not virtually have any control over sub-national governments with regard to tax collection, organization and allocation of their budgets and accountability, hiring of personnel, investments, and debts. These are under the responsibility of state and municipal governments. 370. The current trend towards the strengthening of municipalities in Brazil may serve as a reference to Angola, even for the initial stage of administrative decentralization. Direct and regular elections (every four years) of the local chief of the Executive and of the members of the Legislative reinforce the political role of this sphere of government in people's lives, in each jurisdiction. The same process applies to the election of the president and state governors, respectively chief of the central and provincial Executive, as well as to the congressmen in charge of the central and provincial Legislative. The only difference lies in the interval between elections: national, state and municipal elections take place alternately every two years, always in the month of October. 371. Sub-national governments have always had autonomy to levy their own taxes (through state and municipal laws which determine the form of collection, the tax burden, and the period of collection), to collect them directly and also to decide on resource allocation, and to manage their accounts. Due to remarkable regional disparities, vertical distribution of the revenues has always occurred from the higher to the lower government levels, whereas horizontal redistribution has been reserved for less developed regions. In centralizing cycles, while the central government tries to build more solid bases and influence sub-national tax collection, the strongest control is over the reduction of automatic transfers and allocation of voluntary transfers, and over sub-national public indebtedness ­ with a secular history that 23In the case of income tax, and taxes on industrialized products, approximately half of the tax collections is transferred to state and municipal governments. - 144 - combines strong growth, also at the international level, with experiences with declaring moratoria and with debt management by the central government. b) Colombia 372. Colombia, as Angola, is organized around a unitarian regime with a long- established centralist tradition. 24 After 1969, but more strongly in the 1980s, Colombia decided to redefine the distribution of tasks and responsibilities for the provision of public services, as well as of revenue sharing and tax power between the three existing levels of government. 373. The roots of the decentralization process in Colombia differ substantially from the Brazilian process. In the case of Colombia, economic reforms served as the basis for the process, in an attempt to improve the efficiency of spending and of fiscal outcomes. The process has received a decisive support from multilateral financial organizations in the last decades. Since these organizations have also considerable importance in Angola, this is an additional reason to follow the good and bad results obtained by Colombia closely. Despite that, the financial authorities of the central government exerted full control over all stages of the decentralization process, dictating the pace and nature of the changes. 374. As a consequence of the decentralization process, the transfers from the central government to sub-national ones, which amounted to 1% of GDP in 1967, increased to 4.7% of GDP in 2002. These transfers also accounted for most of the revenue of sub-national governments, which corresponded to 34% of their revenue in 1967 and to 171.2% in 2002. 375. A more cohesive set of legal regulations that further strengthened decentralization arose in the country in the second half of the 1980s. Laws 12, of 1986, and 14, of 1983, reinforced the local governments' cash flow and the fiscal control over it as well. Law 11/1986 established the principles for the operation and modernization of municipal government. At the same time, the direct elections for mayors demanded higher accountability from local governments. 376. A new round of institutional acts related to decentralization occurred in the early 1990s. A constitutional reform in Colombia in 1991 introduced a schedule for the gradual increase of transfers, which corresponded to 46.5% of the central government's revenues by the end of the process. Law 141/1994 established that the revenues from oil royalties should be shared between the three levels of government. Law 60/1993 clearly defined the responsibilities of the three levels of government. According to this law, the responsibilities of municipalities should include: (i) preschool to high school education, including capital expenditures; (ii) health, including the maintenance of clinics, hospitals and investments in the construction of new units; (iii) electricity, telephony, water and sanitation, including expansion of 24 Until the mid-1980s, the central government used to appoint mayors and state governors. The first election for mayor took place in 1986 and the first one for governor, in 1991. - 145 - these networks; (iv) housing programs for low-income families; (v) city streets, agricultural extension programs, and rural development; and (v) environmental protection, culture, sports, and recreation. In the case of electricity, telephony, water and sanitation services, local governments could choose between providing such services directly, allowing the private sector to take them on, or allowing the state companies from other municipalities or from the state to provide them. 377. Another by-product of the Colombian decentralization process was a new division of taxing powers and own revenues between the levels of government. Currently, it is the central government's duty to collect import taxes, value added taxes, taxes on financial transactions, income tax, and taxes on gasoline and social security payments. In 2002, the revenues collected by the central government accounted for 16.1 percent of GDP and for 81.7 percent of the tax revenue. The state governments, in their turn, were in charge of collecting selective taxes on beer, alcoholic beverages and cigarettes, in addition to a local rate on gasoline added to the federal tax. 378. The states ­ intermediate government levels ­ have the lowest tax collection. In 2002, their tax revenue corresponded to 0.9% of GDP and to 4.5% of the national tax revenue. The difference between the Brazilian and Colombian cases is noteworthy. In the Colombian case, the intermediate level of government was not allowed to levy and collect a comprehensive indirect tax, as the value added tax implemented in Brazil; Colombia falls within the international standards, and this will probably be the case of Angola as well (Brazil is an exceptional case). 379. Finally, the tax revenue of Colombian municipalities amounted to 1.9% of GDP and to 13.7% of the national tax revenue in 2002. This amount was collected from taxes on property, industry and commerce, and also from an additional rate on gasoline. The Colombian success may be encouraging to Angolans, since Colombia did not use to have a tradition of politically, administratively and fiscally autonomous municipalities ­ and it is also an emerging economy with civil war problems. c) Malawi and Tanzania 380. Malawi and Tanzania have actually used social action funds (similar to the one currently being implemented in Angola) to support their ongoing decentralization process. Social Funds can play an important role in supporting decentralization. In general, the consensus is that they can be a useful instrument to pilot a more decentralized approach to development and build capacity of local governments and civil society organizations for accountability and co-provision of services. However, it is important to note that Social Funds are the correct mechanism to foster policy reforms of the inter-governmental structure needed to make decentralization work. 381. Malawi was the first to implement a SAF in 1995, as part of a process of reforms and democratization that included the implementation of a multiparty system in 1993 followed by elections in the subsequent year. The fund aimed at reducing poverty and providing the population with public services. Moreover, the fund aimed - 146 - to strengthen institutions at the three levels of government, so as to sustain the poverty reduction measures. 