PUBLIC EXPENDITURE REVIEW Gabon Improving public spending quality to foster inclusive growth Document of The World Bank FOR OFFICIAL USE ONLY GABON PUBLIC EXPENDITURE REVIEW Improving public spending quality to foster inclusive growth ABBREVIATIONS AND ACRONYMS 6 ANGTI National Agency for Major Infrastructure Works (Agence Nationale des Grands Travaux d’Infrastructures) BEPC First-Cycle Certificate of Study (Brevet d’Etudes du Premier Cycle) BOP Operational program budget (budget opérationnel de programme) CAR Central African Republic CEMAC Central African Economic and Monetary Community (Communauté Économique et Monétaire de l’Afrique Centrale) CEP Certificate of Primary Studies (Certificat d’Etudes Primaires) CNAMGS National Health Insurance and Social Security Fund (Caisse Nationale d´Assurance Maladie et de Garantie Sociale) CP Payment appropriation (credits de payment) CSS Social Solidarity Contribution (Contribution Social Solidaire) DCAF Central Department of Financial Affairs (Direction Centrale des Affaires Financières) DGI General Tax Directorate (Direction Générale des Impôts) DHS Demographic and Health Survey FCFA Central African CFA francs GCI Global Competitiveness Index GEF Economically weak Gabonese (Gabonais économiquement faible) GOG Government of Gabon HRMIS Human resource management information system IMF International Monetary Fund IT Information technology LOLFEB Organic Law on the Laws of Finance and Budget Execution (Loi Organique relative aux Lois de Finances et à l’Exécution du Budget) MENEC Ministry of National Education and Civic Education (Ministère de l'Education Nationale et de l'Education Civique) MESRSFC Ministry of Higher Education, Scientific Research, and Management Training (Ministère de l’Enseignement Supérieur, de la Recherche Scientifique, et de la Formation des Cadres) MOH Ministry of Health MPI Multidimensional Poverty Index MTEFTPIJ Ministry of Labor, Employment, Technical and Vocational Training and Youth Integration (Ministère du Travail, de l’Emploi, de la Formation Technique et Professionnelle et de l’Insertion des Jeunes) NHA National Health Account OECD Organisation for Economic Co-operation and Development OGR Gabonese Revenue Authority (Office Gabonais des Recettes) OU Operational unit (unité opérationnelle) PAP Annual Performance Plan (Plan Annuel de Performance) PBF Performance-based financing PFM Public financial management PIP Performance-based wage premium (prime d ’incitation à la performance) PNDS National Health Development Plan (Plan National de Développement Sanitaire) PPP Public-private partnership PRE Economic Recovery Plan (Plan de Relance Economique) PSGE Emerging Gabon Strategic Plan (Plan Stratégique Gabon Emergent) RAP Annual Performance Report (Rapport Annuel de Performance) ROAM Obligatory fee for health insurance (redevance obligatoire à l’assurance maladie) RPROG Manager responsible for programs (responsables de programme) RUP Recognized as public utilities (reconnaissance d’utilité publique) SEEG Water and Energy Company of Gabon (Société d’Energie et d’Eau du Gabon) SEZ Special economic zone SOGARA Gabonese Refinery Company (Société Gabonaise de Raffinage) SSA Sub-Saharan Africa SYDOGEP Customs System for Managing Economic and Privileged Regimes (Système Douanier de Gestion des Régimes Economiques et Privilégiés) TVET technical and vocational education and training UNESCO United Nations Educational, Scientific and Cultural Organization VAT Value-added tax WDI World Development Indicators WEO World Economic Outlook WHO World Health Organization Table of Contents ACKNOWLEDGEMENTS .......................................................................................................... i EXECUTIVE SUMMARY .......................................................................................................... ii 1. FISCAL POLICY AND PUBLIC FINANCIAL MANAGEMENT ................................. 1 Introduction ............................................................................................................................................. 1 Context ..................................................................................................................................................... 1 Fiscal Trends: 2006-2017 ........................................................................................................................ 1 Public Spending and Economic Growth .............................................................................................. 1 Fiscal Policy......................................................................................................................................... 3 Fiscal Outlook: 2017-2020 ...................................................................................................................... 5 Public Financial Management................................................................................................................ 6 The Legal and Institutional Framework for PFM ................................................................................ 7 Fiscal Discipline .................................................................................................................................. 7 Strategic Resource Allocation .............................................................................................................. 8 Efficient Service Delivery and Financial Transparency ...................................................................... 8 Recommendations .................................................................................................................................... 9 2. NON-OIL REVENUE MOBILIZATION ........................................................................ 10 Introduction ........................................................................................................................................... 10 Revenue Levels, Trends, and Composition ........................................................................................... 11 Revenue Collected at the Border ........................................................................................................ 15 VAT on Domestic Goods and Services ............................................................................................... 16 Income Taxes ...................................................................................................................................... 17 Property Taxes ................................................................................................................................... 19 Tax Administration ................................................................................................................................ 21 Recent Tax Policy and Administration Reforms: Efficacy and Impact .............................................. 23 Recommendations .................................................................................................................................. 26 3. OVERALL EXPENDITURE TRENDS AND QUALITY ANALYSIS ......................... 28 Introduction ........................................................................................................................................... 28 Overall Expenditure Trends .................................................................................................................. 28 Public Investment .................................................................................................................................. 33 Overall Investment Spending Trends ................................................................................................. 33 Public Investment and Infrastructure Quality .................................................................................... 35 Public Investment Management............................................................................................................ 38 Economic Analysis of Investment Proposals ...................................................................................... 38 Investment Project Selection .............................................................................................................. 38 Investment Project Costing ................................................................................................................ 39 Investment Project Monitoring........................................................................................................... 39 The Wage Bill and Public Employment Management ......................................................................... 40 Recommendations .................................................................................................................................. 44 4. EDUCATION ...................................................................................................................... 47 Introduction ........................................................................................................................................... 47 Education Sector Overview ................................................................................................................... 47 Education System and Service Delivery ............................................................................................. 47 Education Outcomes .......................................................................................................................... 49 Overview of Education Expenditures: Funding, Sources, and Trends .............................................. 54 Budget and Spending Processes ......................................................................................................... 54 Education Expenditure Levels and Trends......................................................................................... 55 Sources of Education Expenditures.................................................................................................... 58 Analysis of Expenditures: Efficiency Issues......................................................................................... 59 Internal Efficiency Issues ................................................................................................................... 60 Economic and Functional Allocation Issues ...................................................................................... 61 Analysis of Expenditures: Equity Issues............................................................................................... 64 Analysis of Expenditures: Cross-cutting Issues ................................................................................... 67 Public Financial Management Issues ................................................................................................ 67 Human Resource Management Issues................................................................................................ 68 Infrastructure Utilization Issues......................................................................................................... 70 Recommendations .................................................................................................................................. 71 5. SOCIAL PROTECTION ................................................................................................... 76 Introduction ........................................................................................................................................... 76 Overview of the Social Protection Sector .............................................................................................. 78 Safety-net System ................................................................................................................................ 78 Outcomes ............................................................................................................................................ 81 Overview of Social Protection Expenditures: Structure, Sources, and Trends ................................... 82 Analysis of Expenditures: Adequacy Issues ......................................................................................... 86 Analysis of Expenditures: Efficiency and Equity Issues...................................................................... 90 Cash Transfers ................................................................................................................................... 90 Subsidies ............................................................................................................................................. 93 Free Goods and Services.................................................................................................................... 94 The “Graduation” Program .............................................................................................................. 96 Pensions and Social Insurance .......................................................................................................... 97 Recommendations .................................................................................................................................. 98 6. HEALTH ........................................................................................................................... 102 Introduction ......................................................................................................................................... 102 Context ................................................................................................................................................. 102 The Structure of the Gabonese Health Sector .................................................................................. 102 Health Indicators, Access to Healthcare, and the Quality of Health Services .................................. 105 Health Expenditures ............................................................................................................................ 110 Expenditure Levels and Trends ........................................................................................................ 110 Health Spending by Source .............................................................................................................. 112 Expenditure Analysis: Efficiency Issues ............................................................................................ 115 Expenditure Analysis: Equity Issues .................................................................................................. 121 Recommendations ................................................................................................................................ 122 ANNEX A: EDUCATION........................................................................................................ 126 ANNEX B: SOCIAL PROTECTION ..................................................................................... 130 ANNEX C: EXPENDITURE TRENDS .................................................................................. 131 ANNEX D: THE PERFORMANCE BASED FINANCING MECHANISM IN PUBLIC HEALTHCARE ........................................................................................................................ 133 Figures Figure 1: Real GDP Growth Rate, 2007-2017 (%) ....................................................................................... 2 Figure 2: Public Investment (% of GDP) ...................................................................................................... 2 Figure 3: Contribution to GDP Growth, Demand Side (%) .......................................................................... 2 Figure 4: Contribution to GDP Growth, Supply Side (%) ............................................................................ 2 Figure 5: Key Fiscal and Debt Indicators, 2006-2017 (% of GDP) .............................................................. 4 Figure 6: Government Revenues and Expenditures, 2006-2017 (% of GDP) .............................................. 5 Figure 7: Recent and Projected Trends in the Fiscal and Debt Profile, 2010-2020 (% of GDP) .................. 6 Figure 8: Oil, Non-Oil, and Total Revenue, Gabon, 2010-2017 (% of GDP) ............................................ 12 Figure 9: Changes in Revenue-to-GDP Ratios, Gabon and Comparators, 2013-2015 (%) ........................ 12 Figure 10: Fiscal Revenue, Gabon and Comparators, 2015 (% of GDP) ................................................... 12 Figure 11: Government Oil Revenue Structure, Gabon, 2010-2016 (% of GDP) ...................................... 13 Figure 12: The Non-Oil Revenue Structure, Gabon, 2016 (% of total non-oil revenue) ............................ 14 Figure 13: Border-Tax Expenditures by Regime, Gabon, 2010 (% of total border-tax expenditures) ....... 15 Figure 14: Border-Tax Expenditures by Regime, Gabon, 2016 (% of total border-tax expenditures) ....... 15 Figure 15: Standard Statutory VAT Rate, Gabon and Comparators ........................................................... 17 Figure 16: Revenue from Taxes on Income, Profits, and Capital Gains, Gabon and Comparators, 2014 or latest ............................................................................................................................................................ 17 Figure 17: PIT Rates, Gabon and Comparators .......................................................................................... 18 Figure 18: CIT Rates, Gabon and Comparators .......................................................................................... 18 Figure 19: Property Tax Revenue Structure, Gabon, 2011-2016 (current CFAF millions)........................ 20 Figure 20: Property Tax Revenue, Gabon and Comparators, 2014 (% of GDP) ........................................ 20 Figure 21: Net Foreign Direct Investment Inflows, Gabon and Comparators, (% of GDP) ....................... 24 Figure 22: Gabon’s Global Rank on Selected Doing Business Indicators, 2016-2017............................... 24 Figure 23: End-of-Period Inflation Growth, Gabon and Comparators, 2013-2016 (percentage points) .... 25 Figure 24: Harmonized Household Consumption Price Index, Gabon, 2011-2015 ................................... 25 Figure 25: Government Expenditure by Economic Structure, 2010-2017 (% of GDP) ............................. 30 Figure 26: General Government Expenditure, 2015 (% of GDP) ............................................................... 30 Figure 27: Comparative Expenditure Structure, 2015 or latest (% of GDP) .............................................. 31 Figure 28: GOG Budget Execution, 2014-2017 ......................................................................................... 31 Figure 29: GOG Executed Expenditure by Function, 2015 (% of total expenditures) ............................... 32 Figure 30: Expenditure Functional Breakdown, 2015 or latest (% of total expenditures).......................... 32 Figure 31: Capital expenditure, 2012 or latest (% of GDP) ........................................................................ 34 Figure 32: Capital expenditure, 2015 or latest (% of GDP) ........................................................................ 34 Figure 33: Functional Distribution of Planned Investments 2015-17 ......................................................... 35 Figure 34: Quality of trade and transport-related infrastructure (1=low to 5=high), 2016 ......................... 35 Figure 35: Quality of Port Infrastructure (1=low to 7=high), 2016 ............................................................ 36 Figure 36: Access to electricity, rural (% of rural population), 2014 ......................................................... 36 Figure 37: Quality of Electricity Supply, (1-7), 2016 ................................................................................. 37 Figure 38: Electric Power Transmission and Distribution Losses (% of output), 2014 .............................. 37 Figure 39: Access to Improved Water Sources (% of Total Population), 2015 .......................................... 38 Figure 40: Access to Improved Sanitation Facilities (% of Total Population), 2015 ................................. 38 Figure 41: Emerging Economies: Wage Bill of Public Administrations, 2014 (% of tax revenues).......... 41 Figure 42: Personnel Spending by Sector ................................................................................................... 43 Figure 43: Average Wage by Sector ........................................................................................................... 43 Figure 44: Per Age Distribution for Grade 1 Pupils in Primary Education ................................................ 50 Figure 45: Per Age Distribution for Grade 6 Pupils in Secondary Education ............................................ 50 Figure 46: Net Attendance Rate in Primary, and Junior, and Senior Secondary Education in Gabon and Peer Countries ............................................................................................................................................. 50 Figure 47: Completion Rates at Primary, Lower Secondary, and Upper Secondary Education Levels ..... 51 Figure 48: Schooling Status of Gabonese Children and Youth by Age, 2012 ............................................ 51 Figure 49: Average Scores for Mathematics and French at Grade 5, 2006 ................................................ 52 Figure 50: Mincerian Regression Result: Earnings Increase by Level of Education (Percentage) ............ 53 Figure 51: Net Attendance Rate in Primary and Secondary Schools, by Gender, Income Quintile, and Place of Residence, 2012 ............................................................................................................................ 54 Figure 52: Public Expenditure on Education as % GDP in Gabon, 2010-2016.......................................... 56 Figure 53: Public Expenditure on Education as a % of GDP and Total Government Expenditure in Gabon and Peer Countries, 2009-2015 ................................................................................................................... 56 Figure 54: Public Expenditure on Education as a % Public Expenditures in Gabon, 2010-2016............... 57 Figure 55: Private Primary Enrolment share in Compared to Income Level and Education Spending, 2010-2016 ................................................................................................................................................... 59 Figure 56: Efficiency of Government Expenditures based on Primary Completion Rates, 2010-2014 ..... 60 Figure 57: Repetition Rate in Primary and Lower Secondary Education in Gabon and Peer Countries .... 61 Figure 58: Distribution of Committed Education Expenditures and the Education Budget by Economic Category, 2015 ............................................................................................................................................ 62 Figure 59: Distribution of Expenditures (2015) and Education Budget (2015 and 2017) by Function ...... 63 Figure 60: Share of Tertiary Education Spending compared to Income Level, 2014 ................................. 63 Figure 61: Distribution Educational Expenditures* by Function, 2015...................................................... 64 Figure 62: Percentage of Enrollment by Quintile and Education Level, 2012 ........................................... 65 Figure 63: Enrollment Concentration Curves by Education Level, 2012 ................................................... 65 Figure 64: Benefit Incidence Analysis of Government Education Spending, 2015 ................................... 66 Figure 65: Government Subsidy to Education from Current Expenditure by Quintile, 2015..................... 66 Figure 66: Gender Equality in Education by Wealth Quintile, 2012 .......................................................... 66 Figure 67: Government Subsidy to Education from Current Expenditure by Gender and Wealth Quintile, 2015 ............................................................................................................................................................ 67 Figure 68: Deficit of Mathematics Teachers in Secondary Education by Provinces, 2013 ........................ 70 Figure 69: Distribution of Student-to-classroom Ratios in Public Primary Schools by Provinces, 2013 ... 71 Figure 70: Safety-net Beneficiaries and Budget Allocations, SIHG........................................................... 81 Figure 71: Financing of Social Safety Nets as a % of GDP in Comparator Countries ............................... 86 Figure 72: Safety Nets for an Average Household (FCFA/year) ................................................................ 89 Figure 73: Safety Nets Implemented in 2015 by Type of Intervention ...................................................... 90 Figure 74: Cash Transfers Implemented in 2015 by Type of Beneficiary .................................................. 93 Figure 75: The Distribution of Healthcare Workers and the Distribution of the Population, 2014 .......... 104 Figure 76: Infant and Maternal Mortality Rates, Gabon and International Comparators ......................... 106 Figure 77: Disability-Adjusted Life Years (DALYs) and Causes of Mortality, 2012 .............................. 107 Figure 78: Total Health Spending per Capita, Gabon and SSA Comparators, 2014 ................................ 110 Figure 79: General Government Health Spending as a Share of Total Health Spending, Gabon and SSA Comparators, 2014 .................................................................................................................................... 111 Figure 80: Total Health Spending per Capita and GDP per Capita, Gabon and International Comparators, 2014 .......................................................................................................................................................... 111 Figure 81: Public Health Spending and Income per Capita, Gabon and Comparators, Latest Available Data ........................................................................................................................................................... 112 Figure 82: Health Spending in Gabon, 1995-2014 ................................................................................... 113 Figure 83: Public Health Financing by Source, 2010-2015 ...................................................................... 114 Figure 84: Health Spending and Health Indicators, Gabon and Upper-Middle-Income Comparators, 2014 .................................................................................................................................................................. 116 Figure 85: Trends in Recurrent and Capital Spending, 2010-2015........................................................... 117 Figure 86: Ministry of Health Non-salary Budget, 2015, Planned and Executed, FCFA ......................... 118 Figure 87: The Planned and Executed Investment Budgets of the 2011-2015 PNDS .............................. 118 Figure 88: The Distribution of Health Spending (public and private) by Provider Type, 2014 ............... 119 Figure 89: The Distribution of Current Health Expenditures by Disease Group, 2014 ............................ 120 Figure 90: Current Health Expenditures by Disease or Type of Care, 2014 ............................................ 121 Figure 91: The Public Health Budget per Capita, 2015 (US$) ................................................................. 122 Figure 92: Population Distribution versus Public Health Expenditure Distribution (%) .......................... 122 Tables Table 1: Overview of Key Recommendations ............................................................................................. xi Table 2: Government Revenue by Source, Gabon, 2010-2017 (% of GDP) .............................................. 11 Table 3: CIT Incentives under the Income Tax Act.................................................................................... 19 Table 4: Estimated Annual Gains from Improvements in Tax Policy and Administration, 2017-2021 (% of GDP) ........................................................................................................................................................... 26 Table 5: Regional Comparison of Gabon’s Payroll Spending with other Natural Resource-rich Countries .................................................................................................................................................................... 40 Table 6: Average Annual Growth Rates of the Wage Bill, Total Public-sector Headcount, and Oil Revenues (2000-2015) ................................................................................................................................ 41 Table 7: Percent Change in the Wage Bill and Staffing by Sector ............................................................ 43 Table 8: Enrollment in Pre-primary, Primary, and Secondary Education, 2001-2015 ............................... 48 Table 9: Primary School Unit-cost Comparisons between Gabon and SSA Countries .............................. 57 Table 10: Execution Rates by Education Level and Economic Category, 2015......................................... 67 Table 11: Social Protection Mechanisms in Gabon .................................................................................... 76 Table 12: Entities Responsible for Social Protection Programs by Household Type ................................. 79 Table 13: Social Protection Expenditures, 2014-2016 (FCAF billion)....................................................... 83 Table 14: Social Security Budget, 2015-2017 (FCFA million) .................................................................. 84 Table 15: Coverage of Gabon’s Social Protection System (Health Insurance), 2016 ................................ 87 Table 16: Benefits of Different Non-Contributory Programs for an Average Household (FCFA/year) .... 89 Table 17: Cash-transfer Programs............................................................................................................... 90 Table 18: Targeted Subsidies in 2015 ......................................................................................................... 94 Table 19: In-kind Transfer Programs in 2015 ............................................................................................. 95 Table 20: Geographic Coverage of Social Assistance, October 2013 to December 2014 .......................... 96 Table 21: Economic Self-Sufficiency Program in 2016 ............................................................................. 