382. The legal framework and the financing process of the Malawi social action fund (MASAF) and the Tanzania social action fund (TASAF) are similar. One organization is in charge of receiving and approving the local government projects. In Malawi, there is a Ministry in charge of the SAF and of decentralization. After the Ministry approves the projects, local governments are entitled to the funds. The local community, consisting of 40 districts, is accountable for the projects and for their inspection. An essential feature is that local governments are democratically elected. In addition, the local levels of government include a full-time civil service staff for the management of services. 383. In Tanzania, the first stage of the TASAF began in November 2000, with financial support from the central government, local governments, and mainly from the International Development Association (IDA).25 As in Malawi, in Tanzania, the local community elects the local governments, which are accountable for the use of the funds and for the project results. The main difference between the TASAF and MASAF is that Tanzania has chosen to implement a larger decentralization, empowering the local authorities and village councils with the appraisal and approval of projects to be financed, whereas the central government was only responsible for the management of the funds. C. A PROSPECTIVE DECENTRALIZATION AGENDA 384. A readily available model of administrative and fiscal decentralization that could be recommended for Angola is out of the question. The experience of each country regarding the organization of the state and of its fiscal system reveals discrepant aspects, even in the case of countries with geographical, economic or social similarity. This section discusses three sequential phases for a gradual decentralization process in Angola. Table 5.1 shows a comparison of the major features of each phase. The proposed process starts with deconcentration of power, suggests a move towards administrative decentralization, and recommends fiscal decentralization as the final arrival point. In practice, these three phases should be seen as parts of the same scenario, with clearly defined goals, scope, and time for implementation. The third phase is only achievable in the long run, and this may take one decade or even a whole generation to be achieved. Its major feature is the concession of full autonomy to sub-national governments. Underlying the different phases is the notion that the sharing of power, resources and accountability to sub- national governments should prioritize the local sphere (municipal administrations). 25In four years of the SAF, 1,704 subprojects were financed, benefiting between 2 million and 2.7 million people, as the fund was initially implemented only in a few districts and in the two islands of Zanzibar. Education, health and water with 45.9%, 19.2% and 14.9% of the funds were the areas that most benefited from the project. - 147 - Table 5. 1: A Possible Decentralization Agenda in Angola with 3 Phases Phases of the Decentralization Process A B C Nature of Measures Specific and gradual Transition Structural changes Time frame Short term Medium term Long term Implementation First 3 years 4th to 6th year After the 6th Programs covered 3 local basic services: Extension of social All social and urban basic education, services, education, services delivered health centers, health, social locally maintenance of assistance, garbage streets and roads collection and solid waste treatment Transfers from the central government: Type Specific purpose Block grants Revenue sharing with free application Sources Budget allocation Budget allocation Functions proportional to revenues Amount Corrected past Corrected past Percentage of total expenditure (inflation expenditure plus federal tax collection and GDP) plus decreasing costing costing margin. Or margin per capita value Instrument Agreement Fund by fund General sharing fund Supervision Normal federal Initial inspection locally Auditing by budgetary process independent organization and local political decisions Institutional Plan Municipalities as Juristic entity created Totally and largely partially independent for each municipality autonomous entities government Financial plan Bank accounts Implementation of own Own administration financial and independent from the accounting services central government Own revenues Registry and follow Property tax collection Property tax and up of local tax (regulated and levied business tax national collection by the central legislation and with government) additional local rate Indebtedness Denied Denied Optional, subject to limits and control by the central government Phase 1: Short-term Administrative Decentralization 385. The first phase is characterized by the implementation of administrative changes. The proposed changes are quite specific and selective. In political terms, the first phase is indifferent about how sub-national authorities are to be chosen, whether they should be appointed or directly elected. With regard to government functions, the objective is to contemplate some few basic services ­ such as basic education, health centers, and garbage collection. The beneficiaries of decentralization will be the sub-national governments to be chosen among those already defined in Law 02/07 - 148 - as potential candidates for decentralization. An initial suggestion is that only 5 provinces and 20 municipalities be included in the first three years. This small number is due to the necessity to test and monitor closely the implemented changes. However, one may say that there are natural candidates for the selection to this first stage: Luanda, Huambo, Benguela, Huila, Cabinda and Kwanza Sul, for instance due to the presence of both governmental organizations, civil society, international and private organizations in these provinces. The Legal Framework 386. Revision of the existing legal framework is of paramount importance during the first three years of the process. In general terms, the current legal framework includes the following: · 1992 Constitution: proposition underway, still undefined · Executive Decree 80/99: Ministry of Finance (recently revised, see below) · Law of the High Authority Against Corruption: not implemented · 2004 Poverty Reduction Strategy · Long-term National Development Strategy (2005-2025) · Law 02/07 387. Other specific regulations that must be scrutinized and adapted to the decentralization process are: · 2004 Land Law and its 2006 regulation · 2005 Law on Urban Planning · Law of Associations (14/91) · Decree of Nongovernmental Organizations (84/02) 388. The revision and/or preparation of a legal apparatus that governs the decentralization process requires better use of existing fora for discussion and "concertation" in the provincial and local levels. In this regard, the strengthening and clear definition of the complementary relevant roles of some entities deserve special attention, including: · Fourth National Assembly Committee · Decentralization Work Group · IFAL · FONGA · Employers' and workers' organizations · Universities · Civil society organizations · Media - 149 - 389. A legal measure that is essential for the implementation of administrative reforms is the concession of autonomy to sub-national governments selected to manage the decentralization programs in this stage. It is advisable to give such governments legal personality, outside the current administrative framework and independent from the central government. In other words, they will not act as a budgetary and fiscal unit, but rather as an entity that receives transfers from the central government and later manages these funds, undertakes expenditures, makes disbursements and maintains accountability in a process that is separate from the current one. 390. Albeit autonomous, this entity would be extremely limited: only for the administration of decentralized programs (elementary education, health centers, social assistance, basic urban services ­ e.g., garbage collection) within the sub-national governments selected as pilot cases. The remainder of the actions undertaken by such governments will follow the current institutional routine and the same fiscal, budgetary and financial process. Even with regard to decentralized actions, they should not be authorized to hire and have their own staff in the first stage of the process. 391. The same legal revision that will grant the status of autonomous units to the selected sub-national governments and to the specified programs should also regulate the contract concerning the details of the implementation of decentralized actions. A suggestion is to create the legal entity of the agreement as an instrument to strengthen the relationship established between governments. In this case, the entity has to decide how responsibilities and funds should be transferred from the central government to sub-national ones; how sub-national governments should carry out the delegated tasks; and how sub-national governments should be accounted for the achievement of goals and acquisition of expenditures. The Political-Institutional Framework 392. The political-institutional plan should include issues related to political, financial and administrative autonomy. It is not possible to state when political autonomy will be granted to the sub-national level of government, since the recent Law 02/07 mentions that municipal administrators still have to be appointed. Administrative autonomy relies on the revision of the legal framework that regulates decentralization. Some of the aspects that have to be dealt with within the three-year period include: · Organizational structure: according to the responsibilities to be transferred to the local governments, each municipal administration will have to organize its administrative structure based on their interests and needs, with regard to service delivery and flexibility set up by Law 02/07. Some units are essential, such as planning; financial; personnel; equity; relationship with communes and local authorities, among others. The personnel unit is very important. The plan for the first stage of decentralization establishes that the personnel in - 150 - charge of the services would still be under the responsibility of the central government. · Competences: in this three-year period, municipal autonomous governments are expected to have exclusive responsibilities regarding elementary school (first to fourth grades); basic health services (health centers); urban infrastructure (roads, public lighting services, garbage collection and waste treatment, parks and gardens, markets, graveyards, and fairs); social and humanitarian assistance, after the solution to concurrent competences. All of these services are regarded as local, and the province should provide regional guidance and technical support to municipalities. Communes should be considered administrative branches of the municipal autonomous government, especially with respect to the direct relationship with the population and to the follow-up and control of service delivery. 393. In order for the decentralization process to develop in a coherent and effective manner, a National Plan for Institutional Strengthening should be devised for all government levels, civil society organizations, and private corporations. The work done by the Brazilian Institute of Municipal Administration (IBAM) in Angola in 2005 is noteworthy. Specific recommendations were provided based on an assessment of sub-national competences and needs, such as support for improved planning and management capacities; establishment of partnerships; production of knowledge; interiorization; and capacity development plan. The latter includes short, medium and long-term training courses, in addition to the qualification of trainers and on-site exposure. 394. The scope of the capacity development plan for the first three years should target: · central government organs directly related to sub-national governments: Ministry of Finance, MAT, Ministry of Planning, Ministry of Employment and Social Security, Ministry of Agriculture and Rural Development, Ministry of Urban Development and Environment and IFAL; · non-governmental and community organizations that operate with and within these municipalities; · private companies operating in the provinces and municipalities. 395. In an initial stage of the plan, the authorities should assess opportunities (potentials) and challenges (needs) of a few provinces and municipalities (the IBAM proposal mentions 5 provinces and 20 municipalities). Among several aspects, this assessment should include: systematization of existing assessments26; definition of priority areas for each provincial and municipal administration according to the responsibilities to be assumed; and the mapping of existing public and private (profitable or not) technical support services. 26Special attention should be given to IBAM (2005) study - 151 - 396. A second stage of the plan should address the compatibility between the supply and demand of technical support services in order to identify where and which limitations need to be overcome. 397. The following basic elements are suggested for the plan: · justification of the plan in the context of the initial decentralization process (first three years); · systematization of supply and demand according to the type of agent; · objectives and goals according to the type of beneficiary agent; · types of qualification and technical support activities to be offered; · estimate of expenditures and possible financial support sources; · entities eligible for service supply; · schedule for implementation of activities, making the supply of qualification compatible with the supply of technical support services; 398. Lessons from other experiences indicate that citizen participation cannot be restricted to consultations. Citizen participation is and will always be a crucial means to the advancement of any public reform and development process. Allowing for participation in the planning, decision-making, management, and monitoring of service delivery contributes to more effective results. In this regard, the legally defined CACS should be strengthened to include those aspects and put in place in all selected sub-national governments from the very beginning. The Fiscal and Financial Framework 399. The concession of specific transfers from the central government to the selected provincial and municipal governments is the basis of the fiscal strategy for this initial stage. The problem with the design of a fiscal strategy for decentralization in Angola is that the size of public administration (in terms of tax burden and of the overall expenditure-to-GDP ratio) is extremely large in comparison to other African countries and even to emerging economies. This means that a future allocation of public resources for sub-national governments direct management should, in principle, be based on some kind of reallocation of existing resources that are currently used by the central government. 