97 Table 22: Pensions, 2015 ............................................................................................................................ 97 Table 23: Key Health Indicators, 2001-2013 ............................................................................................ 107 Table 24: Health-Service Indicators by Consumption Quintile, 2012 ...................................................... 110 Table 25: Health Spending in Gabon, 1995-2015..................................................................................... 113 Table 26: 2011-2015 PNDS for Investing in Periphery and Primary Care Services ................................ 119 Table 27: 2017-2021 PNDS Public Health Budget by Expenditure Type, Level, and Year .................... 123 Boxes Box 1: Options for Reforming Investment-Related Tax Incentives ........................................................... 24 Box 2: Gabon’s National Infrastructure Master Plan.................................................................................. 33 Box 3: Public Investment Management in the Health Sector .................................................................... 40 Box 4: Restructuring the Social Protection System in Gabon .................................................................... 78 Box 5: Cash Transfers as a Long-term Poverty Reduction Tool ................................................................ 92 Box 6: Ghana’s Graduation Program ........................................................................................................ 100 ACKNOWLEDGEMENTS This report has been prepared by a team co-led by Maurizia Tovo, Emanuela di Gropello and David Elmaleh under the guidance of Jehan Arulpragasam, Gaston Sorgho, Sona Varma, Francisco Galrao Carneiro, and Seynabou Sakho. Team members include Karen Coulibaly, Leif Jensen, Rick Emery Tsouck Ibounde, Sonia Barbara Ondo Ndong, Ousmane Kolie, Celestin Adjalou Niamien, Lida Bteddini, Harisoa Danielle Rasolonjatovo Andriamihamina, Hamoud Abdel Wedoud Kamil, Nelly Rakoto-Tiana, Catherine Laurent, Paul Jacob Robyn, Jean Claude Taptue Fotso, Elena Celada, and Soule Sow. The report also benefitted from very valuable comments from Elisabeth Huybens, Olivier Godron, Mazen Bouri, Souleymane Coulibaly, Norbert Matthias Fiess, Dominic Haazen, Jan Loeprick, and Dimitris Mavridis. The team would also like to thank the peer reviewers Virginia Alonso Albarran (IMF), Carine Clert, Andreas Blom, Jamele Rigolini, Moulay Driss, and Ulrich Bartsch for their very valuable insights. The research, mission organization, preparation of the report and dissemination counted with valuable administrative support from all colleagues at the Libreville country office, including Sonia Vanecia Boga, Antoinette Pongui, and Firmin Nkoghe. The final version of this report was edited by Oscar Parlback and Sean Lothrop. Photo credit: Odilia Hebga Cover page Graphic design: William Ursenbach i EXECUTIVE SUMMARY Introduction 1. The analysis presented in this Public Expenditure Review (PER) is designed to support the efforts of the government of Gabon (GOG) to enhance public expenditure efficiency and improve the quality of public goods and services. The report reviews public expenditure data over an 8-year period, examines trends in the composition and efficiency of spending, and assesses whether spending patterns are consistent with Gabon’s development objectives. The report includes in-depth analyses of public investment and the public wage bill, the two largest budget categories. It also reviews the composition and evolution of non-oil revenue which could play a key role in closing the fiscal deficit. Finally, the report reviews education, social protection, and public health expenditures which are crucial for socioeconomic development and poverty reduction. 2. The PER’s recommendations aim to support the government’s current fiscal consolidation effort as well as the implementation of the national development strategy. In 2009, the government adopted the Emerging Gabon Strategic Plan (Plan Stratégique Gabon Emergent, PSGE), an ambitious program of economic and social transformation that focuses on accelerating economic diversification by investing in infrastructure and enhancing the business climate. However, since the 2014 oil price collapse which brought about a deep recession across CEMAC countries, Gabon has joined other Central African Economic Community (CEMAC) member states in pursuing fiscal consolidation and expenditure rationalization. Gabon adopted the Economic Recovery Plan (Plan de Relance Economique, PRE) for 2017-2019, which adapts the economic diversification agenda to a context of sustained fiscal consolidation. The PER offers specific recommendations for increasing expenditure efficiency multiple sectors of the public administration so as to improve the quality of service delivery despite a context of limited resources. Macroeconomic Context 3. Gabon’s most pressing development challenges include high levels of unemployment, poverty and inequality, and population growth that regularly outpaces economic growth. Per capita income has fallen over the last decade and a half, as aggregate income growth has failed to keep up with population growth. Gabon was unable to fully capitalize on the positive macroeconomic conditions that prevailed between 2010 and 2014, and despite robust GDP growth, roughly one-third of the population still lives below the poverty line. While no recent poverty statistics are available, World Bank estimates suggest little to no improvement in poverty indicators over time. 4. The 2014 oil-price shock has severely curtailed the available fiscal space for development spending and made fiscal consolidation an urgent priority. Between 2010 and 2014, high oil prices financed large-scale public investment, which accelerated growth. After oil prices collapsed, the government cut public investment sharply, and by 2016 the GDP growth rate had fallen to 2.1 percent per year. However, reduced public investment was partially offset by a rising public sector wage bill. Revenues fell faster than expenditures, widening a preexisting fiscal deficit. ii 5. Driven by mounting arrears and statutory advances from the regional central bank, the public debt stock has more than tripled since 2013. At 64.2 percent of GDP in 2016, Gabon’s debt stock is far above the government’s strategic threshold of 35 percent. In 2016, domestic arrears and total arrears reached 7.7 percent and 9.5 percent of GDP, respectively, while statutory advances stood at 5.4 percent of GDP. Although the fiscal deficit is projected to narrow from -2.3 percent of GDP in 2017 to -1.2 percent in 2018, a combination of stagnant revenues and modest expenditure cuts will prevent a return to surplus. Consequently, debt levels are projected to remain high at around 62 percent of GDP in 2017 and 2018. 6. Economic growth is expected to remain modest in the short term, but ongoing diversification efforts contribute to a more favorable medium-term outlook. Real GDP is expected to grow by 0.8 percent in 2017 and 2.7 percent in 2018. Over the medium term, the expansion of the nonoil sectors is expected to push the overall growth rate to 4.1 percent by 2019. Gross investment is projected to increase in 2018, driven by renewed dynamism in the agricultural and mining sectors, and manganese production should accelerate as international prices recover. 7. Reduced spending will be necessary to curtail the deficit and contain the debt stock, even as the government strives to increase revenue. Large public investment projects remain at the core of the authorities’ diversification strategy. The government has devoted 40 percent of the budget to investment projects designed to boost growth and to create a more favorable business environment. However, the authorities must sustain fiscal discipline to weather the current crisis. This will limit fiscal space for the foreseeable future. The authorities requested International Monetary Fund (IMF) support in 2016 and signed a program in June 2017. Implementation is proceeding on schedule, and early progress in improving the fiscal and external balances is encouraging. The IMF program, which is built on a well-balanced fiscal consolidation, will run from 2017 to 2020 and provides US$628 million in total financing. Under the agreed program framework, public spending will fall to about 19 percent of GDP by 2018, a 3 percentage-point decline from 2016 level. 8. Since 2014 mounting fiscal imbalances have forced the government to cut public spending: there is still scope for further fiscal consolidation but it is limited by rigid expenditure categories such as the large public-sector wage bill. Spending on public investment, subsidies, and transfers bore the brunt of the adjustment, even as administrative reforms and additional debt increased the wage bill and interest payments. The general public sector wage bill (i.e. Ministries of Budget, Economy, Public Service, Relations with Parliament, and Justice) accounted for nearly half of the public-sector wage bill and almost a third of total government expenditures in 2015. The level of public spending in Gabon is currently below the average for peer countries, despite the high public sector wage bill, and further cuts to capital investment could weaken economic growth and worsen the already poor quality of the country’s infrastructure. 9. Gabon should allocate less funding to public administration and more to housing, health, education, and social protection. Almost half of Gabon’s government expenditures are devoted to public administration and only 20 percent to health, education, and social protection. Improving the country’s roads, seaports, energy efficiency, power generation and distribution capacity, communications infrastructure, agriculture, housing quality, water and sanitation networks, and social development outcomes will require significant adjustments in the distribution iii of public funds. Additionally, the large sums currently allocated to sports and cultural events inflate spending on recreation, culture, and religion and should therefore be curtailed and closely monitored. 10. The government top priorities should be twofold: to create more fiscal space where possible, and to improve spending efficiency across the board. To create more fiscal space, the growing wage bill will have to by contained by improving long term human resources management in the public sector. In parallel, resource mobilization will have to shift from a model that relies mostly on the oil sector to one that increasingly relies on the entire economy, with a tax policy that generates sustainable tax revenue in an efficient way. Greater spending efficiency is critical to providing quality service delivery to the public. Efficient public services are not only key to protect the poor and the vulnerable from the impact of the crisis in the short term, but also to build the human capital necessary for diversified growth in the longer term. This requires the reallocation of funds to the social sectors, while at the same time ensuring that these funds deliver the maximum impact. 11. Spending efficiency and fiscal sustainability could significantly be enhanced by strengthening public financial management. Priorities are (i) addressing gaps in the regulatory framework, (ii) promoting institutional ownership of the reformed budget process, and (iii) enhancing accountability, transparency, and rigor in both the public procurement and public investment management systems. The introduction of program-based budgeting, the creation of a Public Procurement Regulatory Agency, and improvements in the quality and comprehensiveness of financial information are promising steps. These measures should be complemented by further efforts to increase transparency, engage a wider variety of actors from both the public and private sectors in the budget process, build the capacity of program managers, budget officers, financial controllers, accountants, and other staff involved in budget implementation, and develop and publish guidelines clarifying the selection and prioritization process for investment projects, as well as the appraisal criteria and monitoring requirements. Lastly, arrears monitoring and clearance mechanisms should be improved. Nonoil Revenue Mobilization 12. Low global oil prices since 2014 have sharply depressed public revenue. Oil revenue plunged from 17.5 percent of GDP in 2012 to 7.1 percent in 2015, while non-oil revenue remained broadly stable at about 14 percent of GDP. Consequently, total revenue fell from 30.2 percent of GDP in 2013 to 21.1 percent in 2015. Total revenue further declined to an estimated 17.1 percent of GDP in 2016 and is projected to rise only slightly to 17.6 percent in 2017, as oil prices marginally recover. 13. The government has adopted measures designed to enhance nonoil revenues, including the introduction of an electronic tax-declaration and payment (e-tax) platform, upgrades to the customs service information systems, and an increase in the frequency of physical customs inspections. However, serious deficiencies in tax and customs administration continue to undermine collection efforts. Priority challenges include improving the quality of taxpayer information, removing obstacles to voluntary compliance, developing compliance-risk management systems, improving reporting accuracy in tax and customs declarations, strengthening the administration and oversight of tax expenditures. iv 14. Domestic revenue mobilization will become increasingly important over the medium- to-long term, as taxes collected at the border—Gabon’s largest source of non-oil tax revenue—are set to decline. Revised EU trade agreements are expected to reduce revenue from tariffs and excise taxes on imports. To offset this decline, the government will need to improve the collection of direct taxes, including personal income tax, corporate income tax and property tax, as well as domestic value-added tax (VAT) and, potentially, domestic excise taxes. 15. Tax expenditures—including exemptions, allowances, credits, reduced rates, tax holidays and deferrals—substantially reduce public revenue and erode the tax base, and many tax expenditures lack a clear rationale. In 2014, domestic VAT, border tax, and CIT expenditures cost Gabon an estimated CFAF 315 billion, or 3.55 percent of GDP, in foregone revenue. Border-tax expenditures alone cost the government an estimated 2.35 percent of GDP each year. Domestic VAT expenditures cost the government an estimated 0.74 percent of GDP, while CIT expenditures cost about 0.46 percent of GDP. There is no evidence that tax exemptions have substantially increased foreign direct investment, and that the VAT and tariff exemptions granted as “measures to contain the cost of living” did not contribute significantly to controlling food-price inflation. Moreover, most of these tax expenditures violate rules of the Central African Economic and Monetary Community (CEMAC). 16. Weaknesses in tax administration compound inefficiencies in tax policy. The complexity of the tax system and the need for multiple, redundant contacts with revenue authorities discourage voluntary compliance. The launch of the e-tax platform has lessened but not eliminated the administrative burden on the 400 small and medium enterprises that used the platform in 2017, and the number of taxes that taxpayers can file and pay online continues to grow. 17. The government is considering plans to foster closer cooperation between the tax and customs agencies. More effective coordination would not eliminate the need to address systemic weaknesses in tax administration, but it could provide an opportunity to implement the necessary actions in parallel with complementary measures to control the cost of tax collection. Improving the articulation between the tax and customs administrations could help reduce administrative costs, simplify and improve service delivery, eliminate overlapping roles and responsibilities, and accelerate the integration of information systems. 18. The government has launched a wide range of revenue reforms as part of the PRE 2017-19, but the impact of these measures remains uncertain. The authorities expect to increase tax collection by about 1.2 to 1.4 percent of GDP over the medium term by completing a full inventory of tax expenditures, clarifying the legal basis for tax exemptions, allowances, credits, reduced rates, tax holidays, and tax deferrals, and specifying the authorities who can grant them. The government has appointed a committee to implement this process, and a “fiscal justice” operation was officially launched in February 2018. 19. The government should complement these efforts by adopting three additional medium-term reforms. First, the General Tax Directorate should develop compliance-risk indicators, classify taxpayers according to compliance risk, and introduce risk-based auditing. Second, the General Tax Directorate, ANUTTC, and local authorities should move forward with plans to strengthen the fiscal cadaster and to create comprehensive taxpayer files. Third, the v government should shift from profit-based to cost-based tax incentives and eliminate tax holidays to encourage longer-term investment. Cross-cutting Expenditure Measures 20. Shrinking the public-sector wage bill could yield significant savings. In 2017 payroll expenditures totaled 8.3 percent of GDP, a large share by the standards of peer countries. Several complementary reforms are required to better manage the wage bill. A more performance-oriented approach, combined with better controls on recruitment in line with ongoing budgetary reforms and the adoption of a new human resource management information system (HRMIS), could improve public administration and reduce the public-sector payroll. 21. Improving budget management and accounting could enhance oversight of public- sector recruitment. Creating budgetary items in line with the “BOP” (operational program budgets, budgets opérationnels de programme) budgetary nomenclature (classified by each mission program) could improve the accountability of program managers and financial controllers and strengthen incentives to enforce hiring controls. Developing a provisional job-management and staffing document could provide financial controllers with a short- or medium-term forecast of payroll and staffing changes. Establishing “prior approval” for individual positions that will not be filled without a vacant budget item and a monthly reporting requirement for using departmental remuneration credits would further bolster the integrity of the recruitment process. 22. Creating a HRMIS could greatly improve public service delivery while also reinforcing control over the wage bill. Current wage data are scarce and inconsistent. A HRMIS could not only consolidate compensation and employment figures, but also disaggregate the payroll by compensation type, sector, employment terms, and administrative classification. 23. Upgrading public investment management systems could enable the government to continue implementing its ambitious investment agenda in a context of fiscal consolidation. Three specific reforms could be effective: (i) revising the national infrastructure plan to ensure greater consistency between sector plans, incorporate regular reviews and updates, and link planning to a realistic allocation of estimated resources; (ii) systematically analyze the economic feasibility and financial sustainability of the highest-value projects; and (iii) more actively monitor the financial and nonfinancial performance of investment projects, beginning with the largest and most important projects in the national development plan. Social Sectors Key Challenges 24. Funding for social services tends to be insufficient and is often not guaranteed. Despite the significant increase in public health spending after the implementation of the national health insurance system in 2008, total expenditure on education, social protection, and health represent only 20 percent of the total budget; this is low compared to peers. Furthermore, allocations to social services vary from year to year and funding for non-contributory social protection programs is insufficient to meet Gabon’s social development objectives. Execution rates can be low and in some sectors, such as health, they have fallen to 70 percent for some categories. Fiscal adjustment should not be borne by these categories; on the contrary, allocation for social services should be vi secured and, provided that the necessary fiscal space is created and that adequate systems are improved, it should even be increased. 25. Within social sectors, allocative efficiency needs to be improved and spending needs to be progressive. The current allocation favors tertiary over primary services, and cities over rural areas. In education, most of the spending goes to the upper-secondary and tertiary levels, such as universities. In public health, too much funding flows to tertiary hospitals, which end up providing services which could be supplied at a much lower cost by primary and secondary facilities which lack adequate funding. In health, operations, maintenance, and public outreach are systematically underfunded, and there are large disparities in per capita spending across regions. Funds for social assistance have been used arbitrarily and cash transfers programs often fail to reach intended beneficiaries. 26. The management of and data on public services needs to be improved to address serious technical inefficiencies. Staff providing social services are not adequately managed or trained. Teachers in basic education and health care workers particularly require more and better training. Health care workers are overly concentrated in hospitals and urban areas and too few are in primary cares facilities which are key for local outreach programs to communities. Improving the quality of primary health care facilities is critical, as low-quality service provision pushes patients to go to hospitals. Social protection programs are also too fragmented and targeting systems are ineffective. Finally, there is a lack of quality data to support better social services delivery; this is a direct consequence of the absence of a recent household survey and of limited and poor-quality institutional data. Education Spending 27. Gabon’s public education outcomes are below what the country’s level of per capita GDP and public spending would predict. Despite significant progress at the primary level, an inadequate supply of human and physical capital and an excessive emphasis on higher education contribute to low completion rates and high repetition rates. Gabon spends less on education than many comparable upper-middle-income countries, and the overall allocation of public education spending in Gabon is regressive. While public spending at the primary level appears to be pro- poor, there is evidence that it is regressive at the secondary and tertiary levels, which tend to serve students from wealthier households. Although girls outperform boys in the overall education system, girls from very poor households are underrepresented in tertiary education. 28. The government should boost public education spending and increasing the share of spending devoted to primary and secondary education. The authorities could create additional fiscal space by improving the allocative and technical efficiency of education spending. At present, most public education spending goes to the upper-secondary and tertiary levels. Underspending at the primary and lower-secondary levels, and among technical and vocational education and training (TVET) institutions, reduces both the quantity and quality of basic education services. As many higher-education institutions would be viable without public support, increased private- sector involvement at this level could allow policymakers to reallocate scarce public resources to primary and lower-secondary education and TVET. vii 29. Investing in teacher training could improve the technical efficiency of education spending by reducing repetition and dropout rates. Building the human capital of teachers would improve to quality of basic education and enhance the coordination of curricula between education cycles, improving the rate at which students successfully transfer to upper-secondary and tertiary education. Better-trained teachers could also more accurately identify students at risk of dropping out and intervene as needed. Extending the school year and strengthening school governance could complement increased teacher training. 30. Concerted policy reforms will be necessary to improve secondary education, higher education, and TVET. The authorities should work to improve the governance of public TVET and higher-education institutions. Expanding the role of the private sector could help to increase the quality of higher education and TVET by aligning curricula with the needs of employers, establishing internship, and developing dual-training programs. Private education providers should be encouraged to invest further in upper-secondary and tertiary education within an appropriate regulatory framework. Given Gabon’s improving level of internet connectivity, the government should explore innovative education strategies such as distance-learning initiatives. 31. Measures to enhance public administration and personnel management would yield especially positive results in the education sector. Assigning staff to specific programs would help teachers focus on priority activities and improve the deployment of human resources. Developing a comprehensive education management information system backed by robust statistical data could provide vital information to guide future policy decisions. Social Protection Spending 32. Although Gabon’s social protection system contains all the necessary components to help households manage risk, it is inefficient, underfinanced, and covers only a small fraction of the poor. Non-contributory programs, including public health insurance and cash transfers, are essential to reduce poverty. While these programs received 61 percent of social protection spending in 2015, their current funding level is insufficient to meet the objectives of Gabon’s Human Investment Strategy (Strategie d’Investissement Human du Gabon, SIHG). Meanwhile, spending on pensions is rising, even though pension beneficiaries are disproportionately non-poor and much fewer than beneficiaries of non-contributory programs. 33. Funds for social assistance have been used arbitrarily, resulting in significant inefficiencies and inequities. Contrary to what is envisaged in the SIHG, subsidies account for 89 percent of expenditures devoted to supporting “economically weak Gabonese” (Gabonais économiquement faible, GEF) households, while cash transfers represent only 10 percent. No cash transfers were made to GEF households in 2015 and 2016, except for two small, short-term programs. However, widows, widowers, and persons with disabilities received monetary support in 2015 regardless of their income level. Even before 2015, cash transfers to GEF households were limited and irregular. In 2014, most safety-net programs were only available in Libreville and Estuaire Province, and only four out of fourteen programs reached the rest of the country. 34. Non-contributory programs are highly fragmented. The SIHG contains 19 different social protection programs designed for seven specific vulnerable groups. These programs range viii from cash and in-kind transfers to fee exemptions and water and electricity subsidies. Adopting a systemic approach in coordination with other sectors could improve implementation efficiency. 35. Better targeting would enhance the impact of non-contributory social protection programs on poverty reduction. To increase the share of resources that reach GEF households, the authorities should: (i) update the poverty line based on data from the 2017 household budget survey; (ii) create a new methodology to identify vulnerable households based on consumption levels and living standards; (iii) determine the instruments used to classify beneficiary households and verify their status; (iv) produce an implementation manual that clearly defines the responsibilities of different parties and describes grievance and redress mechanisms; (v) restructure the GEF database to include socioeconomic indicators of vulnerability; and (vi) ensure that the information for each household in the GEF database is unique and complete. 36. Consolidating and modernizing social transfer programs could boost their efficiency. The government should begin by merging programs of the same type: for example, the eight cash- transfer programs should be consolidated into one, and the “newborn kit” should be monetized and merged with the cash-transfer program. Leveraging modern technologies to transfer money, such as cell phones and ATM cards, could further increase technical efficiency. 37. Social protection programs should promote self-reliance and facilitate households’ transition out of poverty. The design of the current GEF database does not reflect the expectation that households will eventually escape poverty, and it does not contain the information necessary to reclassify households that do. The authorities should: (i) determine when GEF households are eligible for special assistance in transitioning out of poverty; (ii) ensure that households participating in entrepreneurship programs receive regular cash transfers that allow them to stabilize their consumption as they build microenterprises; (iii) provide frequent and regular technical and psychological assistance to beneficiaries; and (iv) develop a robust monitoring system based on objective indicators. 38. Regular, predictable funding for social protection programs is essential to achieve sustainable poverty reduction. The government should guarantee a minimum funding allocation each year for its social transfer budget. This would make it possible to implement a long-term cash-transfer program for poor households complemented by support to households transitioning out of poverty as well as specialized services for particularly vulnerable groups. 