400. Public investments could continue to be totally financed, managed and monitored directly by the central government. Due to the understaffed institutional structure of municipal and communal governments, it is necessary to prepare and implement an administration that at least warrants the management of the current expenditures. The specific objective of the agreements should be therefore the acquisition by sub-national governments of goods, materials, and services that are necessary for decentralized service delivery. 401. Rules for future calculation of these transfers should consider the following: - 152 - · Firstly, the central government must include a specific work program in its annual budget for each of the services to be decentralized. The allocation of the transfer of basic education management, for instance, should not mix up with that of health centers (in opposition to what will be suggested for the subsequent scenarios); · Secondly, the allocation of each program in the central government's annual budget must correspond at least to the funds disbursed in the previous two or three years for the respective action, in all locations targeted by the decentralization process, plus inflation and real GDP oscillation, estimated in the year immediately before and in the year of the budget. It is also necessary to determine an additional margin for new expenditures of sub-national governments on the establishment of the new administrative structure; · Thirdly, the revision of the legal framework should establish formulas for the sharing of funds among the sub-national governments for each program, using technical and objective criteria that cannot be easily changed. In practice, the annual budget plan should include the calculation for the amount anticipated to each program and apply the proportional distribution table per provincial and municipal government as established by the legal framework. · Finally, all administrative measures contemplated by the decentralization plan should be largely advertised. Besides publication in the local press, the central government should provide provincial or municipal authorities with all relevant operational details through a certified statement. For instance, the calculation of the share to be allocated to sub-national government should be published and notified, informing the parameters that have been used, and giving the interested parties time and conditions to occasionally contest the calculations, request corrections or further information. 402. In financial terms, the revision of the legal framework should also include rules for both the entity that is transferring the funds and the entity that will apply them. There is a common ground between these entities that involves the demand for a specific and exclusive bank account for each program, obeying the following principles: the deposits into this account should be separated from all other financial transactions; the provincial or municipal government which received its own legal personality should be the account holder, regardless of whether the autonomy is partial or exceptional; the provincial or municipal authority should be in charge of all bank account movements, and every check or payment order must be signed by the chief of the regional or local Executive branch. 403. There must also be a legal provision requiring the central government to regularly transfer the funds to be decentralized. A simple and efficient operational mechanism is to determine that the grants established in the annual budget should be paid using the payment system based on the release of a twelfth-share of the total annual transfer (duodecimo) with a preset date in the legal framework (e.g., on the fifth business day). - 153 - 404. The recipient governments should also be obliged to have their own treasury, accounting and auditing services for the management of the accounts of each decentralization program. Bank balances should serve as ex-post controls, which may still be carried out by the central government at the initial stage of decentralization. Phase 2: Medium-term Administrative Decentralization 405. The second stage is an extension of administrative decentralization. Ideally it should ensure the continuity of the previous phase instead of being a different or parallel process. The scope of the decentralization process could now be broadened in order to include a larger number of sub-national governments and a longer time frame ­ six years. The Legal Framework 406. In the second phase, the direct election of sub-national authorities would be ideal. In terms of government competencies, the goal is to increase the number of basic social services to be decentralized ­ in areas such as education, health, social assistance, garbage collection, and solid waste treatment. 407. The deployment of the initial phase should go through the expansion of fiscal, budgetary, and financial autonomy granted to sub-national governments. This is expected to happen when the scope of the decentralization programs is broadened. For instance, in the case of education, if the initial phase included only basic education it should now include also preschool education and maybe even college education which could be decentralized to provinces. 408. The most relevant legal difference in the second phase consists in authorizing the legal personality of provincial and municipal governments included in the decentralization plan to fully assume all rights and duties of any governmental entity. This means permission for three actions: (i) hiring their own servants; (ii) collecting taxes directly (although regulated and introduced by the central government); and (iii) issuing their own balance sheets. The Political-Institutional Framework 409. The revision of the legal framework in the second phase of administrative decentralization would grant the same rights and duties of the central Executive to the municipal Executive regarding the administration of public accounts and matters of public concern. This is not dependent upon the choice of the municipal authority but it would be preferable to have the municipal authority elected directly. However, if the municipal authority is appointed by the center instead of being elected, he/she should have absolute authority to administer the province or municipality. 410. A more substantial change would be the local government's right to collect typically local taxes ­ property tax and occasional taxes to be paid by real estate owners, taxes on garbage collection, cleaning of the streets, and public lighting. The - 154 - central government should continue to regulate or define the tax rates or parameters for the collection of the property tax (such as zoning of cities and mapping of generic values), as well as the tax entries. The municipal governments should be in charge of collecting the taxes and maintaining the respective registers. If that is not possible, then the responsibility should be handed over to the provincial government (respecting the territory of the municipality that refused to or could not administer the revenue directly). 