39. Improved public financial management processes, greater data availability, and enhanced institutional capacity would strengthen the efficiency and effectiveness of social protection policies. The authorities should produce more accurate budget and expenditure data, better align policies with program objectives, ensure more consistent reporting, and clarify the budgetary nomenclature to enable proper expenditure planning an evaluation. Conducting regular household surveys would greatly improve data quality. Finally, the government should consider managing the health-insurance system separately from the rest of the social safety net. Health Spending 40. While public health spending has increased substantially in recent years, Gabon’s expenditure levels remain low and volatile by the standards of comparable countries. Public ix health spending more than doubled between 2008 and 2015, driven by the rollout of the national health-insurance system, but Gabon still spends less on public health than its peers. Rising expenditure levels have not been matched by commensurate improvements in health indicators, underscoring the importance of prioritizing expenditure types that generate the most value for money. In addition, large annual fluctuations in public health spending as a share of the national budget highlight the government’s lack of a consistent strategy for managing the impact of revenue volatility. 41. Reorienting expenditures toward preventive care and public outreach efforts executed through primary and secondary care facilities could greatly enhance the allocative efficiency of health spending. Currently, about 80 percent of health resources are devoted to curative care, which contributes less than 10 percent to improvements in health indicators. By contrast, preventive care and public outreach account for only about 20 percent of health spending, yet they contribute more than 60 percent to improvements in indicators targeted under the 2017- 2021 National Health Development Plan (Plan National de Développement Sanitaire, PNDS). The distribution of health spending does not reflect either Gabon’s epidemiological profile or the demographic distribution of its population. 42. As Gabon strives to meet its objectives for the health sector in a context of fiscal consolidation, policymakers will need to improve the efficiency and distributional equity of sectoral resources. To achieve its goals of the PNDS, the government must: (i) operationalize health districts; (ii) improve referral systems; (iii) strengthen pharmaceutical management; (iv) develop transparent criteria for allocating resources; (v) establish oversight mechanisms for the implementation of investment plans; and (vi) create more efficient budget-management and incentive systems. Other priority reform measures include promoting greater financial decentralization, expanding copayments for preventive and outreach services, allowing health facilities to deposit user fees into dedicated bank accounts, and increasing copayments for primary care to help match the supply of health services with demand. The authorities should also develop a financing strategy for the health sector that reflects the objectives of the PNDS. 43. Improving the quality of primary care facilities could reduce costs by encouraging patients to seek treatment at the appropriate level of the health system. Patients routinely bypass underfunded and understaffed primary care clinics in favor of hospitals. Strengthening the primary care system could alleviate the burden on hospitals and lower overall costs, both for the health system and for patients. 44. Adopting innovative techniques such as performance-based financing (PBF) could help improve the quality of care, especially at the primary level and address some challenges related to quality, efficiency and equity. The international experience has demonstrated the value of PBF in strengthening both preventive and curative health services, enhancing the managerial and supervisory activities of health district teams, and encouraging citizen engagement in the delivery of health services. PBF can facilitate the transfer of resources to the point of service delivery and improve accountability for the utilization of these resources. PBF is a reform strategy, which aim to improve the quality, the efficiency and the equity of health system. Gabon will soon launch a PBF pilot in four health regions and two health departments Libreville, and the authorities should closely monitor its results. x 45. The government should increase funding for health-facility maintenance and public outreach and convert standard budget lines to performance-based payments. The current health budget is based on distributing the available resources, and it does not adequately reflect demand for services or health outcomes. Consequently, operations, maintenance, and public outreach are systematically underfunded, and there are significant disparities in per capita health spending across regions. Introducing PBF principles, including direct transfers to service providers based on results achieved, would help increase value for money in health spending. 46. Reorganizing tertiary facilities could enhance the efficiency of the hospital system. Large-scale investments in hospital construction have resulted in an excess of hospital beds, and many hospitals have occupancy rates as low as 40 percent. The authorities could increase efficiency by merging hospitals and converting underutilized facilities to primary care centers for maternal health, child health, and emergency services. In addition, many regional hospitals lack appropriate specialists, who could be redeployed from the capital. 47. Higher-quality data would support more informed decision-making. The government relies on irregular household surveys to monitor progress in the health sector, and while institutional information systems exist, they produce limited data. The Ministry of Health has established an information management unit, and efforts are underway to expand population-based data and create geographical information systems. Greater investment in information technology will be necessary to enhance sectoral oversight, and facility-based information systems will be crucial to monitor performance and assess local needs for pharmaceuticals and medical supplies. While CNAMGS has an online platform to register beneficiaries, an online system for managing claims could improve efficiency and facilitate timely reimbursement. The World Bank’s ongoing e-Gabon project is expected to substantially improve Gabon’s health-information systems. Table 1: Overview of Key Recommendations Proposed Reforms Timeframe Expected Impact(s) Institution(s) Public Expenditures and Public Financial Management Reallocate resources away Short to Improved allocative • Min. of Budget from public administration and medium term efficiency of expenditures • Min. of Economy toward economic affairs, • Relevant line Ministries housing, health, education, and social protection Implement the new Short term Stronger public • Min. of Budget procurement code and procurement process • Public procurement empower the public accountability agency procurement agency, notably by providing necessary resources Finalize the update of the PFM Short term Better budget execution • Min. of Budget legal framework by issuing the • Line Ministries remaining order and regulations that will clarify roles, and budget execution procedures Intensify hands-on-support on Short term More ownership of the • Min. of Budget the new budget reform for budget reform among staff • Line Ministries program managers, budget officers, financial controllers, and accountants xi Proposed Reforms Timeframe Expected Impact(s) Institution(s) Public Investment Management (PIM) Update the National Short term Greater consistency • Min. of Budget Infrastructure Plan to ensure between sector plans, and • Min. of Infrastructure, greater consistency between planning linked to realistic Public Works and sector plans, mandate regular estimates of medium- to Territorial Development reviews and updates, and take long-term resources • National Agency for into account realistic estimates Public Works (ANGTI) of long-term resources needed Systematically analyze the Short term Better project quality at • Line ministries proposing economic feasibility and entry and at term the project with technical financial sustainability of the infrastructure quality oversight from the highest-value projects ANGTI Introduce multi-year Medium term Continuity in public • Min. of Budget management processes for investment funding (no • Min. of Economy investment projects more ‘stop’ and ‘go’ and • ANGTI thus lower costs) Improve physical monitoring of Short term Lower cost-overruns, • Min. of Budget on-going projects and ensure that reduced project • Min. of Economy financial disbursements and implementation delays, and • ANGTI corresponding stages of project better infrastructure quality completion match Develop and publish guidelines Medium term Improved investment • Min. of Budget clarifying the project-selection projects pipeline, with • Min. of Economy and prioritization process, along projects closely aligned to • ANGTI with appraisal criteria and strategic priorities and monitoring requirements better monitoring Wage bill Improve hiring controls for a Short term More efficient management • Min. of Civil Service and more efficient management of of wage bill Administrative Reform the wage bill, if needed through New hiring procedures • Line Ministries a recruitment freeze for a rolled out period of twelve months, except for priority ministries Specific measures include: (i) the creation of budgetary items in line with program budgeting nomenclature (classified by each mission program); (ii) creation and submission of a provisional job management and staffing document; (iii) the establishment of a ‘prior approval’ function for individual recruitments; and (iv) the implementation of a monthly reporting on use of departmental remuneration credits Develop a Human Resources Medium term Better wage-bill • Min. of Civil Service and Management Information management Administrative Reform System that can monitor (i) pay composition; (ii) areas of employment; (iii) employment type: and (iv) worker type xii Proposed Reforms Timeframe Expected Impact(s) Institution(s) Create three-year strategic Short to Better control of • Each ministry with staffing plans, including medium term recruitments, and more ad support and oversight employment ceilings for each equation between staffing from the Ministry of Civil function at the departmental level needs and personnel Service and numbers and skills Administrative Reform Non-oil revenue mobilization Implement Economic Recovery Short to Increased non-oil revenue • Min. of Economy, esp. Plan measures, particularly on medium term mobilization Tax and Customs tax expenditure rationalization, Directorates improved taxpayer • Revenue Authority once information, and tax and operational customs IT systems upgrades Develop risk indicators and Short term Improved compliance • Tax and Customs classify taxpayers according to management and Directorates risk to facilitate risk-based potentially higher revenue auditing Shift from profit-based to cost- Medium to Higher investment and • Tax Directorate based tax incentives and long term corporate income tax intake eliminate tax holidays to encourage long-term investment. Move forward with plans to Medium term Better taxpayer (and other) • Tax Directorate improve the fiscal cadaster information • ANUTTC registry and taxpayer files • Local authorities Education Increase public education Short to Better education outcomes • Min. of Budget spending medium term • Min. of National Education • Parliament Increase the share of education Short to Improved primary and • Min. of Budget spending allocated to primary medium term secondary education • Min. of National and secondary education quality and improved Education access to lower secondary • Parliament education in under-served areas Improve the governance of Medium term Narrower employment and • Min. of Labor, public TVET and higher skills gap Employment, Technical education institutions and and Vocational Training, leverage the private sector in and Youth Integration; the financing and/or provision • Min. of Higher Education, of secondary, TVET, and Scientific Research, and higher education Management Training Invest more in teacher training Short to Better quality of teaching • Min. of National medium term Education Improve the planning, budgeting, Medium term Better PFM in the • Min. of National execution, and reporting of education sector, higher Education education spending; modernize efficiency of sector • Min. of Budget human resource management and spending, and better attach staff to programs; and education data quality develop a reliable education management information system Social Protection xiii Proposed Reforms Timeframe Expected Impact(s) Institution(s) Improve the targeting of non- Short to Greater safety nets poverty • Min. of Social Affairs contributory services medium term reduction impact • CNAMGS, FNAS Consolidate and modernize Medium term Higher efficiency and more • Min. of Social Affairs social transfer programs adequate programs • CNAMGS, FNAS, CNSS, (through lower CPPF fragmentation) Promote self-reliance as an Medium term Better exit-strategies and • Min. of Social Affairs integral part of the safety-net faster transition out of • CNAMGS, FNAS system to help poor households poverty for program graduate out of poverty beneficiaries Improve PFM in the sector and Short to Improved, more reliable • Min. of Social Affairs secure funding for safety nets and medium term and consistent service • Min. of Budget non-contributory health delivery quality insurance Health Reorient health expenditures Short to Higher allocative efficiency • Min. of Health toward preventive and medium term of health spending and • Parliament promotional care in primary improved health outcomes and secondary care facilities Operationalize health districts, Medium term Improved health spending • Min. of Health improve referral systems, technical efficiency and strengthen pharmaceutical health outcomes management, establish oversight mechanisms to ensure investment plans are fully implemented, and create more efficient budget- management and incentive systems, such as output-based payments Rebalance both capital and Medium term Higher allocative efficiency • Min. of Health current expenditures away of health spending and • Parliament from hospitals and toward improved health outcomes primary care facilities Ensure that investments in the Medium term Higher allocative efficiency • Min. of Health health sector are aligned with of health spending and population density and local improved health outcomes health needs Explore options to raise Medium term More financial resources • Min. of Health additional revenue for health, for health (and better • Min. of Budget including greater financial outcomes) decentralization and expanded copayments for preventive and outreach services Note: Most critical recommendations are bolded. Short term refers to actions that are feasible within the current fiscal year, Medium term to actions in the next one or two fiscal years. xiv 1. FISCAL POLICY AND PUBLIC FINANCIAL MANAGEMENT Introduction 48. This chapter describes the critical fiscal policy and public financial management (PFM) challenges faced by the government. The chapter discusses the impact of the sharp decline of global oil prices since 2014 on Gabon’s economy and fiscal position. The chapter further assesses the government’s fiscal-adjustment measures and evaluates their impact on Gabon’s macroeconomic and fiscal outlook. It then examines the overall quality of PFM in Gabon and discusses ways in which PFM can reinforce the sustainability of fiscal policy. The chapter concludes with a set of short and medium-term recommendations to enhance PFM. Context 49. In 2009, Gabon launched the Emerging Gabon Strategic Plan (Plan Stratégique Gabon Emergent, PSGE), an ambitious program designed to promote economic and social transformation by diversifying the sources of growth and investing in infrastructure. The PSGE is based on three programmatic pillars: (i) environmental protection, (ii) industrial development, and (iii) the growth of the service sector. The program targets economic growth and diversification efforts to the secondary and tertiary sectors. 50. Despite substantial infrastructure investment over the past decade, the business climate remains poor. Gabon’s global rank in the World Bank’s Doing Business report has not improved since 2015, and in the 2017 edition, Gabon fell three places to 164th out of 190 countries. Although the government has made large investments in energy, access to electricity remains one of the most important obstacles to doing business. 51. Between 2010 and 2014, Gabon largely failed to capitalize on a strong macroeconomic environment, and despite this period of relatively robust growth, high rates of unemployment, poverty, and inequality persist. In 2005, one in three Gabonese lived below the national poverty line of 35,778 Central African CFA francs (FCFA) per capita per month, or roughly US$2.5 per day, and poverty rates were substantially higher in rural areas (44.6 percent) than in urban centers (29.8 percent). While no more recent official statistics for poverty and inequality are available, World Bank staff estimate little or no improvement in these figures over time. In 2010, the unemployment rate also remained high at 20.4 percent, or at 27 percent if discouraged workers are included. Unemployment rates were especially high among workers aged 19-24 (35.7 percent) and among women (60 percent). Fiscal Trends: 2006-2017 Public Spending and Economic Growth 52. During 2010-2014, high oil prices facilitated the implementation of the PSGE, and a surge in public investment and consumption accelerated growth. Following a period of generally low and volatile growth, marked by significant contractions in 2006 and 2008, the launch of the PSGE in 2010 drove a sustained improvement in Gabon’s economic performance. The GDP growth rate accelerated to an average of 5.9 percent per year during 2010-2014 (Figure 1). Public 1 investment increased rapidly as a share of GDP, rising from 4.0 percent in 2009 to 9.7 percent in 2012 (Figure 2). Meanwhile, the contribution of total investment to GDP growth expanded (Figure 3). High oil prices also fueled growth in sectors such as construction and services (Figure 4), and the aggregate growth of the nonoil economy rose to an average of 5.6 percent per year over 2010- 2014. Figure 1: Real GDP Growth Rate, 2007-2017 (%) Figure 2: Public Investment (% of GDP) 15.0 16 14 10.0 12 6.3 6.3 7.1 5.3 5.5 5.0 4.4 3.9 10 1.7 2.1 8 0.0 0.8 -2.3 6 -5.0 4 -10.0 2 0 -15.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Oil sector Non-oil sector Total GDP Gabon SSA CEMAC Figure 3: Contribution to GDP Growth, Demand Side Figure 4: Contribution to GDP Growth, Supply Side (%) (%) 10.0 25.0 20.0 8.0 15.0 6.0 10.0 4.0 5.0 6.3 6.3 7.1 5.3 5.5 2.0 4.4 3.9 1.7 2.1 0.8 0.0 0.0 -2.3 20072008200920102011201220132014201520162017 -5.0 -2.0 -10.0 -4.0 -15.0 -6.0 -20.0 -8.0 Services Private consumption Public consumption Industry (excl. manufacturing, construction, EI) Construction Investment Net exports Manufacturing Extractive industries Total GDP Agriculture Source: National Authorities, International Monetary Fund. 53. In 2014, global oil prices collapsed, sharply reversing trends in investment and growth. Although investment continued to expand in 2014, the government began cutting investment outlays to adjust to lower oil revenues. Capital expenditures fell from an average of 36.2 percent of total spending between 2010 and 2014 to just 22.1 percent in 2016. Public investment dropped to 4.2 percent of GDP in 2015, in line with its pre-PSGE levels. As investment 2 levels plunged, the annual GDP growth rate slowed to 2.1 percent in 2016 and to an estimated 0.8 percent in 2017. 54. The oil sector contracted in 2016, as global oil prices failed to recover. Oil producers attempted to optimize the output of existing fields, but a lack of investment caused annual output growth to fall from 4 percent in 2015 to -2.7 percent in 2016. In 2017, oil prices remained more than 50 percent below their 2011 level, and other commodity prices dropped by about 31 percent over the same period. The industrial sector stagnated in 2016, and the decline in public spending combined with significant public-sector arrears slowed the growth rate of services from 5.2 percent in 2015 to 3.2 percent in 2016. The relatively small agricultural sector bucked the general trend and performed well over the period, as an already robust growth rate of 8.2 percent in 2015 accelerated to 11.9 percent in 2016. 55. The oil sector’s declining contribution to economic growth reflects a downward trend in production that is unlikely to be reversed. The oil sector contracted for most of the past decade, as existing fields aged, investment dwindled, and no major new discoveries were made. With production continuing to weaken and prices projected to remain low, oil is unlikely to drive Gabon’s future growth to the same extent that it has in the past. For Gabon to achieve its development objectives, it will need to leverage new sources of economic growth and public revenue. Fiscal Policy 56. Prior to the launch of the PSGE, debt levels were low, but fiscal and debt indicators began to deteriorate after 2010. Before 2010, the government consistently ran budget surpluses, and debt levels fell to a low of 20.1 percent of GDP in 2008 - below the authorities’ ceiling of 35 percent and well below the CEMAC recommendation of 70 percent. However, the implementation of the government’s ambitious investment program narrowed the surplus to 0.6 percent of GDP in 2011, as nominal revenues grew slowly while expenditure levels rapidly increased. The surplus turned to a deficit of 3.1 percent of GDP in 2013, as public spending hit a peak of 34.7 percent of GDP (Figure 5). 3 Figure 5: Key Fiscal and Debt Indicators, 2006-2017 (% of GDP) 70.0 64.2 61.8 60.0 50.0 44.7 39.6 39.2 40.0 34.1 31.1 30.0 26.0 20.1 21.3 21.4 21.4 20.0 8.7 10.7 7.9 5.9 6.0 10.0 5.4 1.8 0.6 0.0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 -10.0 -3.1 -1.0 -5.0 -2.3 Fiscal balance (commitment basis) Public debt Source: National Authorities, International Monetary Fund. 57. The government’s fiscal position worsened significantly from 2014, as oil prices fell while public investment continued to surge. Fiscal deficits were recorded in 2015 and 2016, at 1.0 and 5.0 percent of GDP, respectively. Although the government began to rein in spending in 2014, revenues fell faster than expenditures, widening the fiscal deficit. Fiscal adjustment focused on capital investment, but the rigidity of the wage bill—which had grown by 70 percent between 2010 and 2015 and which now represents about 36 percent of total public spending—kept expenditure levels high. The public debt stock grew dramatically, due in part to the inclusion of government arrears to suppliers that had accumulated as oil prices fell. In 2016, the debt stock reached 64.2 percent of GDP, breaching the authorities’ policy ceiling. 58. Despite a long history of fiscal surpluses, the government lacked adequate fiscal buffers or other countercyclical measures to mitigate the impact of the crisis. In 2014, Gabon’s fiscal savings amounted to 7.2 percent of GDP; by 2016, they had fallen to 2.4 percent. Of this, 1.7 percentage points of GDP was invested in the Fund for Future Generations and could not be used to finance the deficit. While the fund was created to accumulate excess oil revenues to finance future spending, the protracted slump in oil prices has essentially stopped it from accruing additional resources. 59. Non-oil revenue, which had remained relatively stable between 2010 and 2014, briefly surpassed oil revenue in 2014, then declined. Oil drove evolution of total revenue over the 2000- 2009 period, while nonoil revenue remained broadly unchanged (Figure 6). Non-oil revenue rose to a peak of 15.0 percent of GDP in 2014-2015, then fell to 10.8 in 2016 as the economic crisis spread from the oil sector to other parts of the economy and tax and customs staff went on strike. 1 With the oil sector facing a bleak medium-term outlook, improving non-oil revenue mobilization will be vital to reinforce fiscal sustainability. 1 See Chapter 2 for further details on the tax and customs staff strikes. 4 Figure 6: Government Revenues and Expenditures, 2006-2017 (% of GDP) 40.0 35.0 30.0 25.0 10.7 10.3 11.0 15.0 11.4 14.8 15.0 20.0 11.9 9.6 15.0 14.0 12.0 10.6 10.0 19.1 19.6 19.2 16.1 14.6 16.6 14.7 13.9 13.9 5.0 7.1 5.1 7.0 0.0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Oil revenue Non-oil revenue Total expenditure Source: National Authorities, International Monetary Fund. 60. Public debt levels have increased sharply since 2013 as domestic arrears and statutory advances from the regional central bank have accumulated. The public debt stock, including arrears, rose from 20.1 percent of GDP in 2008 to 64.2 percent in 2016, far above the strategic threshold of 35 percent set by the government, but still marginally below CEMAC’s recommended ceiling of 70 percent. In 2016, domestics arrears and total arrears 2 reached 7.7 percent and 9.5 percent of GDP, respectively while statutory advances from the regional central bank stood at 5.4 percent. Moreover, as the implementation of the PSGE increased the government’s financing needs, it turned to international capital markets, issuing Eurobonds worth US$1.5 billion in 2013 and US$500 million in 2015. Fiscal Outlook: 2017-2020 61. The government needs to maintain fiscal discipline to weather the current crisis; this will increase the fiscal space available for investment. The authorities requested IMF support in 2016 and signed a program in June 2017. Implementation is proceeding on schedule, and early progress in improving the fiscal and external balances is encouraging. Significant expenditure consolidation could, however, negatively impact short-term growth prospects through a significant multiplier effect. The IMF program will run from 2017 to 2020 and provides US$628 million in total financing. Under the agreed framework, public spending will fall to about 19 percent of GDP by 2018, a 3 percentage-point decline from its 2016 level. Reduced spending will be necessary to curtail the deficit and contain the debt stock, even as the government strives to increase revenue. 3 2 Official arrears numbers do not include all public sector arrears: at the time of writing, an audit of the arrears stock was ongoing 3 Revenue measures are discussed in detail in Chapter 2. 5 Figure 7: Recent and Projected Trends in the Fiscal and Debt Profile, 2010-2020 (% of GDP) 70.0 Historical data IMF program data 64.2 62.7 61.8 61.6 58.0 60.0 50.0 44.7 40.0 34.7 34.1 31.1 30.0 24.0 22.8 24.7 23.8 21.3 21.4 21.4 22.1 22.1 19.8 19.1 19.1 18.6 20.0 10.0 0.0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 -10.0 Fiscal balance (commitment basis) Public debt Total revenue Total expenditure Sources: National Authorities and the IMF. 62. Fiscal consolidation needs to be implemented carefully to limit potentially negative effects on growth, employment and poverty. The Economic Recovery Plan defines the government’s fiscal reform strategy and policies. While fiscal consolidation is the top priority, the authorities need to ensure that economic diversification continues while also rebuilding the country’s fiscal buffers. The consolidation process should address three challenges simultaneously: (i) adjusting the fiscal stance to maintain sustainability in a context of lower revenue; (ii) building the foundation for economic diversification over the medium term, and (iii) protecting poor and vulnerable households from the adverse impacts of reduced public spending. 63. To ensure that fiscal policy effectively stabilizes the business cycle and catalyzes inclusive growth, the quality of public expenditure quality needs to improve. The current crisis presents a valuable opportunity to sustainably improve fiscal policy. To achieve this, the government needs to expand available fiscal space by increasing revenue mobilization and enhancing the efficiency of tax administration; the government also needs to maximizing the impact of existing resources. The size and allocation of the public investment budget should be carefully assessed, and important expenditure categories, such as the wage bill, should be reviewed to identify potential savings. Allocative efficiency can be improved by more accurately targeting priority sectors and objectives, such as social development and economic diversification. Finally, improving PFM processes will be vital to strengthen fiscal discipline, improve resource allocation, and enhance financial transparency. Public Financial Management 64. The government is modernizing its PFM system via the enforcement of a program- based budgeting approach, which lays the groundwork for greater fiscal discipline and transparency. However, serious deficiencies in the standard procedures for budget execution have forced the authorities to rely extensively on exceptional procedures. Inadequate budget-execution procedures and systems have also led to the mismanagement of arrears. In addition, the prevalence of single-source procurement weakens the efficiency of public service delivery. 6 The Legal and Institutional Framework for PFM 65. Since 2015, the government has extensively reformed the legal framework for PFM. The authorities formally adopted CEMAC PFM directives via the new Organic Law on the Laws of Finance and Budget Execution (Loi Organique relative aux Lois de Finances et à l’Exécution du Budget, LOLFEB). Program-based budgeting has been in place since 2015, and resource allocations are now based on each line ministry’s core mandate and programs, which are now underpinned by clearly defined objectives and performance indicators. Decree 236 of 2016 established the budget nomenclature, which complies with current CEMAC directives and reflects international PFM standards based on the IMF’s 2001 Government Finance Statistics Manual. The accounting nomenclature has also been overhauled as part of the implementation of the new state accounting plan, and a new book of accounting standards was produced in 2014. The expenditure system in place since 2013 includes four classifications: (i) programmatic spending, (ii) administrative spending, (iii) economic spending, and (iv) functional spending, though the latter has not yet been operationalized. These reforms have redefined the role of stakeholders in the PFM process and reorganized the services of the Ministry of the Budget and Public Accounts, representing a significant step toward making the budget process more credible and clearly linking public policies with their respective objectives and resource envelopes. 66. These reforms are being implemented in a context where budgetary control and execution systems still suffer from persistent weaknesses. The implementing regulations for the LOLFEB have not yet been drafted or updated. Moreover, the normative framework for expenditure execution has not been stabilized, which weakens internal controls and increases risks at a time when new actors—programs managers, operating budget officers, etc.—are assuming roles in the budget-execution process. In addition, the government’s frequent recourse to exceptional procedures and extra-budgetary operations remains a serious challenge. Fiscal Discipline 67. The strongest components of Gabon’s PFM system are its budget-classification, the budget-formulation process, the medium-term budget framework, and debt-management mechanisms. Budget formulation includes discussions between the Ministry of the Budget and Public Accounts and the line ministries held according to a predetermined calendar, which enables the timely submission of the draft finance bill to Parliament. The Annual Performance Plan (Plan Annuel de Performance, PAP) and the Annual Performance Report (Rapport Annuel de Performance, RAP) classify programs and provide the necessary documentation to clarify the budget structure. Budget discussions are based on medium-term expenditure ceilings and forecasts defined at the beginning of the budget-preparation process. Finally, the production of comprehensive, up-to-date debt information and the formulation of a single action plan supports robust debt management. 68. However, weaknesses in budget control and execution, as well as the absence of fiscal risk management, create opportunities to deviate from the parameters of the budget law. Internal controls are based on a legal framework, but their implementation is uneven, which results in erratic budget management and encourages the use of extra-budgetary practices. Arrears management is also problematic, as not all arrears are recorded in the public accounts. However, a comprehensive arrears assessment is currently underway as part of the IMF program. Internal 7 and external audits are still excessively focused on compliance control, and recommendations do not adequately influence future practices. Even in the absence of economic shocks, the composition of executed expenditures often deviates substantially from initial budget allocations, which in turn tend to be inconsistent with the medium-term expenditure framework. Strategic Resource Allocation 69. The necessary mechanisms are in place to ensure the efficient use of public resources. Gabon’s Parliament reviews and discusses a detailed assessment of revenues and expenditures, though this assessment is limited to a single fiscal year. Moreover, collected revenues are transferred relatively quickly to the Single Treasury Account. 70. Weaknesses in economic forecasting, budget strategies, and basic cash management reduce the predictability of resource allocation. The forecasts and budget strategies used to prepare the budget are not based on sufficiently detailed assumptions—particularly in terms of measuring the fiscal impacts of economic trends—to establish reliable revenue and expenditure trajectories. The absence of in-year fiscal reporting does not allow for an adequate year-end review process. Many adjustments (e.g., transfers from one budget line to another) are made during the fiscal year without applying LOLFEB rules for executing credit transfers within the budget. In a context of tight liquidity constraints and depressed revenues, the inadequate implementation of the commitment plan negatively affects cash management. Efficient Service Delivery and Financial Transparency 71. The introduction of the PAP and the RAP could significantly improve public service delivery. Beginning in 2015, the program-based presentation and execution of the budget has compelled the authorities to structure their public service strategy according to targets backed by measurable indicators. While there remains substantial scope to improve service delivery, the government has established a useful framework by which to measure performance. Following the same structure as the PAP, the RAP presents progress on performance indicators and provides explanations for failures to meet targets. 72. However, insufficient transparency in public procurement and public investment management undermines service delivery. The creation of the Public Procurement Regulatory Agency is a positive step, but over 70 percent of public procurement is still conducted through single-source selection. Moreover, the agency is not yet fully operational, which inhibits its ability to strengthen accountability in the procurement chain. In addition, information on investment management is limited: there is no comprehensive and consolidated list of major investment projects, few studies are conducted, and project-selection criteria are not published. 73. Financial information is often unreliable, incomplete, and not easily accessible. Despite progress in improving the quality of the annual financial statements, which are now presented in a format that reflects international best practices, the accuracy and comprehensiveness of the underlying financial data are limited. In addition, implementing agencies 4 do not provide 4 These include the Agence Nationale des Grands Travaux, the Agence National des Infrastructures Numériques et des Fréquences, and the Agence Nationale de Promotion des Investissements, inter alia. 8 systematic, organized feedback. Policymakers lack a clear, up-to-date assessment of service quality, while citizens have difficulty accessing timely budget information. 74. No mechanisms are in place to allow direct public participation in budget management. Public participation is limited to representation in the legislature, and public information on budget-related issues is limited government websites and press reports. In 2016, the internet access rate in Gabon was about 10 percent, or 181,604 people of a population of about 1.8 million 5. While the program-based budgeting approach has made budget documents more understandable and comprehensive, public access to budget information remains limited. Recommendations 75. The government could significantly improve PFM and strengthen financial sustainability. In order to do so the following reforms will be critical. (i) Addressing gaps in the regulatory framework: the Ministry of the Budget and Public Accounts should complete the regulatory framework by issuing the implementing regulations for the LOLFEB, which should clarify the roles and responsibilities of all actors involved in the budget-execution process. (ii) Promoting institutional ownership of the reformed budget process: intensifying administrative and technical support to program managers, budget officers, financial controllers, accountants, and other staff would facilitate the implementation of the reformed budget system. (iii) Enhancing accountability in public procurement: the Ministry of Economy should enforce the new procurement code and operationalize the Public Procurement Regulatory Agency. 5 www.gabonmediatime.com/taux-dacces-a-internet-le-gabon-compte-181-604-connectes/ 9 2. NON-OIL REVENUE MOBILIZATION Introduction 76. Gabonese policymakers have shifted focus to mobilizing non-oil revenues, and this chapter therefore only addresses non-oil revenue mobilization. 6 It assesses revenue collection performance—examining especially tax revenue levels, trends, and composition over 2010- 2017—and evaluates tax expenditure administration and monitoring, and other key tax administration issues (taxpayer information quality, taxpayer compliance, compliance risk management, and reporting accuracy). 7 The chapter considers previous, ongoing, and proposed tax policy and tax administration reforms , and both their individual and aggregate impacts. It concludes by recommending additional measures to strengthen tax policy and administration and boost non-oil revenue. 77. The Gabonese tax (and customs) system is governed by the general tax code, and CEMAC customs code and regulations. Special provisions and exceptions to the standard tax and customs regimes are detailed in the mining code, the forestry code, the investment code, the tourism investment act, and the hydrocarbon code, as well as annual initial and revised budget laws. The General Tax Directorate (Direction Générale des Impôts, DGI) levies internal direct and indirect taxes, while the General Customs Directorate (Direction Générale des Douanes et des Droits Indirects, DGDDI) collects revenue at the border, including customs duties and value-added tax (VAT) on imports. 78. The government has launched several policy initiatives in recent years to boost non- oil revenue, foster investment, and limit air pollution. These policy measures include an import ban on used cars older than three years, an export ban on unprocessed timber, a new electronic tax declaration and payment (e-tax) platform, more frequent physical customs inspections, and DGDDI information technology (IT) systems upgrades. The authorities have also eliminated import duties and VAT on a wide range of staple food imports to ease inflationary pressure on food prices in 2009-2010. Generous tax incentives were also made available to companies setting themselves up in the Nkok SEZ or in the Port-Gentil Free-Trade Zone, including a 10-year corporate income tax holiday, total tariff exemption on imported capital equipment and parts, total VAT exemption for the first 25 years of operation, unlimited and tax-free profit repatriation, and flexible labor laws for seven years to facilitate the employment of foreign workers. 8 79. The Gabonese government is currently implementing additional measures to boost non-oil revenue as part of the 2017-2019 Economic Recovery Plan (Plan de Relance Economique, PRE 2017-19). Total fiscal revenue has decreased considerably since oil prices collapsed in 2014, falling from 30.2 percent of GDP in 2013 to 21.1 percent in 2015. The PRE 2017-19 calls for an easing of the ban on used vehicle imports; the introduction of new manganese, 6 The chapter does not formally assess the overall tax gap, tax efficiency or equity. It does not address the redistributive impact of taxes, nor does it provide a systematic and global assessment of the extent to which taxation distorts investment, consumption, and savings. It does, however, briefly touch on all of these issues. 7 The section on tax administration loosely follows the Tax Administration Diagnostic Assessment Tool (TADAT) methodology. It does not cover all performance criteria and indicators, but focuses instead on the main areas of analysis and indicators identified under the framework most relevant in the Gabonese context. 8 IMF, 2014 Article IV consultation staff report, IMF country report No. 15/47. 10 gold, and gold dust export taxes; a reduction in the number of items eligible for VAT exemptions; improvements to the fiscal cadaster; and measures to boost tax compliance and fight tax evasion (e.g., the adoption of a risk-based tax auditing strategy). Revenue Levels, Trends, and Composition 80. The collapse of global oil prices in 2014 drove a sharp decline in public revenue. While non-oil revenue remained relatively stable, oil revenue plunged from 17.5 percent of GDP in 2012 to 7.1 percent in 2015, even though crude oil production rose in 2015 (Table 2). Meanwhile, non- oil revenue remained broadly stable at about 14 percent of GDP despite the adoption of important revenue reforms, such as the introduction of the e-tax declaration and payment platform, upgrades to the customs service information systems, and an increase in the frequency of physical customs inspections. Consequently, total revenue fell from 30.2 percent of GDP in 2013 to 21.1 percent in 2015. Total revenue further declined to an estimated 17.1 percent of GDP in 2016 and is projected to rise only slightly to 17.6 percent in 2017, as oil prices marginally recover. An expected increase in non-oil revenue in 2017, notably on account of PRE measures, failed to materialize. Instead, non-oil revenue fell by 2 percentage points of GDP between 2016 and 2017, due in part to strikes by tax and customs personnel, which disrupted revenue collection. Table 2: Government Revenue by Source, Gabon, 2010-2017 (% of GDP) 2010 2011 2012 2013 2014 2015 2016e 2017f Total Revenue 25.8 29.0 30.1 30.2 26.1 21.1 17.1 17.6 Of which: Non-oil 11.9 13.1 12.6 14.7 14.6 14.0 12.0 10.6 Of which: Non-oil tax revenue 11.3 12.1 12.0 14.0 13.3 12.3 10.8 9.9 Of which: Oil revenue 13.9 15.9 17.5 15.5 11.5 7.1 5.1 7.0 Tax revenue 13.6 15.1 14.2 16.1 14.9 12.8 10.8 10.1 Taxes on income, profits, and capital gains 5.6 7.1 5.8 6.1 5.9 5.2 3.6 3.7 Of which: non-oil 3.3 4.0 3.5 3.9 4.3 4.7 3.6 3.5 Corporate income tax 4.2 5.6 4.3 4.7 4.6 3.5 2.1 2.5 CIT oil companies 2.3 3.1 2.3 2.2 1.7 0.5 0.0 0.2 CIT non-oil companies 1.9 2.5 2.0 2.5 2.9 3.0 2.1 2.2 of which mining (manganese) 0.0 0.9 1.0 1.2 1.0 1.1 0.0 0.0 Personal income tax 1.4 1.5 1.5 1.5 1.4 1.7 1.5 1.2 Domestic taxes on goods and services 2.3 2.3 2.1 3.2 3.2 2.0 2.7 2.5 General sales, turnover, or value-added taxes 1.3 1.7 1.4 2.5 2.4 1.2 1.9 1.9 Other taxes on goods and services 1.0 0.6 0.7 0.8 0.8 0.8 0.8 0.6 Taxes collected at the border/DGDDI 5.1 4.4 4.6 4.6 4.0 4.2 3.3 3.1 Other taxes 0.6 1.4 1.7 2.2 1.8 1.4 1.2 0.8 Non-tax revenue 12.2 13.8 15.9 14.0 11.2 8.4 6.3 7.5 of which non-tax oil revenue 11.5 12.8 15.2 13.3 9.8 6.6 5.1 6.8 Other non-oil revenue 0.6 1.0 0.7 0.7 1.4 1.8 1.2 0.7 Source: Gabonese authorities and IMF staff projections. Note: Taxes on goods and services exclude VAT on import (which is instead accounted for in the DGDDI total. 81. Revenue growth was also negative, negligible, or barely positive over 2013-15 for most CEMAC and non-CEMAC comparator countries. Chad’s revenue-to-GDP ratio fell by almost half, while the Republic of Congo’s fell by 40.7 percent, Ecuador’s by 15.3 percent, and Malaysia’s by 7.9 percent. Equatorial Guinea’s revenue-to-GDP ratio remained at about 25 percent of GDP, as both revenue and GDP contracted sharply, while Cameroon’s hovered around 18 percent. Exceptions included Côte d’Ivoire and the Central African Republic (CAR), both of which are 11 recovering from conflict, and South Africa and Botswana, two of the region’s top economic performers. CAR’s revenue growth of over 70 percent was especially robust, though this was primarily due to a low base effect resulting from years of violent conflict and political instability. Botswana’s revenue-to-GDP ratio increased by 12.2 percent, while South Africa also enjoyed positive revenue growth, supported by improvements in tax administration and private income tax policy reforms (Figure 9). Figure 8: Oil, Non-Oil, and Total Revenue, Gabon, Figure 9: Changes in Revenue-to-GDP Ratios, Gabon 2010-2017 (% of GDP) and Comparators, 2013-2015 (%) 35.0 Central African Republic 70.7 29.0 30.1 30.2 30.0 25.8 26.1 Botswana 12.2 25.0 21.2 Cote d'Ivoire 7.2 20.0 15.9 17.5 17.1 17.6 South Africa 7.2 15.5 14.1 13.9 11.5 Equatorial Guinea 1.9 15.0 12.0 7.1 Cameroon -0.5 10.0 11.9 -7.9Malaysia 5.0 5.1 7.0 -15.3 Ecuador 0.0 2010 2011 2012 2013 2014 2015 2016 2017f -29.4 Gabon Republic of Congo -40.7 Total revenue Oil Revenue -41.6 Chad Non-oil revenue -50 0 50 100 Source: Gabonese authorities. Source: IMF WEO database. 82. Gabon’s total revenue is in line with the CEMAC average and above the SSA average, but slightly below the average for a broader set of comparator countries. At 21.1 percent, Gabon’s 2015 revenue-to-GDP ratio was above both the CEMAC average of 19.8 percent and the SSA average of 18.1 percent, yet below both the 23.5 percent average for 11 comparator countries and the 22 percent average for upper-middle-income countries (Figure 10). Figure 10: Fiscal Revenue, Gabon and Comparators, 2015 (% of GDP) Revenue CEMAC average SSA average All comparators 40.0 35.0 30.0 25.0 21.1 20.0 15.0 10.0 5.0 0.0 Sources: Gabonese authorities and IMF WEO database. 12 83. The oil sector’s contribution to total public revenue fell from 53.7 percent in 2010 to about 30 percent in 2016. Most oil revenue comes from non-tax sources (Figure 11), including surface rent, royalties, bonuses, government oil share, and transfers from the Gabonese Refinery Company (Société Gabonaise de Raffinage, SOGARA). Non-tax oil revenue dropped from 15.2 percent of GDP in 2012 to an estimated 5.1 percent in 2016 and is expected to recover only marginally to 6.8 percent in 2017. Meanwhile, tax revenue from the oil sector, which is mostly corporate income tax revenue, slid from 2.3 percent of GDP in 2012 to close to zero in 2016 and is projected to tick up slightly to 0.2 percent in 2017. 9 Figure 11: Government Oil Revenue Structure, Gabon, 2010-2016 (% of GDP) 20.0 18.0 16.0 15.2 14.0 12.8 13.3 12.0 11.5 10.0 9.8 8.0 6.0 6.6 4.0 5.1 2.0 0.0 2010 2011 2012 2013 2014 2015 2016e Oil tax revenue Oil non-tax revenue Source: Gabonese Authorities. 84. Tax expenditures—including exemptions, allowances, credits, reduced rates, tax holidays and deferrals—erode Gabon’s tax base. The general tax code includes 86 different provisions for tax incentives. Beyond the general tax code, there are 11 provisions for special customs categories, 26 provisions for corporate income tax expenditures, and 22 provisions for VAT expenditures. Various laws provide tax incentives for housing, tourism, agriculture, fisheries, small and medium enterprises, and new businesses. There are 28 different customs regimes, and the SEZ and Free-Trade-Zone in Nkok and Mandji Island offer long-term tax exemptions and other tax incentives designed primarily to promote export-oriented investment. 85. Policies designed to keep certain consumer prices low and stable also have negative fiscal implications. The policies, known as “measures to contain the cost of living” (« mesures de lutte contre la vie chère »), include a September 2012 decree 10 establishing the “controlled freedom” of some prices and suspending customs duties and VAT on meat, poultry, fresh and preserved fish, dairy products, fruits, fresh and canned vegetables, pasta and rice, as well as an April 2013 decree 11 blocking the prices of 166 imported food products. These were designed as temporary measures and were set to expire in 2014, but they were extended, albeit with a slightly modified list of products. These measures and other tax expenditures have a considerable fiscal 9 Gabonese authorities’ projections 10 Decree n°2727/MEEDD/SG/DGDDI. 11 Decree n° 241/MEEDD/SG/DGCC. 13 cost: in 2015, the import ban on used vehicles more than three years old alone cost the government 25 billion Central African FCA francs (CFAF), or about 0.3 percent of GDP. 86. Most Gabonese tax expenditure provisions violate CEMAC rules. CEMAC Act no. 2/92-UDEAC formally banned exemptions other than those explicitly mentioned in Article 40. Exceptional exemptions granted by ministers of finance or national customs directors are expressly forbidden, and no exemptions are permitted unless their social and economic utility has been clearly established. 87. Taxes account for almost all of Gabon’s non-oil revenue: it is skewed toward consumption taxes, particularly indirect tax revenue collected at the border. However, the share of non-tax revenue in non-oil revenue doubled between 2010 and 2016. While non-oil non- tax revenue rose from 0.6 percent to 1.2 percent of GDP, it remains a modest contributor to total revenue. Revenue collected at the border by the DGD, (mostly import tariffs, import VAT on imports, and excise taxes on imports) is Gabon’s largest source of non-oil (tax) revenue, at an estimated 28 percent of non-oil revenue and 30 percent of non-oil tax revenue in 2016. Tariffs and import VAT represented roughly 95 percent of all revenue collected at the border, as imports are largely exempt from excise taxes. 98 percent of total imports were zero-rated in 2014. 12 88. Consumption taxes on domestic products are the second-largest contributor to non- oil revenue. Domestic VAT and other taxes on domestic goods and services (including internal excise duties, and gambling, rent and fuel taxes) made up 22.5 percent of non-oil revenue in 2016. 89. Direct taxes account for about a third of non-oil revenue, compared to half for indirect taxes. The third and fourth largest contributors to non-oil revenue were corporate income taxes (CIT), which accounted for an estimated 17.5 percent of non-oil revenue in 2016, and personal income taxes (PIT), which contributed an estimated 12.5 percent (Figure 12). Figure 12: The Non-Oil Revenue Structure, Gabon, 2016 (% of total non-oil revenue) Other Nonoil nonoil corporate revenue income tax 20% 17% Personal income tax 12% Revenue VAT and collected at other taxes the border on domestic 28% goods and services 23% Source: Gabonese authorities. 12 Bank staff estimates based on ASYCUDA data from the DGDDI for 2014. 14 Revenue Collected at the Border 90. Increased tax expenditures, contributed to a fall in border revenue collected by the DGDDI between 2010 and 2016 even as the ratio of imports to GDP grew. DGDDI revenue decreased from 5.1 percent of GDP in 2010 to 3.3 percent of GDP in 2016, while the ratio of imports to GDP rose from 30 percent in 2010 to 34 percent in 2016. Border tax expenditures include tariff and VAT reductions and exemptions on imported food products as part of “measures to contain the cost of living” and food-distribution programs implemented through the CECA- GADIS supermarket chain, 13 as well as oil and mining research equipment and military equipment. 91. In 2014, border-tax expenditures cost the government approximately CFAF 211 billion. This amount is equal to 2.35 percent of GDP or 3.78 percent of non-oil GDP, about 9 percent of total revenue, and about 62 percent of total DGDDI revenue. Border-tax expenditures increased by an estimated 67 percent between 2010 and 2015, and their composition shifted, with state market (government purchases of goods and services), “measures to control the cost of living,” ad hoc, and oil-sector VAT exemptions accounting for a growing share of foregone DGDDI revenue (Figure 13 and Figure 14). Figure 13: Border-Tax Expenditures by Regime, Figure 14: Border-Tax Expenditures by Regime, Gabon, 2010 (% of total border-tax expenditures) Gabon, 2016 (% of total border-tax expenditures) CECAGADIS Petroleum Petroleum and mining Military (food Military and mining research equipment distribution)equipment research CECAGADIS equipment (food , French Other 4% , national equipment 1% 6% distribution) forces 6% forces Other 3% Exemption 4% 5% 2% budget laws, 5% decrees, reduced etc. 5% rate 8% reduced 23% Military rate State equipment 52% High cost market , national of living' 20% forces 15% 7% State market 11% zero-rated VAT - oil sector Ad-hoc 10% exemption s 13% Source: DGDDI. Source: DGDDI. 92. Weaknesses in border-tax administration cost the government an estimated 0.7 percent of GDP per year. In 2015, the effective tariff rate was 9 percent, and the import-to-GDP ratio was 28.3 percent. 14 The customs duties to GDP ratio should therefore be about 2.5 percent of GDP. However, only 1.8 percent of GDP was collected, indicating a revenue leakage of about 0.7 13 Compagnie d’Exploitations Commerciales Africaines – Société Gabonaise de Distribution. 14 World Bank’s World Development Indicators (WDI). 15 percent of GDP. This leakage likely arises from the undervaluation of imports, the abuse of exemption regimes, and the duty-suspension or relief mechanisms. 93. Over the medium-to-long term, the implementation of revised trade agreements with the EU is expected to decrease border-tax revenue, underscoring the importance of boosting domestic revenue mobilization. The gradual liberalization of selected tariff lines under EU- CEMAC trade agreements is projected to decrease tariff revenue by between 16 and 47 percent and total DGDDI revenue by between 7 and 19 percent, depending on the reform scenario. 15 VAT on Domestic Goods and Services 94. Gabon’s statutory VAT rate is high by the standards of comparable countries, but the internal VAT and other indirect taxes levied on domestic goods and services yield relatively little revenue. Gabon’s VAT rate of 18 percent is higher than the averages for CEMAC (17.3 percent), SSA (15.2 percent), and upper-middle-income countries (14.7 percent) (Figure 15). However, in 2016 Gabon collected 2.7 percent of GDP in taxes on domestic goods and services, well below the 4.8 percent collected by Cameroon in the same year. Revenue collected by the DGI from indirect domestic taxes also decreased over the 2013-2016 period, from a high of 3.2 percent of GDP in 2013 to an estimated 2.7 percent in 2016. 95. Gabon’s numerous VAT expenditures partly explain its relatively low level of domestic indirect tax revenue. The list of VAT-exempt, zero-rated, and reduced-VAT-rate commodities is long and includes newspapers, textbooks, medicines, pharmaceuticals, fertilizers, and capital goods for the tourism, agriculture, and livestock sectors. In 2015, foregone revenue from domestic VAT expenditures totaled CFAF 62.55 billion, equivalent to 0.74 percent of GDP or about 3.48 percent of total public revenue. 16 96. Though the government lowered the VAT registration threshold, from CFAF 80 million to CFAF 60 million (roughly about US$ 100,000), Gabon still has the highest threshold among comparator countries, which may incentivize businesses to remain or appear small. Equatorial Guinea, CAR, the Republic of Congo, and Ecuador have no VAT threshold, while Cameroon’s (50 million CFAF or about US$ 83,000), Côte d’Ivoire’s (US$ 40,000 for services and US$80,000 for goods), and South Africa’s (about US$ 75,000) are significantly lower (Figure 15). 17 While Gabon’s high threshold reduces taxpayer compliance costs and DGI administrative costs, it also increases the risks that medium-sized businesses abuse the system by attempting to appear smaller than they really are to remain under the threshold. 15 These are World Bank staff estimates based on DGDDI ASYCUDA data. 16 These are World Bank staff estimates based on DGI and supply-use data. 17 Crowe Horwath International, Africa VAT Guide 2016; KPMG’s Americas indirect tax country guide. (for Ecuador). 16 Figure 15: Standard Statutory VAT Rate, Gabon and Comparators VAT rate CEMAC average SSA average Upper-middle income 20 18 18 16 14 12 10 8 6 4 2 0 Malaysia Congo, Republic of Chad Cote d'Ivoire Ecuador South Africa Gabon Central African Botswana Equatorial Guinea Cameroon Republic Source: USAID Collecting taxes database, 2012-13. Income Taxes 97. Revenue from Gabon’s taxes on income, profits, and capital gains is in line with that of comparable countries, but well below that of the best performers. Gabon’s revenue from taxes on income, profits, and capital gains was 4.7 percent of GDP in 2015. The best performer in CEMAC, the Republic of Congo, collected 6 percent of GDP, while South Africa collected 15.1 percent of GDP, Botswana 9.3 percent of GDP and Malaysia 8.8 percent of GDP the same year (Figure 16). Figure 16: Revenue from Taxes on Income, Profits, and Capital Gains, Gabon and Comparators, 2014 or latest 16.0 15.1 14.0 12.0 8.8 9.3 10.0 8.0 6 6.0 4 4.1 4.7 4.0 2.1 2.0 0.0 Cote d'Ivoire Malaysia South Africa Gabon Republic of Congo Equatorial Guinea Cameroon Botswana Source: IMF GFS. 98. Gabon’s weak income tax revenue performance partially reflects its below-average maximum PIT rate and intake. There is no minimum PIT rate in Gabon, and the 35 percent maximum rate is below the 40 percent average rate for 11 comparator countries (Figure 17). In 17 2016, Gabon collected 1.5 percent of GDP in PIT revenue, significantly less than Côte d’Ivoire, which collected 2 percent of GDP. 18 99. However, Gabon’s statutory CIT rate is on par with those of comparator countries and cannot explain its below-average CIT revenue. Gabon’s statutory CIT rate of 30 percent is in line with the SSA average of 30.9 percent and below the CEMAC average of 34.7 percent, but it is significantly higher than the 23.2 percent average for upper-middle-income countries (Figure 18). Figure 17: PIT Rates, Gabon and Comparators Figure 18: CIT Rates, Gabon and Comparators CIT rate CEMAC average Minimum PIT rate Max PIT rate SSA average Upper-middle income 70 45 60 40 35 30 50 30 40 25 20 30 15 20 10 5 10 0 Malaysia Congo, Republic of Chad Cote d'Ivoire Ecuador South Africa Gabon Central African Republic Botswana Cameroon Equatorial Guinea 0 Source: USAID Collecting taxes database, 2012-13. Source: USAID Collecting taxes database, 2012-13. 100. Rising tax expenditures are compounding the negative impact of low oil prices on CIT revenue. CIT revenue dropped from 4.7 percent of GDP in 2013 to 2.1 percent in 2016. Gabon’s numerous CIT expenditures are sector- and location-specific, and almost all are profit-based. The standard CIT rate is 30 percent, but firms holding intellectual property rights, licensed tourism operators, land developers building low-income housing, public companies, nonprofit organizations, and the Gabonese Development Bank, inter alia, are subject to a reduced rate of 25 percent. Other CIT expenditures include exemptions from the minimum tax, tax holidays, accelerated depreciation, and deductions for training costs or capital reinvestment. For SEZ-based companies, the 10-year CIT holiday is followed by a reduced CIT rate of 10 percent for an additional five years. Gabonese-owned small and medium enterprises with annual turnover under CFAF 2 billion are exempt from CIT for the first five years of their operation. Oil and gas company profits are often exempt from CIT during the exploration and development phase (Table 3). Almost all CIT expenditure provisions are profit-based: they either offer lower tax rates or reduce the portion of revenue subject to taxes. 18 Gabonese and Ivorian authorities’ data. 18 Table 3: CIT Incentives under the Income Tax Act Activity or Sector Tax Incentive Farming Temporary tax holiday for cooperatives engaged in the production, transformation, storage, or sale of agricultural products Cement manufacturing 7-year tax holiday SEZ-based firms 10-year tax holiday followed by a reduced rate of 10% for 5 years Hotels Tax holiday during construction, and an additional 3-year tax holiday if the investment is greater than CFAF 1.8 billion Tourism 5-year tax holiday; 25% reduced rate for companies with special authorization Mining Tax holiday for research and development; 5-year tax holiday for exploitation New company 1-year tax holiday; 50% reduced rate for the second year. Firms eligible under the investment code also receive a 3-year tax holiday. Small and medium enterprises 5-year tax holiday under conditions defined in Law N°16/2005 of 2006 Petroleum subcontractors Special tax rate of 5.25% of turnover Companies listed on the Gabon 20%, 25%, or 28% rate for the first three years Stock Exchange Public companies 25% rate Firms that own intellectual 25% rate property rights Land developers 25% rate for authorized firms Nonprofit organizations 25% rate Source: Gabon 2015 Tax Code. 101. CIT expenditures cost the government the equivalent of about 12 percent of CIT revenue. In 2014, CIT expenditures totaled CFAF 41.27 billion, equivalent to 0.46 percent of GDP or 1.76 percent of total government revenue. Oil and mining companies received 45 percent all CIT tax expenditures, or CFAF 18.57 billion. 