411. It would be desirable if governments, at least those included in the second stage of the decentralization process, could take good advantage of the implementation and initial operation of a local Advisory Council and develop a new fiscal and political culture. It is important that all members of this Council live in the same municipality. It is also important that they at least discuss issues related to the local administration and approve the municipal budget before the beginning of the accounting period; inspect its execution during this period; and then analyze the accounts and take a stand, albeit more political than technical. This may be done prior to the analysis and decision processes undertaken outside the territory (at the provincial or national level). 412. Another substantial change that could be implemented in the second stage of decentralization is the creation of an independent and technical body for the analysis of annual accounts. Ideally, this accounting office should be created outside the central government's structure, as a supportive agency to the national Legislative power. Finally, the agreements ­ institutional instruments to regulate intergovernmental relationships in the first stage of the process ­ should be replaced in the second stage with sectoral programs. Each program should be regulated by a central government act, with detailed nationwide operational rules (thus avoiding different treatments by sub-national governments). The sub-national government should implement the national program as a prerequisite to receive transfers regularly. The Fiscal and Financial Framework 413. The scope of the second stage should be restricted to the decentralization of the so-called current expenditures incurred to comply with the programs selected as object of this process. Expenditures on investments should still be undertaken by the central government, including the functions and responsibilities extended in the scope of the second stage of decentralization. Unlike the previous stage, which did not allow for investments, sub-national governments should be free to carry out works and to purchase machines and facilities. This is expected to occur marginally due to the massive dependence of sub-national governments on transfers from the central government; there are only two sources of resources left for direct local investments - the product of revenue from local taxes or the release of funds by the central government. 414. The system of transfers should continue to be contingent upon conditional transfers, but with a more general approach, such as the so-called block grants. If the first stage demanded very specific transfers, with detailed conditionalities (e.g.: for - 155 - maintenance of the public elementary school system), the second stage may include transfers with sectoral applications (e.g.: education). The budgetary and financial strategy defined in the first stage should be maintained, but the managerial instruments have to be replaced. 415. Another important characteristic of this second stage is the decentralization of control over local tax revenues. It is important to note, however, that this does not mean transferring competence in order to be able to impose and regulate taxes, which should still be under the full responsibility of the central government. The objective is to transfer only the collection and inspection of some taxes (e.g.: property tax) to the local government, and to maintain the regulation and definition of tax collection by the central government. On the other hand, the tax revenue obtained in the territory would be part of the local government's own revenues and could be freely applied. This alone may be an important incentive for good performance in the collection of the taxes mentioned in contrast to the applications of sectoral transfers. 416. In this second phase, the contemplated sub-national governments are expected to assume full responsibility for budgetary, fiscal and also equity management. In summary, they are expected to detach themselves from the central government's management and budget structure and to become totally autonomous in terms of fiscal management. In practice, they remain dependent upon the central government's structure in that nearly half of their resources would still come from the center. 417. This medium-term stage is one more step in furthering the administrative decentralization process. The aim is to allow sub-national governments to acquire fuller knowledge of all the stages of fiscal management and to ensure a smooth transition from an administrative to a fiscal decentralization. Changes would still need to be made carefully, since this is a more quantitative expansion phase rather than the qualitative changes in the decentralization process. Restructuring can only occur in the long run, and based on the experience obtained from the two aforementioned stages. Phase 3: Fiscal Decentralization 418. Fiscal decentralization is the ideal arrival point. The third phase of the decentralization process should include direct elections for sub-national authorities and the transfer of all basic services to sub-national governments. It would also be necessary to establish a national system for the transfer of funds, control the indebtedness capacity of sub-national governments, and to grant more autonomy over the collection of typically local taxes. The major objective is to provide reasonable autonomy for sub-national governments to exercise their power and execute their expenditures through a more autonomous fiscal administration. This, however, is considered to be the hardest part. - 156 - The Legal Framework 419. The revision of the legal framework is crucial in this last phase. First of all, unlike the changes implemented through the Law enacted in early 2007, the changes should be more radical. This may mean a more comprehensive political process, encompassing discussions as broad as possible at the political and community levels. 420. The revision of the legal framework should include two key principles for the implementation of fiscal decentralization: · Firstly, the new process should be implemented on a nationwide basis, being extended to all sub-national governments. There should be a transition period for municipalities not contemplated in the previous two phases, allowing some time for adaptation. The central government should be definitively dismissed from the delivery of services. · Secondly, it is necessary to have a new national system for fund transfers, no longer based on conditional transfers, but on tax revenue sharing, where recipient governments may apply the funds as they wish. A quite simple alternative to define the amount to be regularly transferred is to apply a percentage value on the central government's total revenue from taxes and from oil. Moreover, the competence to collecting (instead of regulating or imposing) all municipal taxes and duties should be delegated to municipal governments all over the country, integrating the budget of the tax collector with the resulting revenue (thus encouraging good management of tax collection), whereas municipalities that fully collect these taxes and are able to pay for their expenditures should be allowed to contract long-term debts (including bank loans). The Institutional Framework 421. The previously suggested revision of the legal framework includes factors that mark out the political and institutional plan. It is important to stress, though, that unlike previous scenarios, fiscal decentralization combines with the definition of fixed tenures and direct elections by the population residing in each jurisdiction of the municipal chief executive and also of the respective members of the legislative branch. 422. The budgetary process may be totally entrusted to the municipal government, as well as the financial, equity and personnel management. However, one should recall that most civil servants would be those employed by the central government and would still get their pay from it, with all of the workers' rights preserved, but new employees should be hired directly by municipal governments. 