19 Oil companies alone received CFAF 13.96 billion, or 33.8 percent of total CIT expenditures. Commercial banks were the second-largest beneficiary of CIT expenditures, receiving CFAF 8.57 billion, or 20.76 percent of total CIT expenditures. 20 Property Taxes 102. Gabonese public revenue from property taxes is relatively low, and there is considerable scope to strengthen property tax collection, particularly taxes on land and other immovable property. The government derives most of its property tax revenue from the income tax on securities, the IRVM. Revenues from land taxes, the TSIL, and taxes on real estate are marginal (Figure 19). Gabon imposes various levies on property. While property taxes generally refer to annual taxes on immovable property, they can also include recurrent taxes on net wealth, taxes on estates, inheritances, and gifts, taxes on financial and capital transactions or transfers of securities and checks, taxes on the sale of immovable property, and other one-time property 19 These are World Bank staff estimates based on DGI data. 20 Ibid. 19 taxes. 21 Some property taxes are based on income from assets, including the special real estate tax on rents (taxe spéciale immobilière sur les loyers, TSIL) and the income tax on securities (impôt sur le revenu des valeurs mobilières, IRVM), while others are based on gains from the sale of movable and immovable assets, or on the value of assets, including taxes on real estate and registration fees. Figure 19: Property Tax Revenue Structure, Figure 20: Property Tax Revenue, Gabon and Gabon, 2011-2016 (current CFAF millions) Comparators, 2014 (% of GDP) 140,000 0.12 120,000 0.1 0.1 100,000 0.08 0.08 0.07 80,000 0.06 60,000 0.04 100,227 0.04 0.03 40,000 45,421 0.02 20,000 40,935 39,169 24,224 26,541 0 - Cote d'Ivoire Gabon Central African Republic Equatorial Guinea Botswana 2011 2012 2013 2014 2015 2016f Special rent tax (TSIL) Property and land tax Tax on movable assets/securities Real estate tax (15% IRPP/revenus professionnels) Source: DGI. Sources: Gabonese authorities and IMF WEO database. 103. Over the medium term, strengthening property tax collection could yield much- needed revenue for local authorities. At 0.07 percent of GDP in 2014, Gabon’s property tax revenue is on par with comparator countries such as CAR (0.1 percent), Côte d’Ivoire (0.08 percent), Botswana (0.04 percent), and Equatorial Guinea (0.03 percent), but improved collection could significantly increase revenue (Figure 20). Collection costs are high, and property tax revenues are likely to remain low in the short term, but reforms could significantly boost the income of local authorities over time. Additional revenue from property taxes could augment local revenue from user fees for services and from leasing, operational transfers for vehicle maintenance, removal-of-household-waste tax rebates, and fines and pecuniary judgements. The main obstacle to improved property tax collection is the absence of a comprehensive tax cadaster, and establishing one would require close collaboration between the DGI, the National Agency for Urban Planning, Topographic Works, and Land Registration (Agence Nationale de l’Urbanisme, des Travaux Topographiques et du Cadastre, ANUTTC), and local authorities. 21 For further details, see the IMF Government Financial Statistics or Organisation for Co-operation in Europe (OECD) statistics manuals. 20 Tax Administration 22 104. Compounding deficiencies in the policy framework, weaknesses in tax and customs administration significantly reduce non-oil revenue. While many aspects of tax and customs administration could be improved, reforms in several key areas could yield especially substantial gains over the near term. These priority areas include the quality of taxpayer information, obstacles to voluntary compliance, compliance-risk management, reporting accuracy in tax and customs declarations, and the administration and oversight of tax expenditures. 105. Inaccurate and incomplete taxpayer information reduces collection efficiency and does not support effective interaction between the DGI and taxpayers. The DGI’s taxpayer registry includes numerous duplicate taxpayers, and it does not accurately record the business and/or residential addresses of registered taxpayers. The proportion of inactive taxpayers is thought to be large, though it has not been formally assessed. 106. The DGI makes only limited use of third-party information to detect unregistered businesses and individuals, or to differentiate between delinquent and inactive taxpayers. The DGI’s IT system cannot distinguish between firms that are not up to date on their tax payments and those that have permanently closed. The DGI also lacks sufficient staff and financial resources to physically identify and contact the relevant businesses. 107. The complexity of the tax system and the need for multiple and redundant contacts with revenue authorities inhibit voluntary compliance. Gabon ranked 161st out of 190 countries on the 2017 Doing Business report’s “paying taxes” indicator. The launch of the e-tax platform in 2015 reduced but did not eliminate the compliance burden on taxpayers, particularly firms. About 400 small and medium enterprises filed declarations and paid corporate taxes online, on the e-tax platform in 2017. Not all taxes can be filed and paid on the e-tax platform, but its comprehensiveness is increasing with each upgrade, the most recent of which was in February 2018. However, the large amount of time and numerous contacts with the tax administration necessary to fully comply with tax obligations still negatively impact voluntary compliance rates. An average of 488 hours is required to prepare, file, and pay corporate, labor, and consumption taxes in Gabon. While the time cost of tax compliance in Gabon is significantly below the CEMAC average of 577 hours, it is well above the SSA average of 307 hours and the averages for individual comparator countries such as Côte d’Ivoire (270 hours), South Africa (203 hours), Malaysia (164 hours), and Botswana (152). The frequency with which Gabonese companies have to file and pay taxes and fiscal contributions (26 payments per year) is the lowest in CEMAC and below the regional average (36.7), but significantly above the frequencies observed among Gabon’s aspirational peers, including South Africa (7), Malaysia (9), and Ecuador (8). 23 108. Compliance-risk management is ineffective, as tax-fraud and avoidance risks do not affect the likelihood of audits. The authorities should design methodological approaches and administrative procedures to assess, rank, and quantify taxpayers according to the associated compliance risks. The DGDDI’s Customs Information System with Advanced Relational 22 This section draws on findings from interviews with DGI and DGDDI staff led by Rick Tsouck Ibounde. 23 Doing Business, 2017 21 Architecture (Système Informatique Douanier à Architecture Relationnelle Avancée, SINDARA 24) flags 75 percent of tax declarations for audit without adequately accounting for risk and risk profiles, and all fines and penalties collected by the DGDDI’s Directorate of Investigations and Litigation represent just 0.17 percent of total customs revenue. Similarly, DGI desk audits are carried out randomly, and the DGI audits only 10 percent of large taxpayers and roughly 30 percent of all active taxpayers each year. The DGI does not have access to third-party data or external intelligence databases, and its information system does not include any indicators that could help target audits more efficiently. 109. The DGI lacks a dedicated department tasked with coordinating tax audits, and its intelligence-gathering and research efforts do not focus on identifying compliance risks. The various units involved in the DGI’s auditing process are scattered across four national directorates and nine provincial directorates, and they do not communicate with one another. In principle, the Investigations and Recoveries Department (Service des Enquêtes et Recoupements) should directly inform the targeting of audits, but in practice it only marginally supports the audit process, and in 2014 it brought just six cases for audit. 110. Much more could be done to ensure accurate reporting in tax declarations. For example, there was until very recently no concerted, systematic effort to identify cases in which tax write-off amounts exceeded assessed tax amounts in a given tax year, or cases in which businesses imported goods for a value exceeding their turnover figure. The DGI’s Integrated Software for Enforcement and Collections (Logiciel Intégré pour l’Imposition et le Recouvrement) could be upgraded to highlight inconsistencies, perform large-scale automated cross-checks, flag unusually low net incomes, chronic net losses, or recurrent noncompliance, and compare reported turnover with third-party information. 111. Greater transparency could improve the administration and oversight of tax expenditures. Government officials have considerable discretionary power over tax expenditures in general, and tax exemptions in particular. There is no comprehensive list of legitimate tax exemptions, and several authorities have the capacity to grant exemptions through special agreements. Exceptional tax exemptions are, by Gabonese law, only the prerogative of the Minister of Economy, even though this violates CEMAC rules, but in practice Gabonese legal and administrative safeguards are often bypassed. 112. Increased connectivity between IT systems could further strengthen the administration of tax expenditures. The DGDDI has established an interface between its two computer systems, SINDARA and the Customs System for Managing Economic and Privileged Regimes (Système Douanier de Gestion des Régimes Economiques et Privilégiés, SYDOGEP). However, customs exemptions may still be granted twice on only one legal basis. Because customs-exemption requests are not systematically archived, an application rejected by one customs inspector could be reintroduced and examined by another. There is no process for following up on temporary exemptions. SYDOGEP stores a list of all applications for temporary exemptions that have reached their end date, as well as the exemption decision for each, but the corresponding declarations are in SINDARA. As a result, the relevant files are not automatically closed, and unless taxpayers themselves acknowledge the end of their period of exemptions, the 24 Gabon’s SINDARA system is based on the Automated System for Customs Data (ASYCUDA), version 2.7. 22 administrative measures necessary to revoke the temporary exemptions are not systematically taken. 113. The government is considering future plans to bring the tax and customs administrations into a semi-autonomous Gabonese Revenue Authority (Office Gabonais des Recettes, OGR). A steering committee is preparing the overarching strategy. The merger would not eliminate the need to address weaknesses in the tax administration. Rather, it would provide an opportunity to implement the necessary actions in parallel with complementary measures to reform the public administration payroll and control the cost of tax administration. Successfully merging the tax and customs administrations could help reduce administrative costs, simplify and improve service delivery, eliminate overlapping roles and responsibilities, and accelerate the integration of IT systems. DGDDI and DGI staff members also do not have to automatically be detached to the new agency and transfers to the OGR can be conditioned on good performance, and salaries and pensions renegotiated. 114. However, the creation of the new revenue agency could temporarily divert scarce financial and administrative resources away from the implementation of planned and ongoing tax and customs reforms. Critical, urgent and costly tax IT infrastructure investments (to notably better monitor corporate income tax expenditures) and customs’ move from SINADARA/ASYCUDA ++ to ASYCUDA World could compete for funding with activities necessary to establish the OGR. Protracted strikes by tax and customs staff in 2017 exacerbated tensions, and personnel reorganization or downsizing undertaken as part of the creation of the OGR could further erode staff morale. Recent Tax Policy and Administration Reforms: Efficacy and Impact 115. In 2014, internal VAT, border-tax, and CIT expenditures cost Gabon an estimated CFAF 315 billion, or 3.55 percent of GDP, in foregone revenue. Border-tax expenditures alone cost the government an estimated 2.35 percent of GDP each year. Internal VAT expenditures cost the government an estimated 0.74 percent of GDP, while CIT expenditures cost about 0.46 percent of GDP. 25 116. Despite their exorbitant cost, tax expenditures do little to increase foreign direct investment, as they cannot compensate for a relatively poor investment climate. The evolution of foreign direct investment in Gabon has mirrored the CEMAC trend, particularly in recent years, which suggests that regional rather than country-specific factors are at work (Figure 21). Meanwhile, the government has made little to no progress in improving Gabon’s business climate (Figure 22). 25 These are World Bank staff estimates based on DGI and DGDDI data. Tax expenditures are estimated using the foregone-revenue method, which estimates the loss of revenue incurred by the government due to tax expenditures but does not account for taxpayers’ response to tax expenditures. 23 Figure 21: Net Foreign Direct Investment Inflows, Figure 22: Gabon’s Global Rank on Selected Doing Gabon and Comparators, (% of GDP) Business Indicators, 2016-2017 Gabon CEMAC SSA average Upper-middle income Overall 164 8.0 161 Resolving insolvency 123 121 7.0 Enforcing contracts 177 172 6.0 166 Trading across borders 165 5.0 Paying taxes 161 160 Protecting minority 158 2017 4.0 investors 157 118 2016 Getting credit 109 3.0 Registering property 175 175 2.0 Getting electricity 158 154 1.0 Dealing with 167 construction permits 164 Starting a business 152 0.0 144 2010 2011 2012 2013 2014 2015 0 50 100 150 200 Source: WDI. Source: DB 2016-17. 117. Shifting from profit-based to cost-based tax incentives would enhance the efficiency of tax expenditures. International experience shows that replacing tax exemptions with an accelerate-depreciation mechanism, explicit targeted investment subsidies embedded in the tax code, and/or provisions extending the carryover period for losses from three years to five could make tax expenditures more effective in attracting foreign investment. Better yet, shifting altogether from profit-based to cost-based investment can lessen the cost of incentives while fostering investment. Tax holidays should be avoided, as they encourage short-term low upfront investment projects (Box 1). However, measures to improve the business climate—especially by streamlining procedures for starting a business, registering property, and trading across-borders and by providing better protection to minority investors—would have a more positive impact on investment than even a comprehensively reformed tax incentive regime. Box 1: Options for Reforming Investment-Related Tax Incentives Tax incentives are not the most effective or efficient way to attract investment, especially in developing countries. According to investment-climate surveys, investors typically rank tax incentives low on their list of concerns when deciding whether to invest in a developing country, and good infrastructure, macroeconomic stability, and the rule of law are all more important factors. A UNIDO survey of 7,000 companies operating in the agriculture, mining, manufacturing, utilities, construction, and services sectors of 19 SSA countries found that tax incentives ranked 11th out of 12 concerns influencing investment decision. The quality of the legal framework was the 5th most important issues, and political stability was the most important. While tax incentives tend to be less effective in attracting investment than structural reforms and improvements in the business climate, some types of tax incentives are more effective than others. Incentives that lower the cost of investment, such as accelerated depreciation schemes and special tax deductions and credits, tend to make a greater number of investment projects more profitable at the margin, facilitating investments that would not otherwise have been made. By contrast, profit-based tax incentives such as tax 24 holidays, preferential tax rates, or CIT exemptions generally reduce the tax rate applicable to taxable income, and thus they tend to benefit projects that would already be financially viable. While many developing countries have attempted to attract investment by adopting costly tax holidays and CIT exemptions, investment tax credits and accelerated depreciation tend to yield more investment per dollar of foregone revenue. Tax holidays tend to favor short-term projects with low sunk costs that can be relocated rather than stable, long-term investment, while tax incentives that target extractive industries or sectors producing primarily for domestic markets often have little impact. Source: IMF, World Bank, OECD and UN (G-20), 2015, Options for Low Income Countries' Effective and Efficient Use of Tax Incentives for Investment. 118. The effectiveness of Gabon’s VAT and customs exemptions on certain food products is also questionable. Gabonese authorities credit the adoption of “measures to contain the cost of living” with the low rates of food- and nonfood-price inflation observed since these measures were introduced, in 2012, Gabon’s inflation rate has been below the CEMAC average and one of the lowest among comparator countries (Figure 23). However, the food-price inflation rate rose from -0.4 percent 2013 to 2.8 percent in 2014 and 3 percent in 2015 (Figure 24). Food-price inflation is estimated to have increased sharply in 2016, even though the measures remain in effect. Moreover, the measures do not necessarily benefit the poorest households, which purchase much of their food from small informal shops that are not VAT- registered. Figure 23: End-of-Period Inflation Growth, Gabon Figure 24: Harmonized Household Consumption and Comparators, 2013-2016 (percentage points) Price Index, Gabon, 2011-2015 Chad 480.0 Food price index Cote d'Ivoire 189.2 Housing and Energy price index Republic of Congo 118.6 CPI CEMAC 84.2 7.0 5.7 Cameroon 5.0 4.0 33.2 4.7 3.0 2.8 South Africa 24.1 3.0 2.7 Ecuador 1.0 1.2 8.2 0.5 -0.4 Botswana -20.0 -1.0 2011 2012 2013 2014 2015 -3.0 -0.4 -23.4 Gabon Central African Republic -32.0 -5.0 -34.8 Malaysia -7.0 Equatorial Guinea -71.2 -9.0 -100 0 100 200 300 400 500 600 -11.0 Source: IMF WEO. Source: Gabonese authorities. 119. To address its widening fiscal deficit, the government has launched a wide range of reforms as part of the PRE 2017-19 designed to reduce the cost of tax expenditures and boost non-oil revenue through improved tax policy and administration. As noted above, these measures include easing the ban on used vehicle imports, introducing new manganese, gold, and gold dust export taxes, reducing the number of VAT-exempt items, rationalizing CIT, customs, and VAT expenditures, strengthening the fiscal cadaster, upgrading the DGDDI’s IT system, and boosting tax compliance through the adoption of a risk-based tax auditing strategy. The authorities expect that cuts to tax expenditures will yield the bulk of both short- and medium-term revenue gains (Table 4). 25 Table 4: Estimated Annual Gains from Improvements in Tax Policy and Administration, 2017-2021 (% of GDP) Total revenue collection gains from all areas, of which: 1.2-1.42 Easing of the ban on used vehicle imports 0.12 Customs service information and IT system upgrades 0.15-0.3 Reduced tax expenditures linked to the ‘high cost of living’ initiative; cuts in 0.2 discretionary franchises’ and state market associated tax expenditures VAT expenditure rationalization 0.15 CIT expenditure rationalization 0.1 Introduction of new manganese, gold and gold dust export taxes 0.08-0.15 Fiscal cadaster improvement 0.1 VAT registration campaigns, risk-based auditing and other measures to boost tax 0.2 compliance and fight tax evasion Withholding tax on bank deposits 0.1 Sources: Gabonese authorities. 120. The government is also attempting to increase property tax revenue by taxing indexed asset values rather than income from assets. Expanding the cadaster will help broaden the tax base and raise additional revenue for local authorities. In total, the government expects property tax revenue to rise by around 0.1 percent of GDP once the reforms have delivered their full impact. The government is also planning to index the TSIL on water and electricity consumption, though this will have a limited impact on total revenue. 121. Additional PRE revenue measures were added to the 2018 budget law. The budget law eliminates all tax exemptions without a legal basis and mandates the full application of CEMAC tariffs. Excise tax measures were also included in the 2018 budget law. The government reduced the number of commodities fully exempt from taxation under “measures to contain the cost of living” from 191 to 171. The authorities also created a new consolidated tax on firms with a turnover of less than FCFA 30 million, designed in part to compensate for the low VAT threshold. The government is planning to create a new active taxpayer file, which will be updated monthly and published on the DGI’s website and in local newspapers. Companies that have import values that exceed the VAT threshold, but which are not on the list of taxpayers subject to VAT, will no longer be able to clear their imports through customs until they take the necessary steps to register for VAT. Recommendations 122. Merging the tax and customs directorates into a semi-autonomous revenue authority would provide policymakers with an opportunity to address weaknesses in tax administration while also reforming the public payroll and adopting measures to reduce the cost of tax collection. However, a merger would also entails important risks. Devoting scarce resources to administrative reorganization could delay important tax and customs reforms, including vital upgrades to IT platforms. Moreover, personnel restructuring could further damage the morale of tax and customs staff. 123. PRE 2017-19 revenue measures were steps in the right direction and maintaining the depth, breath and pace of reform, particularly on tax expenditure rationalization, taxpayer information and tax and customs IT systems upgrades, will be crucial for improved fiscal 26 sustainability in the medium and long term. The GOG should follow through and effectively eliminate all tax exemptions without a legal basis, resist the temptation to re-extend the number of commodities fully exempt from taxation under “measures to contain the cost of living”, and prevent companies that have import values that exceed the VAT threshold but are not on the list of taxpayers subject to VAT, from clearing their imports through customs until they take the necessary steps to register for VAT. Critical, urgent and costly tax IT infrastructure investments (to notably better monitor corporate income tax expenditures) should be made, and the customs administration IT systems transition from SINADARA/ASYCUDA ++ to ASYCUDA World fully funded and prioritized. 124. The tax directorate should develop compliance-risk indicators, classify taxpayers according to compliance risk, and introduce risk-based auditing. Compliance-risk management is currently ineffective, as tax-fraud and avoidance risks do not affect the likelihood of audits, and fines and penalties are too low and unlikely to effectively deter fraud and non- compliance. The IT system upgrades should help support efforts to highlight inconsistencies in declarations, perform large-scale automated cross-checks, and for instance, flag unusually low net incomes, chronic net losses, or recurrent noncompliance, and compare reported turnovers with third-party information. 125. The tax directorate, ANUTTC, and local authorities around the country should move forward with plans to strengthen the fiscal cadaster and create comprehensive property taxpayer files. Collection costs are high, and property tax revenues are likely to remain low in the short term, but better property taxpayer information could significantly boost the income of local authorities over time. Additional revenue from property taxes could augment the currently meager local revenues from user fees for services and from leasing, operational transfers for vehicle maintenance, removal-of-household-waste tax rebates, and fines and pecuniary judgements. 126. Finally, the GOG should shift from profit-based to cost-based tax incentives and eliminate tax holidays to encourage longer-term investment. GOG tax incentives are almost all profit-based. Yet, profit-based tax incentives such as tax holidays, preferential tax rates, or CIT exemptions generally reduce the tax rate applicable to taxable income, and thus they tend to benefit projects that would already be financially viable. Tax holidays favor short-term projects with low sunk costs rather than stable, long-term investment, while tax incentives that target extractive industries or sectors producing primarily for domestic markets often have little impact. Shifting away from profit-based to cost-based investment could lessen the cost of incentives while fostering investment. 27 3. OVERALL EXPENDITURE TRENDS AND QUALITY ANALYSIS Introduction 127. This chapter gives an overview of the Government of Gabon’s (GOG) expenditure trends over 2010-17. It identifies key drivers of changes in expenditure levels, expenditure economic and sectoral composition, and execution performance. The chapter then focuses on payroll and public investment, the two largest categories of public expenditure, and assesses their management, effectiveness, efficiency, and associated outcomes. The chapter addresses public investment management processes and the adequacy of capital budget allocations and utilization across sectors given Gabon’s investment needs and national priorities. Finally, the chapter describes structural public employment challenges, reviews recent public-sector employment reforms, and makes recommendations to improve remuneration and recruitment policies and rein in the wage bill. Overall Expenditure Trends 128. Gabon’s main strategic and long-term goals are outlined in the PSGE. The three-pronged strategy seeks to provide the country with the requisite socio-economic infrastructure for development and economic diversification, as well as to leverage industry, sustainable natural resource management, and services to transform Gabon into an emerging economy by 2025. The PSGE calls for substantial public investments (e.g., US$13 billion 2013-16) in national and local road infrastructure, ports, energy efficiency, power generation and distribution capacity, communications, agriculture, housing, and water and sanitation to address infrastructure bottlenecks and foster economic diversification. Concrete government objectives include developing a nationwide fiber optic network, doubling the country’s energy capacity to 1,000 MW, and expanding national transport networks up to 3,600 kilometers of paved roads and 3,000 kilometers of waterways by 2025. 129. Public expenditures rose significantly in 2010-13 following the 2008-09 global financial crisis. Gabonese government spending rose from 24.9 percent of GDP in 2010 to 28.4 percent in 2013, as buoyant oil prices drove up expenditures on fuel subsidies and the GOG recruited more civil servants and increased public sector wages. The GOG also scaled up public investments to meet the goals established in the PSGE and build the infrastructure to host the 2012 soccer Africa Cup of Nations. 130. The GOG then cut public spending in 2014-15 as oil prices declined and mounting fiscal imbalances emerged. However, these expenditure cuts were unevenly distributed across spending categories. The GOG reduced capital expenditures by half, from 10.6 percent of GDP in 2013 to 5.0 percent of GDP in 2015. It also reduced spending on subsidies and transfers, including on fuel subsidies, from 5.5 percent of GDP in 2013 to 3.8 percent of GDP in 2015. Higher external debt (e.g., US$1.5 billion and US$500 million Eurobonds were issued in December 2013 and June 2015) increased interest spending from 1.7 percent of GDP in 2013 to 2 percent of GDP in 2015 (Figure 25). However, the wage bill rose from 6.4 percent of GDP in 2013 to 8.4 percent of GDP in 2015, as many civil servants’ employment status was changed from temporary to permanent, a 28 new performance-based wage premium (prime d ’incitation à la performance, PIP) was introduced, and roughly 30,000 additional civil servants were recruited between 2012 and 2014. 131. Gabonese authorities suppressed most fuel subsidies 26 in 2016, but the growing wage bill and a sudden increase in local government and CNAMGS spending drove total government expenditures up. The GOG cut most fuel subsidies in early 2016, except for kerosene (“Pétrole lampant”) and natural gas which are consumed by a large share of the population. Subsidies spending declined from 0.8 percent of GDP in 2015 to 0.3 percent of GDP in 2016. However, the wage bill grew from 8.4 percent to 8.8 percent of GDP and local governments and National Health Insurance and Social Security Fund (Caisse Nationale d´Assurance Maladie et de Garantie Sociale, CNAMGS) spending increased by more than 1 percent of GDP in 2015-16. As a result, total government spending rose from 22.4 percent of GDP in 2015 to 23.1 percent in 2016. 132. The government began to implement the medium-term 2017-19 PRE in 2017. The PRE calls for: i) the systematic retirement of about 1,200 civil servants who have reached retirement age; ii) updates of payroll records; iii) a freeze on the recruitment of civil servants (except in education and healthcare); iv) an audit of housing grants and transport allowances; and v) renewed efforts to identify ‘ghost workers’ still on the government payroll. 133. Interest spending continued to rise in 2017, but cost containment measures under the PRE helped lower total public spending. The wage bill decreased from 8.8 percent of GDP in 2016 to an estimated 8.3 percent of GDP in 2017. Subsidies and transfers fell further in 2017, as the government eliminated flour grants, allocated education grants on a merit and means basis, and reduced its support to refinery activities and fuel subsidies. Spending on goods and services also declined from 2.9 percent of GDP in 2016 to 2.3 percent of GDP in 2017 thanks in part to tighter control over government water and energy consumption and expenditures. Consequently, total government expenditures decreased from 23.1 percent of GDP in 2016 to an estimated 21.1 percent of GDP in 2017. 27 26 Most fuel subsidies were abolished in February 2016 and an international fuel price passthrough mechanism was implemented. Products that are the most widely consumed, kerosene and natural gas, are however still subsidized. Despite this measure, authorities have frozen the passthrough mechanism on gas (and hence continued to subsidize fuel) for several months in 2017, amid social tensions and rising oil prices. It is critical that the new mechanism remains in place for fuel subsidies spending to remain under control. 27 Gabonese authorities’ figures; latest IMF 2016 and 2017 estimates differ slightly, at 22.1 percent and 19.8 percent of GDP respectively. 29 Figure 25: Government Expenditure by Figure 26: General Government Expenditure, 2015 (% Economic Structure, 2010-2017 (% of GDP) of GDP) 30.0 60.0 48.3 25.0 50.0 44.5 11.4 10.6 38.7 20.0 11.6 40.0 36.2 10.3 6.8 5.0 4.9 33.2 4.9 30.0 25.1 15.0 23.1 3.8 2.9 20.6 22.4 5.7 5.5 3.9 17.0 3.3 5.1 2.2 20.0 14.9 10.0 10.0 5.0 8.8 7.9 8.4 8.3 5.8 5.2 5.9 6.4 0.0 0.0 2010 2011 2012 2013 2014 2015 2016e 2017e Payroll Goods and services Subsidies and transfers Interests Capital Other Source: Gabonese Authorities and IMF staff estimates. Source: IMF World Economic Outlook (WEO), April 2017. 134. Public expenditures in Gabon were above average in 2013 but are now comparatively low; there might therefore be little room for further expenditure cuts and reducing the growing fiscal deficit will also require revenue measures. Gabon spent 22.4 percent of GDP in 2015, 7.8 percent less than the average for the selected group of 10 peer countries (Figure 26). Chad, CAR, and Cameroon spent less, 17.0, 14.9, and 20.6 percent of GDP, respectively, while the governments of Equatorial Guinea and the Republic of Congo spent as much as 44.5 percent and 48.3 percent of GDP, respectively. However, Gabon’s overall fiscal balance went from a surplus of 1.8 percent of GDP in 2013 to an estimated deficit of 1.1 percent in 2015. The deficit is estimated to have reached 6.6 percent of GDP in 2016, as oil production declined, and oil prices remained low. Fiscal consolidation efforts therefore need to address not only spending but revenue as well (especially non-oil revenue). 135. Yet, the GOG spends more on the wage bill, subsidies, and transfers than its peers. GOG spent 8.4 percent of GDP on payroll in 2015, compared to Cote d’Ivoire’s 7 percent, Malaysia’s 6.1 percent, Cameroon’s 5.3 percent, CAR’s 4.6 percent, and the Republic of Congo’s 3.4 percent of GDP. Among the group of peer countries, only the Government of Botswana spent significantly more than the GOG at 11.7 percent of GDP in 2014. In 2015, Gabon also dedicated 3.8 percent of GDP to subsidies and transfers, more than Cameroon’s 3 percent and CAR’s 2.1 percent of GDP. GOG spending in other economic categories is generally in line with its peers’ (Figure 27). 30 Figure 27: Comparative Expenditure Structure, Figure 28: GOG Budget Execution, 2014-2017 2015 or latest (% of GDP) 140 135 0.3 123 Other 120 112 105 107 107 104 5.0 100 100 Capital 98 99 9896 100 95 89 87 84 2.0 77 79 79 Interests 76 76 80 72 72 70 65 Subsidies and 3.8 60 transfers 47 2.8 40 Goods and services 8.4 20 Payroll 0 0.0 10.0 20.0 30.0 Payroll Goods Subsidies Interests Capital Other TOTAL Malaysia Gabon and and Côte d'Ivoire Republic of Congo services transfers Central African Republic Cameroon Botswana 2014 2015 2016e 2017e Source: IMF GFS, Gabonese authorities, Cameroon Source: Gabonese authorities. BOOST, and Bank staff calculations. Note: Budgetary central government data. 136. Budget execution in Gabon has improved in recent years. Execution on an order to pay basis was low in 2014, at 70 percent, but rose to 89 percent in 2016, before reaching an estimated 104 percent in 2017. Capital expenditure was under-executed by a wide-margin in 2014-16 (common for countries in crisis), as the government cut its investment budget to respond to the fall in crude oil prices and its impact on the fiscal balance. The execution of capital expenditures has nonetheless improved, mostly because the capital budget was halved, falling from FCFA 1,321 billion in 2014 to FCFA 644 billion in 2015. 2017 estimated budget execution rates are generally very high, except for subsidies and transfers, mostly because the 2017 budget was extremely conservative and factored in PRE-related cuts (Figure 28). 137. Almost half of Gabon’s public spending is public administration spending; 20 percent of government expenditure go to health, education, and social protection. Approximately half of Gabon’s public expenditures (47 percent in 2015) is general public-service sector spending.28 GOG's second largest category of spending is education (12.9 percent of total expenditures in 2015). The other large categories are housing and communal services (11.6 percent of total expenditures in 2015), followed by recreation, culture and sports, and defense, each at 6.5 percent of GOG total spending in 2015. At the same time, the GOG only spends 6.2 percent of total 28 General public services expenditures include public administration spending (i.e., legislative bodies and judiciary expenditures) and external affairs spending along with any other spending not classified elsewhere. 31 expenditure on economic affairs, 29 4.7 percent on health, and 2.7 percent on social protection (Figure 29). Figure 29: GOG Executed Expenditure by Figure 30: Expenditure Functional Breakdown, 2015 or Function, 2015 (% of total expenditures) latest (% of total expenditures) Recreation, culture and South Africa religion Social 7% protection Malaysia Education 3% 13% General Côte d'Ivoire Public Health services 5% Republic of Congo 47% Gabon Housing 0% 20% 40% 60% 80% and communi General Public services Defense ty Defense Economic amenitie 6% Public order and safety Economic Affairs Affairs s 6% Public Environmental protection Housing and community ame Environme 12% order and Health Recreation, culture and religio ntal safety protection Education Social protection 1% 0% Source: Gabonese authorities and Bank staff calculations. Source: IMF GFS, Gabonese authorities, and Bank staff calculations. Note: Budgetary central government data. 138. Gabon spends more on public administration and less on economic affairs, housing, and health than peers. While the share of Gabon’s public spending on general public services (47.2 percent) was lower than South Africa’s (54.1 percent) in 2015, it was higher than the Republic of Congo’s (26.5 percent) and Cote d’Ivoire’s (31.8 percent). Spending on recreation, culture, and religion was also high in Gabon, reflecting large sums allocated to sports and cultural events such as the 2016 Africa Cup of Nations, the New York Forum Africa, and the Tropical Amissa Bongo cycling race. Gabon did however spend a comparatively low share of its public expenditures on economic affairs in 2015 (6.2 percent) compared to the Republic of Congo (22.3 percent), South Africa (8.53 percent), and Malaysia (23.8 percent). Similarly, the GOG spent less than its peers on housing and communal amenities, health, education, and social protection (Figure 30). 139. Further expenditure savings would likely come from payroll. Gabon dramatically cut subsidies and transfers spending in recent years, down to only an estimated 2.2 percent of GDP in 2017 so that subsidies and transfers spending in Gabon is now be closer to the peer average. 2017 payroll spending (8.3 percent of GDP) is however still comparatively high; savings could therefore likely come from cuts in payroll. Payroll rationalization could also help rebalance Gabon’s distribution of spending across sectors. The general public-sector wage bill (i.e., payrolls of the 29 Economic affairs include the water and sanitation, housing, fuel and energy, agriculture, industry and mining, transport and telecommunication sectors. 32 Ministries of Budget, Economy, Public Service, Relations with Parliament, and Justice) made up nearly half the public-sector wage bill and almost a third of total government expenditures in 2015. 30 Public Investment Overall Investment Spending Trends 140. The PSGE’s strategic investment priorities are operationalized in Gabon’s National Infrastructure Master Plan. The plan, which was put together in 2012 with support from the engineering firm Bechtel, operationalizes and estimates the cost of key projects planned under the PSGE. It also serves as a roadmap for mid- to long-term infrastructure strategies across major sectors. For instance, the plan emphasizes the need to expand and improve transport and communications networks to better connect the country’s major economic centers with the population and export zones, especially deep-water ports. The National Infrastructure Master Plan’s US$13 billion 2013-2016 budget included a focus on mining (US$3.3 billion), transport (US$3.5 billion), and power and energy (US$3 billion) (Box 2). Box 2: Gabon’s National Infrastructure Master Plan The investment priorities of the National Infrastructure Master Plan included: • Basic housing and infrastructure (including urban road networks and water and energy networks) in Libreville, Port Gentil, Lambarene, and Franceville; • Completing the Transgabonese (road and rail) transport corridor that links Libreville and Franceville; • Linking Port Gentil and the SEZ to Lambaréné and to Ndjolé by first waterways and then by road and/or rail; • Linking Belinga (and its iron ore deposits) to the Transgabonese corridor; and • Relieving congestion at Libreville airport and developing plans for a new airport. Key projects included: • Franceville ‘s Poubara 1 and 2 hydroelectric plants; • The Grand Poubara dam; • The Empress and Fe 2 dams (to supplement the Ntoum plant and increase Libreville and the Nkok SEZ's supply); • Linkages of power transmission lines between dams and cities; • Construction of a third water-supply pipeline between Ntoum and the capital; and • Improvement to the Port of Owendo. Source: Gabon National Infrastructure Master Plan. 141. Gabon’s capital expenditure (which was scaled-up to meet PSGE and Master Plan’s objectives) was close to the high end among peer countries prior to the 2014-15 oil price downturn; it is now well below average. The GOG spent 11.4 percent of GDP on investment in 2012, more than all of its peers except for Chad, the Republic of Congo, and Equatorial Guinea (Figure 32). However, Gabon’s capital expenditure dropped down to 5 percent of GDP in 2015, ranking among the lowest in the peer group, only ahead of South Africa and Malaysia (Figure 32). 30 Bank staff estimates based on Treasury data. See the wage bill section of this chapter for more details. 33 Figure 31: Capital expenditure, 2012 or latest (% of Figure 32: Capital expenditure, 2015 or latest (% of GDP) GDP) 25 21.9 60.0 52.9 20 50.0 14.7 40.0 15 12.6 30.0 10 21.5 6.9 5.9 6.0 6.3 20.0 5.0 5.8 11.4 13.7 5 4.1 8.1 10.0 4.1 4.9 5.0 6.1 0.4 0.3 0.0 0 Sources: IMF GFS 2013 yearbook, Gabonese authorities, Sources: Gabonese authorities, IMF GFS and WDI. Cameroon BOOST, IMF Chad 2013 article iv. 142. The GOG’s investment budget’s functional trends over 2015-2017 did not reflect stated government priorities, as most of the spending was allocated to transport and core government functions to the detriment of other priority sectors, such as housing, energy and ICT. An average of 57 percent of budgeted total investment spending was allocated to transport and governance (which includes defense and military, judiciary, and decentralization infrastructure spending) between 2015 and 2017 (Figure 33). Only an average of 11 percent of budgeted investment went to health, education, and employment, and approximately an average of 8 percent went to water and sanitation during the same period. An average of barely 0.5 percent of the public investment budget was allocated to the housing sector, while an average of 10.2 percent of investment spending were allocated to the construction of stadiums. 34 Figure 33: Functional Distribution of Planned Investments Figure 34: Quality of trade and transport- 2015-17 related infrastructure (1=low to 5=high), 2016 100% 4.0 3.8 Other infrastructure 3.4 90% 3.5 Housing 3.0 3.0 2.6 80% 2.5 2.5 2.5 Utilities 2.5 2.2 70% 33.3 2.0 2.1 32.4 38.0 Energy 2.0 60% 1.5 1.5 Transport 50% 1.0 0.5 1.9 IT/Communications 40% 13.0 0.5 28.3 12.4 Stadia 30% 0.0 Education/Employmen 20% 7.6 19.2 32.3 t/Health Governance 10% 15.1 Agriculture/Rural 0% Development 2015 2016 2017 Sources: Gabonese authorities/PLF 2015-2017. Source: WDI, April 2017. Note: 2014 CAR data. Public Investment and Infrastructure Quality 143. Despite historically high levels of transport sector investment spending, the quality of Gabon’s transport infrastructure compares poorly to that of its peers, with the exception of seaports. The quality of the country’s trade and transport infrastructure is lower than the average for peer countries (Figure 34), even though the GOG allocated a large share of its investment budget to transport in 2015-16. Moreover, Gabon’s score of 2.19 on the World Bank’s Logistics Performance Index 31 is far lower than that of many of its peers such as Botswana (3.05), Malaysia (3.43), and South Africa (3.78). However, the country’s seaports are of higher quality than those of comparator countries (Figure 35), although their quality has degraded in recent years (from a score of 5.9 on the Global Competitiveness Index (GCI) in 2010 to 5.3 in 2016). 31 The Logistics Performance Index is an interactive benchmarking tool created to help countries identify the challenges and opportunities they face in their performance on trade logistics. It consists of a scale from 1 (bad performance) to 5 (good performance). Source: The World Bank’s Logistics Performance Index (https://lpi.worldbank.org). 35 Figure 35: Quality of Port Infrastructure (1=low to Figure 36: Access to electricity, rural (% of rural 7=high), 2016 population), 2014 6.0 120.0 100.0 5.0 80.0 4.0 60.0 42.3 3.0 40.0 20.0 2.0 0.0 1.0 0.0 Source: World Economic Forum,GCI. Source: WDI, April 2017. 144. Road quality is poor, as inefficient spending has hindered the development of a good quality network. Gabon expanded its road network from 9,170 km of roads in 2013 to 10,130 km of roads in 2015. While around 2,000 km of roads were paved (20 percent of total roads) at the end of 2014, the GOG’s objective 32 was to have a total of 3,600 km of paved roads by 2016 and build an additional 2,500 km of paved roads between 2017 and 2025. Gabon’s score of 2.8 on the GCI is lower than that of all peer countries, with the exception Cameroon. Cost overruns, delays in project implementation, and very high unitary costs plague flagship road projects. The enlargement of the 7 km N1 highway into a dual carriageway in Libreville has so far been allocated a total of FCAF 75.6 billion (including FCFA 13.5 billion in 2016 and FCFA 18.5 billion in 2017). The initial cost of this project had been estimated at a third this amount, or FCFA 26 billion in 2012. 33 The GOG paid the equivalent of FCFA 10.8 billion per km to expand the highway, while World Bank estimates 34 the average cost of similar types of highways in similar countries to be FCFA 371 million. 145. There is a high but unequal access to electricity in Gabon, and a supply gap could quickly emerge. Though access to electricity throughout the country is comparatively high (89.5 percent of Gabonese have access to electricity, compared to only 56.8 percent of Cameroonians), only about 42 percent of rural Gabonese have access to electricity (Figure 36). Gabon’s extremely high urbanization rate makes service delivery in highly populated urban centers easier than in low- density rural areas. There was a significant increase in power generation over 2010-15 (+30.5 percent overall), but growth in the electricity supply in Gabon was lower than in comparator countries: Equatorial Guinea increased generation by 131 percent and the Republic of Congo by 96 percent over the same period. Moreover, an electricity supply gap could potential arise in Gabon, as electricity consumption of 1,303 KWh per capita (or 2,200 GWh in total) in 2014 was 32 Gabon Ministry of Infrastructure. 33 Initial cost estimated by the engineering company Bechtel. 34 World Bank, Road Costs Knowledge System (ROCKS) 2.3 database and WDI for exchange rate. 36 barely covered by the 2,213.2 GWh generated. Electricity consumption grew by 5.9 percent per year over 2010-14, much faster than power generation (4.4 percent annual growth) over 2010-15. 146. Power supply efficiency, reliability and service quality are also poor, and the frequent and long outages impair the business environment. The GCI assesses the reliability of the electricity supply by measuring interruptions and voltage fluctuations. Though Gabon’s index value has improved since 2012, its 2.9 index value in 2016 is below the 3.4 average among comparators (Figure 37). Power transmission and distribution losses are high in Gabon and have increased since 2010, with losses of 23.4 percent of output in 2014, up from 19.5 percent in 2010. Losses were among the highest among comparators, with only Congo having higher reported losses in 2014 (Figure 38). Gabonese consumers experienced an average of 52 outages in 2016, compared to 23.3 and 42 outages for Cameroonians and Ivoirians, and only experienced 2.96 and 0.48 outages for Ecuadorians and Malaysians the same year. Outages lasted an average of 32 hours in Gabon, compared to 86 hours in Cameroon, 56 hours in Cote d'Ivoire, 2.1 hours in Ecuador, and only a half an hour in Malaysia. 35 Figure 37: Quality of Electricity Supply, (1-7), 2016 Figure 38: Electric Power Transmission and Distribution Losses (% of output), 2014 7.0 50.0 45.0 6.0 40.0 5.0 35.0 4.0 30.0 23.4 2.9 25.0 3.0 20.0 2.0 15.0 1.0 10.0 5.0 0.0 0.0 Source: World Economic Forum,GCI. Source: WDI, April 2017. Note: 1 = not reliable at all; 7 = extremely reliable. 147. While access to improved water sources is quasi-universal in Gabon, access to improved sanitation facilities is significantly lower. Only 42 percent of the Gabonese population has access to improved sanitation facilities, compared to 74.5 percent of Equatorial Guinea’s population. There is however higher access to improved sanitation facilities in Gabon than in Cote d’Ivoire, CAR, the Republic of Congo, or Chad. Gabon’s rural access to improved water sources and sanitation facilities (67 percent of the population) is lower than urban access (32 percent), as is the case in all peer countries (Figure 39 and Figure 40). 35 Doing Business data. 37 Figure 39: Access to Improved Water Sources (% Figure 40: Access to Improved Sanitation Facilities of Total Population), 2015 (% of Total Population), 2015 120 120 100 93.2 100 80 80 60 60 41.9 40 40 20 20 0 0 Source: WDI, April 2017. Source: WDI, April 2017. Public Investment Management 36 Economic Analysis of Investment Proposals 148. While the GOG conducts feasibility studies of major investment projects, their results are neither published nor accessible and it is difficult to get a good sense of the share of projects formally appraised. The 2015 budget allocated US$77 million (FCFA 47 billion) to feasibility studies, but most of the funds went to the Ministries of Youth and Sports, Internal Affairs, and Budget to appraise projects linked to the Africa Cup of Nations that was held in Gabon in 2017. The National Agency for Major Infrastructure Works (Agence Nationale des Grands Travaux d’Infrastructures, ANGTI) also funds feasibility studies. The only readily available studies are generally those commissioned by international donors who require studies to be made public. 149. Attempts to set up a systematic mechanism to fund the appraisal of economic projects have failed. PSGE projects were given priority access to financing for feasibility studies in 2012- 14 by the Maturation Fund set up within the Directorate of Budget, if they had terms of reference However, this Fund is no longer active for lack of resources, but is expected to be recreated under the auspices of the Office for the Coordination of the PSGE. Investment Project Selection 150. There is a legal framework for project selection and inclusion in the public investment program but no operational guidance, and line ministries do not comply with the 36 This section describes the current processes and the strengths and weaknesses of Gabon’s public investment management system following the World Bank methodology known as “The Eight Must Haves of PIM” (see Rajaram et Al, Editors. 2014. The Power of Public Investment Management: Transforming Resources into Assets for Growth. Directions in Development. Washington, DC: World Bank.). It does not cover all the eight areas, but rather focuses on the most relevant areas in the Gabonese context. 38 prioritization criteria and process for project selection. Decree n° 862 of October 28, 2013 requires a preliminary study, which takes into account the context and justification of the project, its feasibility, its logical framework, and the sustainability of its intended goals, to include a project into the public investment program. There are however no other operational guidelines clarifying the criteria and or procedures for projects selection before their inclusion into the list of investment projects annexed to the Finance Bill. Investment Project Costing 151. Aggregate forecasts for capital investments exist but do not detail the full cost of major investment projects. There is also no reference either to expenditures on equipment or the operating costs (and other recurrent costs) associated with the projects. There is no regulation or national directive imposing a detailed method of calculation of the costs over the lifetime of projects, even though the overall cost of projects should be known in advance to properly plan and include these costs in future budgets. The ANGTI does plan to introduce the multiyear management of investment project costs in 2017. Investment Project Monitoring 152. There is no annual consolidated report that presents the total cost and physical progress of key projects. Only a few 2015 Results and Action Planning documents, such as the one for the Ministry of Housing, include data on the physical progress and total cost of projects since inception. The 2015 general administrative account did provide information on development expenditures by program in 2015. The PSGE Coordination Office is working on the implementation of an information system that would collect comprehensive data on investment projects, and the ANGTI plans to expand its financial monitoring reports to include information on the physical implementation of projects. 153. Despite improvements in strategic tools, guidelines, and institutional frameworks, the Gabonese public investment cycle continues to suffer from weaknesses that adversely impact project execution and infrastructure quality. Weaknesses in project appraisal, planning, and selection are especially apparent in Gabon’s health sector (Box 3). 39 Box 3: Public Investment Management in the Health Sector Health investment planning does not reflect population density or health needs. The Supreme Audit Institution’s review of the major investments projects made in the health sector over 2000-111 flagged the lack of effective planning for hospital investment both in terms of buildings (construction and rehabilitations) and equipment. The construction of health centers is concentrated in the district capitals and follows administrative territorial divisions, with little to no regard to population density or specific health needs. For instance, the Bongoville and Ngouoni medical centers, which are about 20 minutes by car from each other, were built in low population density areas. Latest generation medical centers would have been more appropriate than mere health centers in the Ntem (Bitam) and the Lebombi Leyou (Moanda) districts. The construction of health centers happens outside of scheduled programming. Political leaders build health facilities without consulting the Ministry of Health and surrender the facilities ex-post to the ministry. The Ministry of Health must then allocate medical staff and equipment, with no corresponding budgets appropriations. There is no effective rehabilitation policy for health facilities, and rehabilitation budgets are wasted on infrastructure, which is later destroyed. A number of rehabilitated health facilities (e.g., the Melem regional hospital, the Ntoum health center, and the Nkembo hospital) still do not meet current sanitation and hygiene standards. Some health facilities (including the health center of Bongoville, rehabilitated for about US$1.7 million, and the Libreville Hospital (emergency and pediatric services), rehabilitated for a total of US$2.8 million) have been rehabilitated, but have subsequently been destroyed to be rebuilt. Source: Supreme Audit Institution’s value for money audit report on health infrastructure, 2013. The Wage Bill and Public Employment Management 154. Gabon’s wage bill is high by international standards, has grown beyond CEMAC- mandated ceilings, and is crowding out priority spending. Public spending on civil service pay reached 9 percent of GDP and 36 percent of government expenditure in 2015, while the government only allocated 25 percent of its total budget to investment in the same year. The Gabonese public-sector wage bill accounted for 53 percent of tax revenues in 2015, well above the CEMAC’s 35 percent convergence criteria. Payroll spending in Gabon is also one of the highest in CEMAC and among a broader set of comparable economies (as a share of GDP, a share of total revenue, and as a share of public expenditures) (Table 5 and Figure 41). Table 5: Regional Comparison of Gabon’s Payroll Spending with other Natural Resource-rich Countries Wage bill/GDP Wage bill/Total Revenue Wage bill/Public expenditures CEMAC (%) (%) (%) Congo 4 8 10 Gabon 9 42 36 Equatorial Guinea 2 3 2 Central African Republic 5 25 26 Other resource-rich countries Angola 9 23 28 Namibia 14 54 37 Nigeria 2 46 36 Dem. Repub. of Congo 4 17 19 Sources: WEO (2015). Observations: For Gabon: GDP, revenue, and expenditure statistics are from the IMF (WEO, 2015) and data provided by the Gabonese authorities for the wage bill. For other countries: IMF staff calculations. Latest available data: 2015 for Gabon; 2012 for: Congo, CAR, Angola, Ivory Coast, and Nigeria; 2011 for: Ghana and Namibia; 2010 for the Democratic Republic of Congo; and 2009 for Equatorial Guinea. 40 Figure 41: Emerging Economies: Wage Bill of Public Administrations, 2014 (% of tax revenues) Source: IMF. 155. Containing the wage bill is an immediate challenge for Gabon, but ensuring its sustainability will require structural reforms to improve the quality of the public administration. A reform of the recruitment and remuneration system is important to strengthen the quality of public administration and should include options to reward performance. At the same time, pressing fiscal sustainability concerns make containing the wage bill crucial in the short to medium term. The GOG recognizes the importance of containing the wage bill and is engaging in broader civil-service reform as articulated in both the PSGE and the Ministry of Public Service’s civil service reform strategy. 156. Wage bill growth was driven in part by strong growth in public sector employment. Between 2000 and 2017, the public-sector workforce went from 61,412 to 102,416 agents, and the wage bill has grown by nearly 70 percent, or 3 percentage points of GDP. While the increase in the wage bill was more than matched by buoyant growth in oil revenues between 2000 and 2010, payroll’s average annual growth rate was roughly 5 times higher than that of oil revenues between 2011 and 2015 (Table 6). By 2015, the wage bill accounted for some 42 percent of total revenue (compared to around 23 percent in 2010). Total public-sector employment was around 102,416 in 2015, compared to 81,500 employees in the formal private sector, making the civil service the largest formal employer in Gabon. Table 6: Average Annual Growth Rates of the Wage Bill, Total Public-sector Headcount, and Oil Revenues (2000-2015) Average annual growth rate Average annual growth rate 2000-2010 2011-2015 (%) Wage bill 7.56% 11.5% Public sector headcount 4.03% 6.84% Oil revenues 9.99% 2.34% Sources: World Bank staff estimates based on data provided by the Gabonese authorities, and oil revenue data from the U.S. Energy Information Administration. 157. The hiring of permanent staff, as opposed to contractual workers, drove the rapid expansion of public employment and the wage bill between 2010 and 2015. The number of 41 civil servants grew by about a third between 2010 and 2015, from 77,806 to 102,416 employees. A closer analysis of recruitment data during this period shows that permanent staff (including military personnel) increased by 56.8 percent, while the number of contractual workers decreased by 82 percent during the same period, as many workers were converted from contractual to permanent status. This reduced the net hiring of contractual workers and increased the level of permanent status employees, and put upward pressure on the wage bill, but did not influence overall total public employment as contractuals are included in public sector employment figures, along with permanent employees/civil servants. 158. The rapid growth in public employment threatens fiscal sustainability and limits the supply of skilled labor available in the private sector. Nearly two-thirds of workers with higher education work in the Gabon’s public administration, mostly in the education and health sectors. 37 159. Defense and security employees make up more than half of the public-sector workforce. Employment growth in the public defense and security sector outpaced growth in the public health and education sectors. 54 percent of public sector employees were defense and security staff in 2015. The education sector is the second largest public employer and accounted for around 26 percent of total public employment in 2015. The health sector and other general public services sectors employed 10 percent and 13 percent, respectively, of the remaining public sector workforce. 