423. The federal government's direct interference on decentralized actions and services in terms of public expenditure should be restricted to the investment policy, which continues to be financed and with all expenditures directly controlled by the central government. Nevertheless, the central government should still depend on - 157 - municipal governments to carry out some activity, especially in the social area (it is hard to decentralize investments in infrastructure). The central government may also continue to raise and consolidate information in order to follow and assess the physical and financial performance and to inspect the public accounts and matters of public concern. Ideally, there should be a national accounting council whose technical management is independent from the executive branch, central government or sub- national governments. The Fiscal and Financial Framework 424. An array of measures should be adopted in order to grant sub-national governments full autonomy so that they can perform their functions with regard to the major legal action of the decentralization process. 425. First, a general tax and oil revenue sharing fund should be created. This fund for revenues obtained from the central government and to be distributed to sub- national governments should have the following characteristics: · The resources should be transferred to sub-national governments on a monthly basis to decentralized programs via agreements; in special cases, the transfers may be carried out on a fortnightly or weekly basis; · The criteria for the allocation of resources should also be predefined in the same legal act and should consider the population of each jurisdiction as a basic indicator, in addition to other municipal indicators closely related to the demand for public goods and services (a proxy that can be easily assessed and updated is the household consumption of electricity, at least in urban areas); · An alternative sharing criterion is to use different formulae for different transactions involving general purpose transfers in such a way that each transaction is associated with one of the decentralized government functions. For instance, if it is established that the funds transferred must be used in the maintenance and development of public education, a possible option could be to allocate these funds according to the number of students enrolled in the respective school. 426. As previously mentioned, municipal governments could be allowed to use the revenues from municipal taxes and the most developed ones may even contract loans. This process, however, should not be so decentralized as that of social services. Based on the historical, social and economic characteristics of the country, the power over these two types of revenues should remain in the hands of the central government. Therefore, the implementation, definition of tax rates and regulation of taxes (e.g.: property tax, or even taxes on the delivery of services, business licenses) should still be under the responsibility of the central government, but the management of tax revenues should be delegated to municipal governments. In the third phase, changes may be proposed to the central government, with the definition of different and higher tax rates for better developed regions ­ which could be collected as additional rates. - 158 - - 159 - Selected References Afonso, José Roberto Rodrigues, Rezende, Fernando. "The Brazilian Federalism: Facts, Challenges, and Perspectives". In: Wallack, Jessica S.; Srinivasan, T. N. (Org.). "Federalism and Economic Reform: International Perspectives". New York. Afonso, José Roberto, Araújo, Érika Amorim. "Municipal Organization and Finance: Brazil", In: Shah, Anwar "Local Governance in Developing Countries", The World Bank, Washington D.C. Angola. Balanço da Execução do Programa Geral de Governo 2005-2006 (Ano 2005). Luanda: Angola, 2006. Angola. Ministério da Saúde. Angola: Despesa Pública no Sector da Saúde 2000- 2006. Luanda: Angola, 2006b. Bernard, Jones (1996), Financial Management in the Public Sector, McGraw Hill. Bird, Richard M; Acosta, Olga Lucia. The Dilemma of Decentralization in Colombia. International Tax Project Paper 404, Toronto: University of Toronto, 2003. Boletim Demográfico nº 9, Projecção da População do País por Províncias e Grupos Quinquenais de Idade para o Período 1985/2010. Boyle, S. et al., (2002), Reaching the Poor: The `Costs' of Sending Children to School, a Six-Country Comparative Study, Educational Paper 47, Department for International Development, Policy Division, Sevenoaks, U. K. Breier, M. (2006) Global Competence/ Local conscience: dilemma in the education of professionals in South Africa and a case study on doctors, Cape Town BRUNS, B., A. MINGAT e R. RAKOTOMALALA, eds. (2003), Achieving Universal Primary Education by 2015: A Chance for Every Child. World Bank, Washington, D. C. Budget proposals (recurrent and capital) by functional and economic classification (1999, 2000, 2001, 2002 and 2003). Chaparro, Juan Camilo; Smart, Michael; Zapata, Juan Gonzalo. Intergovernamental Transfers and Municipal Finance in Colombia. International Tax Project Paper 403, Toronto: University of Toronto, 2003. COLCLOUGH, C (ed), 1997, Marketizing Education and Health in Developing Countries: Miracle or Mirage?, Clarendon Press, Oxford. COLCLOUGH, C. and AL-SAMARRAI, S., (2000), Achieving Schooling for All: Budgetary Expenditures on Education in Sub-Saharan Africa and South Asia, World Development. Dobla-Norris, E., Wade, P. The Challenge of Fiscal Decentralization in Transition Countries. IMF Working Paper 103. Washington D.C: IMF, june 2002. - 160 - EICHER, J.-C. (1984), Educational Costing and Financing in Developing Countries: Focus on Sub-Saharan Africa, World Bank Staff Working Paper 655, Washington, D. C. Gamkhar, Shama; Shah, Anwar. "The Impact of Intergovernamental Fiscal Transfers: A Synthesis of the Conceptual and Empirical Literature". In: Boadway, Robin; Shah, Anwar (Ed.) Intergovernamental Fiscal Transfers: Principles and Practices. Washington D.C: World Bank, 2006. Government of Angola ­ Council of Ministers (1997) Decree No17/99 of 29 of October, Decree on the organic structure of the Provincial governments and municipal and communal administrations, Luanda Government of Angola ­ Council of Ministers (2000) Decree No 2/00 of 14 of January, Estatuto Orgânico DU Ministério da Saúde, Luanda Government of Angola ­ Council of Ministers (2000) Decree No 27/00 of 19 of May, Paradigma de regulamento DUs Governos das Províncias, das administrações DUs municípios e das comunas, Luanda Government of Angola ­ Council of Ministers (2003) Decreee No 54/03 of 5 of August. Regulamento Geral das Unidades Sanitárias do Serviço Nacional de Saúde, Luanda IBAM. "Avaliação das Necessidades de Formação da Administração Local". UNDP/IBAM/Governo de Angola, 2006. IMF (2003), Angola: Selected Issues and Statistical Appendix, IMF Country Report No. 03/292, Washington D.C. International Monetary Fund (2006) Angola: 2006 Article IV Consultations Preliminary Conclusions of the IMF Mission, FMI website. International Monetary Fund. Colombia: 2006 Article IV Consultation and the Third and Final Review under Stand-by Arrangement. IMF Country Report No. 06/408, Washington D.C: IMF, 2006. JMJ International. "O Ambiente do Funcionamento da Governação Local em Angola: Reforçar as Ligações entre Descentralização e o Desenvolvimento Dirigido pela Comunidade". Relatório Final para o Banco Mundial, Washington D.C: World Bank, 2006 Lenneiye, N. Mungai. Who has the Yam and who has the Knife? Social Action Funds and Decentralization in Malawi, Tanzania and Uganda? Social Protection Discussion Paper Series 518, Washington D.C: World Bank, 2005. McIntyre, D. Gilson, L. Mutiambyzi, V. (2005) Promoting equitable health care financing in African context: Current challenges and future prospects, Regional Network for Equity in Health in Southern Africa, Equinet discussion paper number 27, South Africa Métier. "Projecto de Descentralização e Governação Local ­ Angola: Relatório da Avaliação de Meio Termo", mimeo. Maputo: Métier Consultoria, jan. 2007. - 161 - MINFIN ­ Direcção Nacional de Contabilidade (2002) Conta Geral do EstaDU 2001, Luanda MINFIN ­ Direcção Nacional de Contabilidade (2004) Relatório de execução orçamental e financeira DU OGE, Luanda MINFIN ­ Direcção Nacional do Orçamento (2005) Manual de Elaboração da Proposta Orçamental para 2006, Luanda MINPLAN ­ Direcção de Estudos e Planeamento (2004 Estratégia de Combate à Pobreza ­ Reinserção social, reabilitação e reconstrução e estabilização económica, Versão revista em 2006, Luanda MOH ­ DNRH (2006) Actividades realizadas durante o ano 2005, Relatório, Luanda MOH ­ GEPE ­ PASS (2005) Estudo sobre a comparticipação da população nos custos das unidades sanitárias do Serviço Nacional de Saúde, Luanda MOH ­ GEPE ­ PASS (2005) Estudo sobre os custos das unidades sanitárias do Serviço Nacional de Saúde, Luanda. MOH ­ GEPE - PASS (2007) Angola: Despesa Pública no Sector da Saúde: 2000- 2006, Luanda. Nações Unidas (2002) O financiamento público dos sectores sociais em Angola, Principia, Cascais O'NEILL, D. (1995), Education and Income Growth: Implications for Cross-Country Inequality, Journal of Political Economy 103, no. 6: 1289­1301. Oates, Wallace E (Ed.). The Economics of Fiscal Federalism and Local Finance. Cheltenham: Edward Elgar, 1998. Oates, Wallace E. Decentralization of the Public Sector: An Overview. In: Bennet R. (Ed.) Decentralization, Local Governments and Markets. Oxford: Oxford University Press, 1990. OECD - Organization for Economic Cooperation and Development. Intergovernamental Transfers and Decentralized Public Expenditures. OECD Economics Department Working Paper Nr. 3, Paris: OECDE, 2006. OXFORD POLICY MANAGEMENT, (OPM) (2002), Unit Cost of Education in Zambia; Final Report, Oxford, England Pavignani, E. (2005) Analysing disrupted health sectors: a toolkit, instrument available on WHO website PNUD - Programa das Nações Unidas para o Desenvolvimento. Relatório de Desenvolvimento Humano 2006 ­ A Água para lá da Escassez: Poder, Pobreza e Crise Mundial da Água. Nova York: ONU, 2006. PNUD - Angola Millenium Goals Report Summary 2005. Luanda: UNDP, 2005. Population data from the 1970 census. PSACHAROPOULOS, G and PATRINOS, H., (2001), Returns to Investment in Education up to the New Millennium, Mimeo. - 162 - PSACHAROPOULOS, G., (1994), Returns to Investment in Education: A Global Update, World Development. REPUBLIC OF ANGOLA (2001), Estratégia Integrada para a Melhoria do Sistema de Educação (2001-2015), Ministry of Education and Culture, Luanda. REPUBLIC OF ANGOLA (2005), Estratégia para o Desenvolvimento Profissional de PEP, Ministry of Education and Culture, Luanda. SCHULTZ, T. (1992), The Role of Education and Human Capital in Economic Development: An Empirical Assessment, Economic Growth Center Discussion Paper 670. Yale University, Economic Growth Center, New Haven, CT. SCHULTZ, T., (1961), American Economic Review. Shah, Anwar A Comparative Institutional Framework for Responsive, Responsible, and Accountable Local Government. In Local Governance in Industrial Countries, ed. Anwar Shah, 1-40. Wasghinton, DC: The World Bank, 2006b. Shah, Anwar, e Shah, Sana. "The New Vision of Local Governance and the Evolving Roles of Local Governments". In Local Governance in Developing Countries, ed. Anwar Shah, 1-46. Wasghinton, DC: The World Bank, 2006. Shah, Anwar. A Fiscal Need Aproach to Equalization. Canadian Public Policy nr. 22, v.2, pp-99-115, Quebec: Toronto University Press, 1996. Shah, Anwar. A Framework for Evaluate Alternate Institutional Arrangements for Fiscal Equalization Transfers. World Bank Policy Research Working Paper 3785, Washington D.C.: World Bank, 2005. Shah, Anwar. A Practitioner's Guide to Intergovernamental Fiscal Transfers. World Bank Policy Research Working Paper 4039, Washington D.C.: World Bank, 2006a. Shah, Anwar. Fiscal Decentralization in Developing Countries and Transition Economies: Progress, Problems, and the Promise. World Bank Policy Research Working Paper 3282, Washington D.C.: World Bank, 2004. Stein, E. Fiscal Decentralization and Government Size in Latin America. Journal of Applied Economics, Volume 2, n. 2, pp. 357-391. Nov 1999. Ter-Minassian, Teresa. Fiscal Federalism in Theory and Practice. Washington, D.C.: International Monetary Fund, 1997. UNESCO (2003), Orçamentos e despesas de Educação ­ Perspectivas em Cabo Verde, Angola, Moçambique, IIPE, Paris. United Nations Development Program (2002) Estudo sobre a "Macro-Estrutura da Administração Local", Luanda United Nations Development Program (UNDP); United Nations Capital Development Fund (UNCDF). Diagnóstico da Descentralização Fiscal em Angola. Luanda: UNDP/ UNCDF, 2006. - 163 - USAID ­ Connor, C. Rajkotia, Y. Lin, Y. Figuereido, P. (2005) Angola Health System Assessment, Luanda Wood, Guillermo. "Angola: Estudio de Revisión de los Gastos Públicos: Revisión de los Gastos e Inversiones Públicas Destinadas al Sector Agropecuario". Informe de consultoria, mimeo. FAO/Banco Mundial, dez 2006. World Bank (2005) Angola Education Sector Expenditure Review, preliminary draft, Luanda World Bank (2006), Angola Higher Education Status Report ­ draft version. World Bank. Angola ­ Country Economic Memorandum: Oil, Broad-Based Growth and Equity. Washington D.C: World Bank, 2006. World Bank. Colombia Public Expenditure Review. Report No. 25163-CO, Washington D.C: World Bank, 2005. World Bank. Decentralization Can Work: Experience from Colombia. Agriculture and Natural Resource Department Dissemination Note 3, Washington D.C: World Bank, 1995b. World Bank. Local Governance Capacity in Colombia: Beyond Technical Assistance. Washington D.C: World Bank, 1995. World Bank. The Tanzania Social Action Fund (TASAF): owning the process of measuring impact and achieving results. World Bank, 2006. - 164 - STATISTICAL APPENDIX - 165 - Angola at a glance 4/5/07 Sub- Lower Key Development Indicators Saharan middle Age distribution, 2005 Angola Africa income (2006) Male Female Population, mid-year (millions) 16.4 741 2,475 70-74 Surface area (thousand sq. km) 1,247 24,265 39,946 60-64 Population growth (%) 2.8 2.1 1.0 50-54 Urban population (% of total population) 53 35 50 40-44 30-34 GNI (Atlas method, US$ billions) 32.5 552 4,746 20-24 GNI per capita (Atlas method, US$) 1,980 745 1,918 GNI per capita (PPP, international $) 2,210 1,981 6,313 10-14 0-4 GDP growth (%) 14.6 5.3 6.9 30 20 10 0 10 20 30 GDP per capita growth (%) 11.4 3.1 5.9 percent (most recent estimate, 2000­2006) Poverty headcount ratio at $1 a day (PPP, %) .. 41 .. Under-5 mortality rate (per 1,000) Poverty headcount ratio at $2 a day (PPP, %) .. 72 .. Life expectancy at birth (years) 41 47 70 300 Infant mortality (per 1,000 live births) 154 100 33 Child malnutrition (% of children under 5) 31 29 12 250 200 Adult literacy, male (% of ages 15 and older) 83 .. 93 Adult literacy, female (% of ages 15 and older) 54 .. 85 150 Gross primary enrollment, male (% of age group) .. 99 115 100 Gross primary enrollment, female (% of age group) .. 