38 160. Since the program budgeting reform, employment ceilings are set within the Finance Law and the annual performance plans. The recently adopted program budgeting reform mandates the establishment of employment ceilings in the performance plans ministries prepare each year. If well enforced, such ceilings can help promote consistency with fiscal and service- delivery performance objectives. However, the government is unable to enforce the ceilings due to a weak hiring control system, which results in recruitments that often exceed the employment ceilings. The GOG’s employment ceilings decreased slightly between 2015 and 2016, as retired officials were not replaced, and the non-permanent workforce declined during this period. Yet, the ceiling for public-sector employment increased to 106,095 in 2017 (101,587 employees at the presidency and ministries and 4,508 staff seconded to institutions and administrative authorities). In the absence of a strategic staffing framework, public-sector recruitment is largely divorced from strategic staffing needs or the identification of actual constraints in the public sector. 161. Gabonese authorities have started to implement reforms to reduce the public-sector payroll. These are essentially PRE measures and include the non-replacement of 1,200 employees reaching retirement age. The statutory retirement age in the public sector is 60 years, with exceptions extending the retirement age to 65 years for specific public-sector bodies. Around 1,796 public sector staff are eligible for retirement during the 2016-18 period, including around 368 contractual workers and 1,426 civil servants (e.g., doctors, oral surgeons, and pharmacists; higher education professors and researchers; and staff of oversight and inspection bodies). 162. Though rising public employment put upward pressure on the public-sector wage bill, wage increases contributed more to the increase in payroll spending. The public-sector wage 37 World Bank, 2013, Report on Growth and Employment. 38 World Bank Staff estimates, Ministry of Budget and Public Accounts. 42 bill increased by 81 percent between 2010 and 2016, while economy-wide payroll spending only grew by 59 percent over the same period. The bill particularly grew in 2013-2014 after the government introduced PIPs. While staffing has impacted the growth of the public-sector wage bill, wage increases played a more important role (Table 7). Table 7: Percent Change in the Wage Bill and Staffing by Sector % change in wage bill (2010- % change in staffing Sector 2016) (2010-2016) Companies 48.2 66.3 Private 50.8 92.2 Parastatal 32.9 -57.3 Public Sector 81.8 39.4 Permanent workforce 88.5 36.9 Non-permanent 27.9 57.3 workforce Other (PIP) Local authorities 47.8 36.3 TOTAL 59.6 50.6 Source: Ministry of Economy, Economic Dashboard. 163. The share of bonuses and allowances in the public-sector wage bill has increased. Bonuses, family allowances, housing allowances, and other allowances grew from 10.4 percent of all wage expenditures in 2005 to around 33 percent in 2016. The GOG made several increases in allowances in recent years to improve civil servants’ standards of living, such as the significant increase of family benefits in 2009 and the increase and harmonization of housing allowances for about 60,000 civil servants in 2012. Sector-specific allowances were also granted in the education and health sectors, and PIPs were introduced in 2014. As a result, under the effective scheme until June 2015, allowances were larger than the base salary. Figure 42: Personnel Spending by Sector Figure 43: Average Wage by Sector Source: World Bank staff estimates based on BOOST Source: World Bank staff estimates, 2015. 2014 data. Governance includes the Ministry of Budget, Ministry of Economy, Ministry of Public Service, Ministry of Relations with Parliament, and the Ministry of Justice. 43 164. Governance and other public administration payroll costs make up the lion’s share of the public-sector wage bill, but the highest wages are paid by ‘institutions’, ‘administrative authorities’ and the education sector. The general public sector wage bill (i.e., the Ministries of Budget, Economy, Public Service, Relations with Parliament, and Justice payroll), made up nearly half of the public sector-wage bill and almost a third of total government public sector expenditures in 2015 (Figure 42). The average wage of a civil servant in Gabon was around FCFA 600,000 the same year. Institutions, administrative authorities, and the education sector paid the highest public- sector wages in 2015, while wages in the defense and security sector were relatively low (Figure 43). Recommendations 165. Several complementary reforms are required in the short and medium term to better manage the wage bill. The combination of a more performance-oriented approach and the introduction of different administrative tools (e.g., a new human resource management information system) should help the GOG better control the public workforce and manage the public-sector payroll. 166. In the short term, improving hiring controls is essential to a more efficient management of the wage bill. PRE measures include the extension of a recruitment freeze for a period of twelve months, except for priority ministries such as health, education, defense, and security. However, previous experiences show that recruitment into exempt ministries during a period of hiring freezes has nonetheless led to de facto recruitment in non-exempt ministries through secondments and internal mobility. Moreover, budgetary savings of previous hiring freezes have been minimal. Better management of hiring processes across the public sector is therefore a prerequisite for the efficient management of the wage bill. 167. Reforms focusing on improving budgetary and accounting management will allow for the adequate oversight of additional recruitment into the public sector. Such reforms could include: (i) the creation of budgetary items in line with the program budgeting budgetary nomenclature (classified by each mission program) to improve the accountability of program managers and financial controllers to enforce hiring controls; (ii) the creation and submission of a provisional job management and staffing document to provide financial controllers with a forecast of the payroll and staffing changes for the coming year (or the next three years); (iii) the establishment of a ‘prior approval’ function for individual recruitments that will not be granted in the absence of a vacant budget item; and (iv) the implementation of a monthly reporting requirement on the use of departmental remuneration credits by the Central Department of Financial Affairs (Direction Centrale des Affaires Financières, DCAF). 168. The Ministry of Civil Service and Administrative Reform has initiated human resource management reforms to improve strategic staffing and control the wage bill. These reforms include establishing an inventory of staff, performing an audit of physical structures, and creating job descriptions. Ongoing reforms concerning human resource directorates and the organization and mandates of the administration are essential steps for better control of the payroll. Additional reforms aim to improve the consistency of the payroll with employment ceilings for each mission at the departmental level, link human resource performance to program performance, 44 and facilitate a better forecasting of needs through skills management using forward planning of employment and skills (GPEEC in the French acronym) and workforce monitoring. 169. A human resource management information system (HRMIS) could greatly contribute to better management of the wage bill and enhanced service delivery in Gabon. The GOG needs a better system to monitor and report on the wage bill, as current data on wage expenditures are scarce and inconsistent. Beyond aggregate compensation and employment data, the HRMIS could monitor the composition of pay (i.e., base wages, allowances, bonuses, in-kind compensation, overtime, and social security contributions); the areas of employment (e.g., by functional classification including education, health, and security); the type of employment (i.e., permanent or contractual; full time or part time); and the type of worker (i.e., clerical, service, professional, or managerial). Modern IT systems could provide disaggregated data across these areas. Access to better quality data could greatly help: i) improve the management of the wage bill through better prepared budgets and payroll forecasting; ii) monitor wage and employment over the medium term and evaluate their consistency with fiscal objectives; iii) achieve better workforce planning; and iv) design and evaluate wage and employment measures to improve service delivery. Finally, several developing countries have introduced biometrics to help them increase public sector efficiency, eliminate double-dipping and ‘ghost workers,’ and monitor employee attendance. 170. Payroll rationalization could help rebalance Gabon’s distribution of public spending across sectors. Payroll and public employment rationalization may decrease the general public services’ share of total government spending. 171. Significant investments are needed to address the poor state of the country’s infrastructure. To reach its infrastructure goals, the GOG needs to allocate additional resources to its strategic priorities and make substantial reforms in the management of public investments. Despite improvements in management tools, guidelines, and institutional frameworks, the Gabonese public investment cycle continues to suffer from weaknesses, adversely affecting the execution and quality of investment projects. As a result, the quality of the country’s roads and transport infrastructure has not improved since 2010, and Gabon’s power sector is marked by an inefficient and unreliable supply of power, poor service delivery, and unequal access to electricity. The Gabonese population could quickly be faced with an electricity supply gap, and frequent and long power outages impair the business environment. Going forward, the GOG should: i) improve allocative efficiency to better mobilize existing resources in support of national priorities; ii) improve the capacity to implement sustainable projects should additional fiscal space for investment be created through expenditure savings (especially in payroll) and/or more revenue; and iii) better plan, appraise, select, and monitor investment projects to improve the quality of the country’s infrastructure. Gabonese authorities should therefore study options to create better public investment management systems and commission or undertake detailed reviews. 172. The GOG should consider three key reforms to improve public investment management. (i) Update the National Infrastructure Plan to ensure greater consistency between sector plans, mandate regular reviews and updates, and take into account realistic estimates of medium- to long-term resources needed to implement projects during the project selection process. (ii) Systematically analyze the economic feasibility and financial sustainability of a small number of the highest-value projects. The government plans to publish cost-benefit analyses for projects 45 whose cost exceeds FCAF 20 billion in annex to annual budget laws. The sophistication of appraisal methods should however be proportional to the scale and complexity of projects and to available skills. The FCAF 20 billion threshold may need to be (periodically) reviewed. (iii) Monitor the financial and non-financial performance of investment projects more actively, starting with the largest and/or key projects under the PSGE. Closer monitoring of investment projects (for which there are currently plans but no concrete actions or dedicated budgets) should help reduce cost overruns, implementation delays, and unitary costs. 46 4. EDUCATION Introduction 173. This chapter examines recent trends in education expenditures. It reviews the key characteristics of Gabon’s education sector, including the status of key indicators and the composition and distribution of education spending. It also provides recommendations for how to improve the efficiency of education expenditures, which is critical during government fiscal consolidation efforts. 174. The unavailability of data and the inconsistency between different budget sources constrain the analysis. Although statistical collection services exist in each provincial directorate, there are no regular statistical yearbooks that could yield information on education indicators. Since there is no one consistent data source for information on education spending in Gabon, the analysis is based on several internal and external sources. First, the Finance Laws provided aggregate-level data for the education sector, with details on programming objectives and expected results available in the PAPs (for 2015 and partially for 2016-17). Second, the 2015 settlement law (Loi de règlement 2015) 39 provided the bulk of the data for the expenditure analysis, complemented by the 2015 PAP and the RAP. 40 Third, the 2012 Demographic and Health Survey (DHS) provided valuable inputs, and data were also collected form the three ministries in charge of education and training as well as the Ministry of Finance. Moreover, data from 2010 to 2014 were obtained from the database of the United Nations Educational, Scientific and Cultural Organization UNESCO, which also provided information on school attendance and completion rates that were comparable to other countries. 41 Finally, The French Development Agency provided school map data. Education Sector Overview Education System and Service Delivery 42 175. The country’s education system is divided into preschool, primary and secondary education, technical education and training, and higher education. 43 Three ministries oversee education in Gabon: the Ministry of National Education and Civic Education, Ministère de 39 Data for 2016 is not available. 40 The BOOST database, based on the Finance and Settlement Laws, PAPs and RAPs, currently covers the years 2014 and 2015, but still requires harmonization in classifications, following the introduction of budgeting by program objectives in 2015. I 2014, the program classification remains to be defined and some captions are missing, making it difficult to compare to 2014 and 2015. The BOOST 2015 data could have been used in this analysis but there is a gap between the expenses recorded in BOOST 2015 and those in the 2015 Settlement Law (Loi de règlement 2015). It turns out that only 93 percent of total expenditure and 84 percent of education expenditure are recorded in BOOST. As for all other social sectors, salary data are also not disaggregated by sector in the executed spending data. The same gap is also true between the published 2015 Finance Law and the budget in BOOST 2015. 41 The UNESCO database is the result of a careful consolidation of public spending data based on the United Nations Institute of Statistics methodology. Care is needed when comparing pre-2015 data with data for 2015 and later because of the transition to budgeting by objectives. 42 Detailed information on the education system and service-delivery characteristics is provided in the Social Spending Annex. 43 The country’s education sector is governed by Law No.21/2011 of February 14, 2012. 47 l'Education Nationale et de l'Education Civique (MENEC) is responsible for primary education and general and technical secondary education; the Ministry of Labor, Employment, Technical and Vocational Training, and Youth Integration (Ministère du Travail, de l’Emploi, de la Formation Technique et Professionnelle et de l’Insertion des Jeunes, MTEFTPIJ) focuses on vocational training; and the Ministry of Higher Education, Scientific Research, and Management Training (Ministère de l’Enseignement Supérieur, de la Recherche Scientifique, et de la Formation des Cadres, MESRSFC) is responsible for public and private higher education. Preschool is not compulsory and essentially private. Since 2005, primary education includes five years of study, and in fifth grade students must pass the Certificate of Primary Studies (Certificat d’Etudes Primaires, CEP) to advance to general secondary education, which spans a seven-year period divided into two cycles. The first cycle lasts four years, while the second cycle is three years. At the end of the first cycle, students must sit for the First-Cycle Certificate of Study (Brevet d’Etudes du Premier Cycle, BEPC) examination, and they must pass the Baccalauréat to access higher education. 176. Enrollment in primary education increased from 265,714 students in 2001 to 317,946 students in 2011, an increase of 20 percent over a ten-year period (Table 8). However, primary education enrollment decreased by about 7 percent between 2011 and 2015. In the 2015-2016 academic year, there were about 296,000 students enrolled in primary schools, of which about 172,000 attended public-sector schools. During the same academic year, there were 11,001 primary and secondary school teachers employed in Gabon’s education system, of which 6,456 worked in the public sector and 4,550 worked in the private sector. Table 8: Enrollment in Pre-primary, Primary, and Secondary Education, 2001-2015 2001 2002 2003 2011 2015 Pre-primary 15,568 na na 45,225 67,400 Primary 265,714 281,871 279,816 317,946 295,870 Lower secondary 75,673 78,546 na na 139,227* Upper secondary 25,045 26,645 na na 46,090* Secondary 100,718 105,191 na na 185,317* Sources: UNESCO database, 2001 to 2011; MENEC for 2015. Note: *Enrollment in secular private secondary education is unknown. 177. The technical and vocational education and training (TVET) subsector represents a limited share of total enrollment in secondary education. Only about 8 percent of students enrolled in post-secondary education attend TVET institutions. This compares unfavorably to other emerging economies where enrollment rates for TVET often reach 30 to 40 percent. Most TVET is provided by public institutions that offer five types of training modules with varying length and duration depending on the subject and profession. 44 In 2015-2016, there were about 8,600 students and 807 teachers in Gabon’s TVET system. 178. The Ministry of Higher Education oversees tertiary education and scientific research and innovation. There are three public universities 45 and four graduate schools. 46 The ministry started transitioning to a Bachelors-Masters-PhD system in 2007, but implementation only started 44 See Annex A for more information. 45 The Université Omar Bongo, Université des sciences et de la santé, and Université des science et techniques de Massuku 46 The Normal School, Technical Normal School, Management School, and the Institute of Technology) 48 in 2016. About 20,000 students are enrolled in tertiary education in Gabon, with a high concentration of students at the University of Omar Bongo. 179. Gabon has well-developed private primary and, to a lesser extent, secondary education sectors. About 42 and 28 percent of all students 47 are enrolled in private primary and secondary education, respectively, which is high compared to comparator countries and many upper middle-income countries. There are two types of private education institutions in the country: denominational or church-based private schools (i.e., Catholic, Christian Alliance, Evangelic, or Muslim schools) and secular private schools. 48 Moreover, private education institutions can be recognized as public utilities (reconnaissance d’utilité publique, RUP), and more than one third of all secondary denominational schools were RUPs in the 2015-16 academic year, which made them eligible for public subsidies to cover their operational budgets and potentially the cost of seconded teachers. 49 Education Outcomes 180. Although literacy rates in Gabon are low by international standards, the country has made substantial progress over the last two decades in increasing enrollment rates in primary education. In 2015, the country’s literacy rate was 89 percent, compared to an average rate of 99 percent for middle-income countries, translating into more than 25,000 illiterate Gabonese aged 20-25. However, with a net enrollment rate (NER) of 93 percent in primary education, the country has practically achieved universal primary enrollment. Moreover, the net school attendance rate in primary education was estimated at 88 percent in 2012, 50 placing the country well above the average of SSA 51 and ahead of many of its peers such as Cote d’Ivoire, the Democratic Republic of Congo, and Cameroon (Figure 46). Also, late entry is not a concern in Gabon’s primary schools, as 62 percent of first graders were six years old or younger, and 83 percent were seven years old or younger (Figure 44). 52 181. The overall attendance rate in secondary education 53 is high by regional standards, although the age distribution in secondary schools is a concern. Gabon’s secondary school attendance rate was 55 percent in 2012, which was better than those of many SSA countries, 54 and its attendance rate in lower secondary schools outperformed those of many of its peers (Figure 46). However, the country performed worse than some its peers, such as the Democratic Republic of 47 Data for secondary education are estimates. 48 In Gabon, there are a total of 765 pre-primary schools, 890 primary schools (528 secular, 226 catholic, 92 Christian Alliance, 40 Evangelic, and 4 Muslim schools), 157 secondary schools (115 secular, 22 catholic, 10 Christian Alliance, 8 Evangelic,2 and Muslim schools), and 12 secular technical secondary schools that are private providers of education. 49 Seconded teachers are provided to schools on a case-by-case basis. They are civil servants allocated to private RUP schools. 50 According to UNESCO’s estimates from the DHS. 51 For example, attendance rates at primary schools are 65 percent in the Gambia, 55 percent in Guinea, 67 percent in Ivory Coast, and 61 percent in Ethiopia. 52 Based on DHS estimations. 53 The 2012 DHS still mentions Grade 6 as primary and Grade 7 as junior secondary. 54 Attendance rates at secondary were 42 percent for Kenya in 2014, 46 percent for Zimbabwe in 2011, 45 percent for Cameroon in 2011, 37 percent for the Gambia in 2013, 25 percent for Guinea in 2012, 21 percent for Mozambique in 2011, 19 percent for Niger in 2012, and 35 percent for Zambia in 2013. 49 Congo and Cameroon, in upper secondary enrollment. Unlike in primary education, late entry in secondary education appears to be a challenge in Gabon, as only 27 percent of students in sixth grade were aged twelve or under, while 39 percent were aged thirteen and fourteen, and 34 percent were over the age of fifteen (Figure 45). 55 Figure 44: Per Age Distribution for Grade 1 Pupils Figure 45: Per Age Distribution for Grade 6 Pupils in Primary Education in Secondary Education 50 Pourcent 42.6 20 Pourcent 17.3 18.7 Grade 6 Grade 1 15.9 16.6 40 15 11.6 30 19.5 21 8.9 10 7.1 20 8.3 3.9 10 4.8 3.9 5 0 0 moins 6 7 8 9 10 ans et <= 9 ans 10 11 12 13 14 15 >=16 de/et 5 plus ans ans Source: Authors’ calculations based on DHS 2012. Figure 46: Net Attendance Rate in Primary, and Junior, and Senior Secondary Education in Gabon and Peer Countries Ecuador 96.5 Côte d'Ivoire 67.4 Upper secondary D.Republic of Congo 83.9 Lower secondary Chad Primary 49.8 Cameroon 78.5 Gabon 88.4 0 20 40 60 80 100 120 Source: UNESCO’s data base from 2011 to 2013 (Gabon, 2012). 182. Less than 20,000 students were enrolled in Gabon’s tertiary education institutions in 2016. In addition, the country’s gross enrollment ratio for tertiary education was only 8.44 percent, compared to over 20 percent in South Africa and Botswana, and over 25 percent in Malaysia and Ecuador. However, meaningful comparisons are difficult to make since data from Gabon are from 2003. 183. While the country’s primary school completion rate is relatively high, the rate in upper secondary education is very low. Gabon’s primary completion rate was estimated at 76 percent in 2012, placing it among the best performing SSA countries (including some of its peers) (Figure 47). However, only 42 percent of students in lower secondary schools graduated during the same year. Moreover, the country’s upper secondary completion rate was a meager 11 percent in 2012, placing Gabon among the least performing countries in SSA. 55 Estimates are based on data from the 2012 DHS. 50 Figure 47: Completion Rates at Primary, Lower Secondary, and Upper Secondary Education Levels Zimbabwe 88 72 Nigeria 50 Kenya Kenya Ghana Namibia Kenya Gabon 76 Nigeria Namibia Zambia Gambia Nigeria Zambia Zambia Cameroon Uganda Ghana Gabon 42 Cameroon Lesotho Malawi Gambia Lesotho Guinea Malawi Côte d'Ivoire Côte d'Ivoire Guinea Ethiopia Guinea Zimbabwe Mozambique Uganda Lesotho Ethiopia Gabon 11 Uganda Mozambique Chad Niger Mozambique Niger 6 Niger 2 Chad 27 0 20 40 60 80 0 20 40 60 0 20 40 60 80 100 Primary Lower secondary Upper secondary Source: UNESCO’s data base from 2010 to 2015 (Gabon, 2012). 184. Gabon’s school dropout rates are initially relatively low but increase with age.56 Nearly 2 percent of thirteen-year-olds were not in school in 2012, while this reached 10 percent for sixteen-year-olds (Figure 48). Among students under twenty years old who dropped out of school, one-third did not complete the primary cycle, 12 percent completed the primary cycle, and more than half did not finish secondary school. Figure 48: Schooling Status of Gabonese Children and Youth by Age, 2012 100% 80% 60% 40% 20% 0% 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 At school Dropped-out Have never attended school Source: Authors’ calculations based on DHS, 2012. 185. Gabonese students performed above average for Francophone African countries in 2006, although performance decreased over the school-cycle. The country had the highest score in French at the end of the primary cycle among the Francophone countries participating in the PASEC VII-VIII-IX evaluations 57 (Figure 49). The average score of Gabonese students in mathematics was also above the group average (44.3 versus 39.6), placing the country in fourth place after Madagascar, Cameroon, and Burundi. The trend continued in the second year of the cycle, where Gabon (53.6) came in second behind Cameroon (65.5) in French and ranked fourth 56 The 2012 DHS. 57 This is the latest PASEC Gabon participated in. The trend in the other countries have changed as per the PASEC 2014. 51 in mathematics. In 2016, the CEP, BEPC, and Baccalaureat pass rates were 69, 37, and 42 percent, respectively, with decreasing performance over the school-cycle. Figure 49: Average Scores for Mathematics and French at Grade 5, 2006 Gabon 61.4 Madagas… Cameroun Cameroun Burundi Burundi Moyenne 40.2 Gabon 44.3 Sénégal Sénégal Burkina Moyenne 39.6 Côte… Burkina Congo Comores Madagas… Congo Comores Tchad Tchad Bénin Bénin Côte… 0 20 40 60 80 0 20 40 60 French Mathematics Source: PASEC VII-VIII-IX results, CONFEMEN 2012. Note: Tchad (2004) - Benin, Madagascar et Cameroun (2005) – Gabon (2006) –Burkina, Congo et Sénegal (2007) – Burundi, Comores et Côte d’Ivoire (2009). 186. The rate of return to education in Gabon is high at the secondary level, pointing to a combination of scarcity of and demand for skilled labor. Wageworkers with completed primary education earn 13 percent more than workers who did not complete primary school (Figure 50),58 which is relatively low compared to other countries, suggesting relatively high completion rates, quality issues, and/or limited demand for primary education graduates. However, workers who completed a full cycle of lower secondary education earn 28 percent more than workers who exited the system after having completed primary school, 59 and workers with TVET qualifications earn 14 percent more than workers who only completed lower secondary school. Moreover, higher education graduates earn 23 percent more than non-graduates, with higher returns for women (34 percent) than for men (19 percent). These results point to a need to improve the quality of primary education as well as increase the coverage of higher education. Additionally, the higher returns to TVET and upper secondary and tertiary education suggest there is a market for private financing and delivery, which could be incentivized through PPPs. 187. There is also scope to improve the employment outcomes of secondary, TVET, and higher education graduates. The unemployment rate is about 24 percent for secondary education 58 The analysis draws on a sample of salaried workers surveyed by the national survey on employment and unemployment (enquête nationale sur l’emploi et le chômage, ENEC) whose earnings information was available. Observations from this sample were complemented with data regarding household consumption. When estimating returns to education based on household consumption, the level of education of the head of household is generalized to the entire household. Analysis was performed using both the number of years of education held by the head of household and categorical dummies for the level of educational attainment. 59 In general, the return accruing to an additional year of schooling is inversely related to the educational profile of the labor force. In other words, as a population becomes generally more educated, the return accruing to an individual through additional years of schooling becomes smaller. In Gabon, only 28 percent of the labor force had completed a full cycle of secondary education in 2010, although the proportion of the labor force that had not completed primary school is relatively low compared to other African countries. 52 graduates and 27 percent for TVET graduates, suggesting there is a miss-match of skills since the demand for skilled workers is relatively high. Figure 50: Mincerian Regression Result: Earnings Increase by Level of Education (Percentage) Completed primary Lower secondary TVET Upper secondary Higher education 120% 111% 104% 100% 86% 79% 82% 80% 69% 72% 70% 67% 63% 62% 55% 55% 58% 60% 50% 52% 42% 45% 41% 42% 35% 40% 35% 34% 23% 23% 13% 14% 14% 20% 8% 9% 0% Total Male Female Total Male Female Wage employment Household Income Source: PAD for the Skills Development and Employability Project, 2016, National Survey of Employment and Unemployment, 2010. 188. While Gabon’s net attendance rate in primary education varied little by gender or income level in 2012, they were slightly higher in rural areas (Figure 51). Although the primary attendance rate was similar across income quintiles and between boys and girls, it was slightly higher in rural areas (92 percent) than in urban (88 percent), which was likely related to the early transition of wealthier and urban students to lower secondary schools. The Woleu-N'tem, Haut- Ogooué, and Libreville/Port-Gentil regions recorded the highest attendance rates (over 90 percent), whereas the Ogooué-Ivindo region recorded the lowest (85 percent). 