87 113 50 Access to an improved water source (% of population) 53 56 82 0 1990 1995 2000 2005 Access to improved sanitation facilities (% of population) 31 37 57 Angola Sub-Saharan Africa Net Aid Flows 1980 1990 2000 2006 a (US$ millions) Net ODA and official aid 52 266 302 442 Growth of GDP and GDP per capita (%) Top 3 donors (in 2005): United States 7 1 37 64 30 Japan 0 0 21 26 20 France 0 11 8 24 10 Aid (% of GNI) .. 3.2 4.1 1.5 0 Aid per capita (US$) 7 25 22 28 -10 -20 Long-Term Economic Trends -30 90 95 00 05 Consumer prices (annual % change) .. 2.4 325.0 12.9 GDP implicit deflator (annual % change) .. 10.9 418.2 8.0 GDP GDP per capita Exchange rate (annual average, local per US$) .. 0.0 10.0 80.2 Terms of trade index (2000 = 100) .. 63 100 136 1980­90 1990­2000 2000­06 (average annual growth %) Population, mid-year (millions) 7.8 10.5 13.8 16.4 3.0 2.7 2.8 GDP (US$ millions) .. 10,260 9,129 44,103 3.4 1.6 11.1 (% of GDP) Agriculture .. 17.9 5.7 7.2 0.5 -1.4 14.3 Industry .. 40.8 72.1 74.0 6.3 4.4 10.5 Manufacturing .. 5.0 2.9 3.6 -11.1 -0.3 13.4 Services .. 41.2 22.2 18.7 1.4 -2.2 6.7 Household final consumption expenditure .. 35.8 .. .. -0.2 .. .. General gov't final consumption expenditure .. 34.5 .. .. 6.7 .. .. Gross capital formation .. 11.7 15.1 14.1 -5.1 .. .. Exports of goods and services .. 38.9 89.6 73.5 2.2 .. .. Imports of goods and services .. 20.9 62.8 48.3 -2.6 .. .. Gross savings .. 9.0 23.8 22.9 .. .. .. Note: Figures in italics are for years other than those specified. 2006 data are preliminary. Group data are for 2005. .. indicates data are not available. a. Aid data are for 2005. Development Economics, Development Data Group (DECDG). Angola Balance of Payments and Trade 2000 2006 Governance indicators, 2000 and 2005 (US$ millions) Total merchandise exports (fob) 7,920 23,724 Total merchandise imports (cif) 3,040 8,667 Voice and accountability Net trade in goods and services 2,448 8,286 Political stability Workers' remittances and compensation of employees (receipts) .. .. Regulatory quality Rule of law Current account balance 795 3,896 as a % of GDP 8.7 8.8 Control of corruption Reserves, including gold 1,198 3,197 0 25 50 75 100 2005 Country's percentile rank (0-100) Central Government Finance 2000 2006 2000 higher values imply better ratings (% of GDP) Revenue 50.2 46.7 Source: Kaufmann-Kraay-Mastruzzi, World Bank Tax revenue 50.0 46.5 Total expenditure 49.6 37.3 Technology and Infrastructure 2000 2005 Overall balance (accrual basis) 0.6 9.4 Paved roads (% of total) 10.4 .. Highest marginal tax rate (%) Fixed line and mobile phone Individual .. .. subscribers (per 1,000 people) 7 75 Corporate .. .. High technology exports (% of manufactured exports) .. .. External Debt and Resource Flows Environment (US$ millions) Total debt outstanding and disbursed 9,408 11,755 Agricultural land (% of land area) 46 46 Total debt service 1,705 2,239 Forest area (% of land area, 2000 and 2005) 47.9 47.4 HIPC and MDRI debt relief (expected; flow) ­ ­ Nationally protected areas (% of land area) .. 10.1 Total debt (% of GDP) 103.1 35.8 Freshwater resources per capita (cu. meters) .. 9,284 Total debt service (% of exports) 20.3 9.3 Freshwater withdrawal (% of internal resources) 0.2 .. Foreign direct investment (net inflows) 879 -1,304 CO2 emissions per capita (mt) 0.48 0.57 Portfolio equity (net inflows) 0 0 GDP per unit of energy use (2000 PPP $ per kg of oil equivalent) 3.1 3.3 Composition of total external debt, 2005 Energy use per capita (kg of oil equivalent) 572 613 IMF, 0 IDA, 319 Other multi- IBRD, 0 lateral, 52 World Bank Group portfolio 2000 2005 Bilateral, 2,817 Short-term, 2,327 (US$ millions) IBRD Total debt outstanding and disbursed ­ ­ Disbursements ­ ­ Principal repayments ­ ­ Private, 6,240 Interest payments ­ ­ US$ millions IDA Total debt outstanding and disbursed 226 319 Disbursements 24 30 Private Sector Development 2000 2006 Total debt service 2 7 Time required to start a business (days) ­ 124 IFC (fiscal year) Cost to start a business (% of GNI per capita) ­ 486.7 Total disbursed and outstanding portfolio 1 2 Time required to register property (days) ­ 334 of which IFC own account 1 2 Disbursements for IFC own account 1 1 Ranked as a major constraint to business Portfolio sales, prepayments and (% of managers surveyed who agreed) repayments for IFC own account 0 0 n.a. .. .. n.a. .. .. MIGA Gross exposure 21 16 Stock market capitalization (% of GDP) .. .. New guarantees 0 15 Bank branches (per 100,000 people) .. .. Note: Figures in italics are for years other than those specified. 2006 data are preliminary. 4/5/07 .. indicates data are not available. ­ indicates observation is not applicable. Development Economics, Development Data Group (DECDG). - 167 - Millennium Development Goals Angola With selected targets to achieve between 1990 and 2015 (estimate closest to date shown, +/- 2 years) Angola Goal 1: halve the rates for $1 a day poverty and malnutrition 1990 1995 2000 2005 Poverty headcount ratio at $1 a day (PPP, % of population) .. .. .. .. Poverty headcount ratio at national poverty line (% of population) .. .. .. .. Share of income or consumption to the poorest qunitile (%) .. .. .. .. Prevalence of malnutrition (% of children under 5) 20 41 31 .. Goal 2: ensure that children are able to complete primary schooling Primary school enrollment (net, %) 50 .. 53 .. Primary completion rate (% of relevant age group) 35 .. .. .. Secondary school enrollment (gross, %) 11 .. 15 .. Youth literacy rate (% of people ages 15-24) .. .. .. 72 Goal 3: eliminate gender disparity in education and empower women Ratio of girls to boys in primary and secondary education (%) .. .. 85 .. Women employed in the nonagricultural sector (% of nonagricultural employment) .. 24 .. .. Proportion of seats held by women in national parliament (%) 15 10 16 15 Goal 4: reduce under-5 mortality by two-thirds Under-5 mortality rate (per 1,000) 260 260 260 260 Infant mortality rate (per 1,000 live births) 154 154 154 154 Measles immunization (proportion of one-year olds immunized, %) 38 46 41 45 Goal 5: reduce maternal mortality by three-fourths Maternal mortality ratio (modeled estimate, per 100,000 live births) .. .. 1,700 .. Births attended by skilled health staff (% of total) .. 23 47 .. Goal 6: halt and begin to reverse the spread of HIV/AIDS and other major diseases Prevalence of HIV (% of population ages 15-49) .. .. .. 3.7 Contraceptive prevalence (% of women ages 15-49) .. .. 6 .. Incidence of tuberculosis (per 100,000 people) 210 228 248 269 Tuberculosis cases detected under DOTS (%) .. 62 76 85 Goal 7: halve the proportion of people without sustainable access to basic needs Access to an improved water source (% of population) 36 .. .. 53 Access to improved sanitation facilities (% of population) 29 .. .. 31 Forest area (% of total land area) 48.9 .. 47.9 47.4 Nationally protected areas (% of total land area) .. .. .. 10.1 CO2 emissions (metric tons per capita) 0.4 0.9 0.5 0.6 GDP per unit of energy use (constant 2000 PPP $ per kg of oil equivalent) 3.7 3.0 3.1 3.3 Goal 8: develop a global partnership for development Fixed line and mobile phone subscribers (per 1,000 people) 7 4 7 75 Internet users (per 1,000 people) 0 0 1 11 Personal computers (per 1,000 people) .. 1 1 .. Youth unemployment (% of total labor force ages 15-24) .. .. .. .. Education indicators (%) Measles immunization (% of 1-year olds) ICT indicators (per 1,000 people) 100 100 80 70 75 60 50 75 50 40 30 25 20 50 10 1998 2000 2002 2004 0 0 1990 1995 2000 2005 2000 2002 2005 Primary net enrollment ratio Fixed + mobile subscribers Ratio of girls to boys in primary & Angola Sub-Saharan Africa Internet users secondary education Note: Figures in italics are for years other than those specified. .. indicates data are not available. 4/5/07 Development Economics, Development Data Group (DECDG). - 168 -