189. However, there were significant differences in attendance rates for secondary schools in 2012 (Figure 51). The attendance rate in rural areas (26 percent) was much lower than in urban areas (60 percent), likely reflecting shortages of secondary schools and/or teachers in rural areas (compounded by a limited supply of private secondary schools). Moreover, only 25 percent of children from very poor households attended secondary school, compared to 72 percent of children from the wealthiest households, pointing to important differences in direct and opportunity costs for poor households (as well as possible overlaps with the urban/rural divide). Finally, girls attended secondary schools at a higher rate (68 percent) than boys (60 percent). 53 Figure 51: Net Attendance Rate in Primary and Secondary Schools, by Gender, Income Quintile, and Place of Residence, 2012 Primary 100 Secondary 100 88.9 89.6 90.6 87.7 88.9 87.7 91.8 88.4 90 84.9 87.8 90 80 80 69.1 72.2 70 70 58.9 57.9 59.5 54.6 60 50.8 60 44 50 50 40 40 24.8 26.1 30 30 20 20 10 10 0 0 Source: Authors’ calculations based on UNESCO’s data base, 2012. Overview of Education Expenditures: Funding, Sources, and Trends Budget and Spending Processes 190. The GOG started transitioning to program-based budgeting in 2015.60 As a result, the government’s budget has been structured into programs by objectives since 2015, and authorities are now required to prepare RAPs and PAPs. Program budgets in the 2017 initial budget law was presented with a multi-year perspective over three years. Moreover, payment appropriations (credits de payment, CP) and commitment authorizations (i.e., committed spending) had the same funding levels, which were very close to the executed spending level. 191. There is a clear budget process for the central government in Gabon. For the three ministries in charge of education and training, the preparation of the budgets is the responsibility of the principal secretary within the DCAF. Each ministry’s needs are expressed by the managers responsible for programs (responsables de programme, RPROGs). The budget framework is prepared by the Ministry of the Budget and Public Accounts in July of each year, and inter- ministerial arbitrations take place in August, followed by internal arbitrations with the RPROGs. Finally, there is a conference with the General Directorate of Budget and Public Finance. Departments may be operational units (unités opérationnelles, OUs) if they are not RPROGs. The external services and the autonomous institutions or operators are either operational program budgets (budgets opérationnels de programme, BOPs) or OUs. Consequently, all the managers responsible for BOPs or OUs can initiate both commitment and expenditure scheduling. 192. There are multiple stakeholders involved in the implementation of MENEC’s programs. The budget of MENEC is divided into four programs: (i) Piloting and Support; (ii) Pre- primary and Primary Education; (iii) Secondary Education; and (iv) Technical and Vocational Education. The Director General of General and Normal Education is responsible for programs (ii) and (iii), and only one person oversees program (iv), although there are two directors: the Director General for Technical and Vocational Education and the Director General of Investment 60 Pursuant to the LOLFEB legislation in 2014. 54 and Equipment Programming. The central services of MENEC are supported by nine provincial academy branches, divided into 24 school districts that are the OUs responsible for the BOPs at their levels. The credits of the OUs are delegated to the provincial treasuries. There is a budget supervisor at MENEC and in each province who performs on-site verification of service delivery. MENEC’s budget is prepared based on a performance conference that is chaired by the Minister of MENEC. The DAPs aggregate the needs of each OU, but the final decision rests with the Directorate General of Budget and Public Finance. 193. The MESRSFC has four programs, each managed by a separate authority. The Directorate General of Tertiary Education, the Directorate General of Scientific Research and Innovation, the Directorate General of Academic Activities, and the permanent secretary are all responsible for the MESRSFC’s programs. The nine public establishments have management autonomy, and their rectors or general managers are responsible for OUs or BOPs, depending on the case. Half of the credits are paid directly by the Ministry of Budget and Public Accounts at the beginning of each financial year, which approaches 100 percent at the end of the budget year. Annual performance contracts are established with each OU or BOP for a period of three years. 194. Spending execution in 2015 was punctuated by freezes and an early closure of the commitment period. The freeze in allocations in the revised Finance Law affected the contracts awarded by public procurement, generating execution stoppages and litigation. The execution of capital expenditure was also limited by the 15 percent freeze on allocations in the revised Finance Law, the early closure of the commitment period for the 2015 financial year, and procurement delays. Education Expenditure Levels and Trends 195. The share of public expenditure allocated to education is difficult to compare after the state budget was restructured in 2015, although it is estimated that spending remained at best flat between 2010 and 2016. From 2010 to 2013, public spending on education was estimated at an average of 3 percent of GDP, before falling to 2.7 percent in 2014 (Figure 52). However, the reforms in 2015 make comparisons to pre-2015 data difficult. Moreover, pre-2015 data likely underestimated the share of the education budget in the country’s GDP. The subsequent restructuring process integrated salaries, bonuses, and executed spending into the 2017 programs, which were previously centralized at the Department of Budget and Public Finance within the Ministry of Finance. These changes predict a slight initial decrease in the share of the budget allocated to education as share of GDP, from an estimated 2.9 percent in 2015 to 2.8 percent in 2016, before increasing to 3.9 percent of GDP in 2017. However, data for 2017 will depend on the revised budget, the budget-execution rate, and the actual GDP growth rate at the end of 2017. 55 Figure 52: Public Expenditure on Education as % GDP in Gabon, 2010-2016 4.0 Pourcent 3.5 3.08 2.99 3.01 3.03 2.90 3.0 2.67 2.77 2.5 2.0 2010 2011 2012 2013 2014 2015 2016 Dépenses publiques d'éducation en % PIB Budget d'éducation en % PIB Source: 2010 to 2014 series are from the UNESCO database. 2015 and 2016 series are based on the credit budget (revised finance law 2015-2016) and GDP series from the IMF’s WEO Database. 196. As a percentage of GDP, Gabon’s level of spending on education is far below the average of both SSA and peer countries. From 2010 to 2014, SSA countries spent an average of 4.3 percent of their GDP on education, 61 compared to 3 percent in Gabon. The country also performed poorly compared to peers such as Botswana (9.6 percent), Malaysia (6 percent), and Ecuador (4.9 percent) (Figure 53). Even with its estimated rate of 3.9 percent for 2017 (which includes salaries and bonuses), Gabon remains behind both peer countries and the average for SSA. Figure 53: Public Expenditure on Education as a % of GDP and Total Government Expenditure in Gabon and Peer Countries, 2009-2015 South Africa 6.0 19.1 Malaysia 4.9 19.7 Ecuador 4.9 12.8 Côte d'Ivoire 5.0 21.2 D.Republic of Congo 2.2 16.9 Chad 2.9 12.5 Central African Republic 1.2 7.8 Cameroon 3.0 13.8 Botswana 9.6 20.4 Gabon 2.7 11.4 0.0 5.0 10.0 15.0 20.0 25.0 Expenditure on education as % of total government expenditure Government expenditure on education as % of GDP Source: UNESCO’s database from 2009 to 2015 (Gabon, 2014). 197. Public education spending as a share of total public expenditures also remained fairly constant in recent years. From 2010 to 2014, education spending represented an average of 11.4 percent of all public spending, ranging from 13.3 percent in 2010 to 10.6 percent in 2012 (Figure 54). After the introduction of program-based budgeting, the education budget increased to 12 percent of the total budget in 2015, before settling at 11 percent in 2016. The forecast for 2017 estimates that education spending reached 19 percent of all government expenditures, as wages were integrated into education programs. 61 The UNESCO database (2010-2014). 56 Figure 54: Public Expenditure on Education as a % Public Expenditures in Gabon, 2010-2016 Pourcent 16.0 13.3 11.3 11.4 11.9 10.6 10.7 10.9 12.0 8.0 4.0 0.0 2010 2011 2012 2013 2014 2015 2016 Dépenses d'éducation en % dépenses publiques Budget éducation en % budget total Source: 2010 to 2014 series are from the UNESCO database. 2015 and 2016 series are based on the credit budget (revised finance law 2015-2016). 198. Gabon spends far less of its public-sector budget on education than most SSA and peer countries. SSA countries and upper middle-income countries in SSA spent an average of 17 and 15 percent, respectively, of their government budgets on education between 2010 and 2014, compared to 11.4 percent in Gabon (Figure 53). Moreover, the country spent less on education than its peers such as the Democratic Republic of Congo (16.9 percent), Cameroon (13.8 percent), and Ecuador (12.8 percent). The forecasted 2017 budget allocated to education may be on track to at least match the regional average. 199. Unit costs in Gabon’s primary education sector is lower than those of upper middle- income SSA countries. Gabon spent $442.8 per primary student in 2011, while upper middle- income SSA countries spent more than $600 (Table 9). Moreover, as a share of GDP per capita, government funding per primary student was 4.4 percent in Gabon, the lowest in SSA (Namibia had the highest at 17 percent). As a result, Gabon’s spending ratios are more aligned with lower income countries. Limited public spending on education coupled with an overblown primary student population point to important allocative and technical inefficiencies in the country’s education system. Table 9: Primary School Unit-cost Comparisons between Gabon and SSA Countries Government Expenditure per Primary Student Initial Government Funding per Primary Student (2011) Nominal Years Data Nominal As share of GDP Constant US$ Available Constant US$ Per Capita Upper middle income Gabon 442.8 2011 438.7 4.4 Botswana 693.7 2009 637.0 10.3 Mauritius 1033.2 2010-2014 897.7 9.7 Namibia 912.8 2010 843.8 17.7 South Africa 1328.0 2010-2014 1172.4 17.4 Lower middle income Cameroon 78.5 2010-2012 76.3 5.8 Côte d’Ivoire 180.6 2011-2014 155.1 12.2 Ghana 282.5 2011-2014 297.5 23.4 Lower income Gambia 81.7 2010-2013 47.9 11.0 Guinea 47.2 2010-2014 43.5 7.9 Malawi 27.3 2010-2014 29.8 5.7 Mozambique 89.9 2013-2014 73.8 12.8 Niger 92.7 2010-2014 81.2 21.0 57 Uganda 25.5 2010-2014 45.3 7.1 Zimbabwe 195.4 2012-2013 193.1 21.5 Source: The UNESCO database (2010-2014). Sources of Education Expenditures 200. Until recently, Gabon has been relying almost exclusively on its internal public resources to finance public education. The presence of international donors in the country’s education system was negligible up until 2016, although a lack of registered data on foreign aid (on-budget as well as off-budget) make it difficult to estimate. The European Union provided the country with US$5.1 million between 2007 and 2011 for two projects focused on enhancing TVET, and it provided assistance for improving governance in the overall education sector between 2008 and 2013. 62 At present, there are several development partners in Gabon that provide education assistance, including the World Bank, which supports the TVET subsector with US$100 million over the 2016-21 period; the African Development Bank, which supports the TVET subsector with an additional $65 million over the 2016-21 period, and the French Development Agency, which provides Euro 150 million to help finance primary and secondary education over a period of eight years. 201. While there are no official estimates of private financing in the country’s education system, the importance of private schools suggests it is significant. The available household survey does not provide information on the education expenditures of households, which is a serious constraint that will be addressed in the upcoming 2017 household survey. Since 40 and 28 percent of students are enrolled in private primary and secondary education, respectively, the level of private financing could be twice the size of the public education budget. As a result, private and public spending on education could be close to 6 percent of GDP in Gabon, although the use of PPPs in secondary education, which involves significant public financing of private providers, likely puts it closer to 5 percent of GDP. Thus, Gabon has achieved a high private enrollment rate compared to its public spending and income level (Figure 55). 62 The project was financed between 2008 and 2013 from the European Union’s 10th European Development Fund. 58 Figure 55: Private Primary Enrolment share in Compared to Income Level and Education Spending, 2010- 2016 Private Primary Enrollment Share 15 Performance relative to government spending on education Gabon Better than average 10 Worse than average Republic of Congo Ecuador 5 Chad Cote d'Ivoire Malaysia South Africa 0 0 .005 .01 .015 Worse than average Better than average Performance relative to income Source: World Development Indicators: average 2010-2016. (*) Performance relative to spending: private primary school enrolment share/government spending (%GDP); Performance relative to income: private primary school enrolment share /GDP per capita. Analysis of Expenditures: Efficiency Issues 202. Based on primary completion rates, Gabon is an “underachiever,” as it does not spend enough on education and has a low rate of return on its education investments. The country is expected to spend more on education and have a higher primary completion rate considering its income level, population, and geographical location. Instead, Gabon is grouped with countries that underperform in terms of education outcomes 63 and have relatively low completion rates and education expenditures, while countries such as Namibia or South Africa (upper middle-income countries like Gabon) are “overachievers” (Figure 56). As a result, Gabon’s primary education sector suffers from both technical and allocative inefficiencies, 64 as illustrated by its high repetition rates, limited emphasis on critical inputs such as teacher training, and over emphasis on higher education. Furthermore, the country is also an “underachiever” based on its secondary completion rates, as there are significant technical inefficiencies in terms of high repetition rates, deficits, and poor allocation of critical human and physical resources in Gabon’s secondary schools. 63 Quality of education is not captured in this analysis, as Gabon might be relatively more efficient than other countries given its better performance on the 2006 PASEC test scores. 64 Analyzing technical/productive efficiency typically entails analyzing: (a) “what” inputs or outputs are bought with the resources; (b) “how” the inputs or outputs bought are used to maximize impact or decrease cost; and (c) “where” resources are invested. Analyzing allocative efficiency typically entails assessing whether resources invested meet society’s needs, maximize the welfare of the community, or whether a society has the right mixture of, say, education/health care programs to maximize the education/health of its members. 59 Figure 56: Efficiency of Government Expenditures based on Primary Completion Rates, 2010-2014 Internal Efficiency Issues 203. Based on available repetition and completion data, Gabon’s education system is internally inefficient. Even though repetition and promotion rates for each primary grade are not available, the percentage of repeaters by school-cycle point to serious internal inefficiencies. At an average of 37 percent, Gabon is among the countries with the highest percentage of repeaters in primary education, and an average of 19 and 24 percent of its students in junior and senior secondary schools, respectively, are repeaters (Figure 57). 65 Although 83 percent of first graders are seven years old or younger, only 36 percent of students in the last year of primary school are aged between ten and eleven, while 17 percent are twelve years old, and 46 percent are older than thirteen years old. 66 The combination of high repetition and low completion rates point to a substantial waste of financial resources, including dedicating classrooms each year to repeaters. 67 Aside from financial reasons, a lower repetition rate in primary education could make more primary teachers available to be retrained as lower secondary school teachers. 65 2012 Country Status Report (RESEN 2012). 66 Estimates based on data from the 2012 DHS. 67 Assuming 37 percent repeaters in primary schools, there would be 37 repeaters for 100 students, resulting in 1.06 classrooms ‘lost’ to repeaters for each 100 students. For a school with 1000 students, more than 10 classrooms would be wasted. However, if there were 10 percent repeaters, only 0.34 classrooms would be ‘lost’ for each 100 students, resulting in less than 4 classrooms wasted. 60 Figure 57: Repetition Rate in Primary and Lower Secondary Education in Gabon and Peer Countries Equatoriel Guinea 19.2 Central African Republic 23.2 Botswana Gabon 4.4 Ecuador 1.2 Côte d'Ivoire 16.1 D.Republic of Congo 11.3 Chad 27.1 Cameroon 12.8 0.0 5.0 10.0 15.0 20.0 25.0 30.0 Lower secondary Primary Source: UNESCO’s data base from 2011 to 2014 – MENEC estimated repetition rate for Gabon. Economic and Functional Allocation Issues 204. Salaries receive the largest share of Gabon’s education budget. In 2015, salaries represented 51 percent of the education budget for the National Education, Higher Education and Scientific Research, and Vocational Training programs (Figure 58). Non-salary spending represented about 37 percent of the budget. Although there seems to be significant space in Gabon’s education budget for goods and services (including inputs to improve academic performance and reduce repetition), 18 percent of the budget was allocated to transfers, of which more than 85 percent was dedicated to non means-tested scholarships in higher education, raising efficiency and equity concerns. Additionally, only 12 percent of the budget was allocated to capital spending. Salaries represented 49 and 54 percent of the 2016 and 2017 education budgets, 68 respectively, which was broadly in-line with the 2015 budget. However, allocations for permanent staff salaries in the 2015 and 2016 budgets for the National Education program were mainly recorded within the transversal program category (i.e., piloting and support), making it difficult to monitor wage allocations by education cycle. 205. The economic classification of education expenditure is problematic due to the centralized execution of wage expenditure. Salaries in the education sector represented only 15 percent of the committed education spending (depenses ‘ordonnancées’), which is similar to the wage bill of the non-permanent labor force (Figure 58). 69 However, wages of all permanent public- sector officials are centralized and executed within the Public Finance Management sector (or mission), and there is no disaggregation of staff across ministries or average wages by category of personnel. The 2017 Finance Law will make it possible to disaggregate salaries across education cycles to derive a better functional classification of education spending. 206. “Variable” expenditure data confirm inefficiencies in the allocation of public education spending. Transfers accounted for 42 percent of committed education expenditures in 68 According to the initial 2016 and 2017 Finance Laws. 69 RAP 2015, Mission Education Nationale and Loi de règlement 2015 61 2015, excluding wage expenditure for permanent staff (Figure 58). This is in sharp contrast to the 18 percent allocated to transfers in the education budget. Moreover, 16 percent of capital spending was committed in 2015 (compared to 12 percent allocated in the budget), of which one-third went to the rehabilitation of secondary schools. 70 Figure 58: Distribution of Committed Education Expenditures and the Education Budget by Economic Category, 2015 100% 12% 16% 80% 18% Investment 60% 19% 42% Transfer 40% 51% 27% Goods and Services 20% 15% 0% Salary Education budget Committed spending Source: Loi de règlement and revised finance law 2015. 207. Higher education accounts for a disproportionate share of education expenditures. Disregarding non-permanent staff salaries, pre-primary and primary education only accounted for 5 percent of education expenditures in 2015, while 21 percent were devoted to secondary education, and 66 percent were absorbed by higher education (Figure 59). 71 The high share of expenditures dedicated to higher education reflected a focus on scholarships and a significant share of capital investment and goods and services for higher education institutions. In 2017, primary (including pre-primary) and secondary education each absorbed about a quarter of the total education budget, 72 while higher education and scientific research accounted for 31 percent, 73 resulting in the National Education mission receiving about 60 percent of the total education budget. As a result, education expenditures in 2017 were generally in alignment with the 2015 budget, although a higher share of the 2017 budget was allocated to vocational training. Moreover, most of the 2017 expenditures for pre-primary and primary education were for the wage bill of permanent staff. 70 RAP 2015, Mission Education Nationale. 71 Due to the lack of disaggregation of wages by sector (or mission) and program/cycle, the analysis of expenditure distribution by function is limited for the financial year 2015 and 2016. 72 Transversal administrative costs needed to be added. 73 When accounting for salaries of permanent staff across education cycles in the 2017 budget. 62 Figure 59: Distribution of Expenditures (2015) and Education Budget (2015 and 2017) by Function 100% 90% Vocational Training 80% 40% 70% 31% Tertiary education and research 66% 60% 50% Other administration (National Education) 40% 24% Technical and Vocational 30% Education 20% 21% Secondary 10% 25% 8% 0% 5% 5% Preschool and Primary Expenditure Budget 2015 Budget 2017 'ordonnancée' 2015 Source: Loi de règlement 2015, revised finance law 2015, and finance law 2017. 208. Gabon’s high share of tertiary education spending makes it an outlier compared to its neighbors and peer countries. From 2010 to 2014, the country allocated 58 percent of its education budget to primary and secondary education, much lower than the 75 percent average for SSA countries and the average of many upper middle-income SSA countries such as South Africa (70 percent) and Mauritius (80 percent). 74 Conversely, Gabon allocated an average of 38 percent of its education budget to higher education during the same time period, much higher than many of its peers and upper middle-income SSA countries such as Namibia (23 percent) (Figure 60). Figure 60: Share of Tertiary Education Spending compared to Income Level, 2014 45 Expenditure on tertiary education (% of government expenditure on education) 40 Gabon 35 30 25 20 15 10 5 0 0 5000 10000 15000 20000 25000 GDP per capita US $ (2014) Source: WDI, 2014. 209. A large portion of expenditures in pre-primary, primary, and secondary education are investments (Figure 61). 75 However, the level of capital spending in pre-primary and primary 74 UNESCO data base (2010-2014). 75 Education expenditures exclude the wage bill of permanent staff. 63 education in absolute terms is small, as the portion going to preschools and primary schools is small in the overall education budget. In 2015, 57 percent of allocations to pre-primary and primary education were used to rehabilitate public elementary schools, and 29 percent were used for goods and services to cover examinations and tests, leaving little room for spending on teacher training (about 4 percent). 76 Secondary education expenditures followed the same trend: 43 percent of non- permanent salary expenditures were allocated to the construction and rehabilitation of secondary schools in 2015, and 39 percent for goods and services (which included school fees for students that attended private schools). 210. However, most expenditures in technical and vocational education are used to pay non-permanent staff, and most tertiary funding is allocated to transfers (Figure 61). Since 57 percent of allocations were used to pay non-permanent staff in 2015, there is likely a shortage of permanent teachers in TVET institutions, similar to the shortage in general secondary education. Moreover, unit costs in technical and vocational education reached US$1,917 per student in 2015, suggesting serious inefficiencies in the allocation of resources. Finally, transfers represented 61 percent of all expenditures in higher education in 2015. Figure 61: Distribution Educational Expenditures* by Function, 2015 100% 4% 90% 22% 27% Investment 80% 43% 70% 57% 8% 60% 22% 61% 50% 5% 22% Transfer 40% 30% 39% 29% 57% 20% 24% 43% Goods and Services 10% 13% 13% 11% 0% Preschool/Primary Secondary Technical and Tertiary Vocational Training Non permanent salary Vocational Education Source: Loi de règlement 2015. Note: * Education expenditures exclude the wage bill of permanent staff. Analysis of Expenditures: Equity Issues 211. The participation of households in the education system is highly uneven at higher education levels, pointing to regressive public spending in upper secondary and tertiary education. 77 All wealth quintiles were equally represented in primary education in 2012, with a slight underrepresentation of the richest quintile (17 percent of enrollments) and an 76 RAP 2015, National Education mission. 77 Due to data constraints, benefit incidence analysis is largely carried out using a simple utilization approach that assumes all individuals that use the service receive the same benefits. An additional limitation is that the analysis should be focused on public schools, but the DHS 2012 does not provide information about private and public attendance. As described earlier in the chapter, many private schools (in particular in secondary education) benefit from subsidies, seconded teachers, or school fees from the government, assuming they receive similar benefits. However, the analysis likely overestimates how regressive public spending is. In higher education, scholarships can be used for both public and private universities. 64 overrepresentation of the first two quintiles (45 percent of enrollments) (Figure 62). Conversely, only 9 percent of pupils in lower secondary education were from the poorest quintile in 2012, while 28 percent were from the richest. At the upper secondary level, the distribution becomes highly unequal, with 41 percent of students coming from the richest quintile, while only 5 percent from the poorest, suggesting regressive public spending. Moreover, higher education was mainly attended by students from the top three quintiles, with over 55 percent coming from the richest households, highlighting extreme education inequities and high levels of regressive spending at this level. 78 Data from concentration curves tell a similar story: the poorest segment of the population was slightly overrepresented in primary education, but there was high inequality in access at the upper secondary and higher education levels (Figure 63). As a result, Gabon’s public spending on education was only pro-poor at the primary education level. Figure 62: Percentage of Enrollment by Quintile and Education Level, 2012 60 55.4 Percent 50 40.9 Quintile 1 40 30.9 Quintile 2 27.9 30 24.1 17.3 Quintile 3 20 8.8 7.8 Quintile 4 10 4.7 0.5 Quintile 5 0 Primary Lower secondary Upper secondary Secondary (total) Higher education Source: Authors’ calculations based on DHS 2012. Figure 63: Enrollment Concentration Curves by Education Level, 2012 100 Cumulative enrolment, % (total) 90 80 Primary education 70 60 Line of equality 50 40 Lower secondary 30 20 Upper secondary 10 0 Higher education 0 20 40 60 80 100 Cumulative wealth index quintiles, % of population Source: Authors’ s calculations based on DHS 2012. 212. An analysis of average benefits across wealth quintiles suggests that half of all benefits go the richest quintile. In 2015, the government spent about FCFA 122 billion on all education levels, excluding investment, administrative units, and salaries of permanent staff. After comparing 2015 government education expenditures with the 2012 enrollment rate 79 (assuming that enrollment patterns did not change since 2012), the richest quintile received 50 percent of expenditures on education in 2015, which represented around FCFA 60.8 billion (Figure 65). 78 It is possible that some scholarships may have benefited students from the middle quintiles, but the overall effect of the country’s scholarship policy in tertiary education appears to be regressive. 79 2012 DHS. 65 However, the lowest and the second lowest quintiles received only 3 and 7 percent of education expenditures, respectively, representing FCFA 12 billion. The same results can be demonstrated in a Lorenz-curve graph that cumulates benefits across wealth quintiles (Figure 64). Figure 64: Benefit Incidence Analysis of Figure 65: Government Subsidy to Education from Government Education Spending, 2015 Current Expenditure by Quintile, 2015 Line of equality 70.0 1 60.8 0.9 60.0 0.8 0.7 50.0 Billion 2015 CFA 0.6 40.0 0.5 27.1 30.0 22.5 0.4 0.3 20.0 0.2 9.0 10.0 3.1 0.1 0 0.0 0 Lowest Second Third Fouth Richest Lowest Second Third Fouth Richest quintile quintile quintile quintile quintile quintile quintile quintile Source: Authors’ calculations based on DHS 2012 and 2015 current expenditure (excluding permanent salary, administrative units, and investment). 213. Gender parity is almost achieved in primary school, while girls are more represented at higher levels of education, except for girls from the poorest households. In 2012, 51 percent of primary students were girls, whereas they were 55 and 56 percent of all secondary and higher education students, respectively. Except for the poorest segment of the population, girls were overrepresented throughout the education system (Figure 66). However, girls from the poorest quintile were less represented in secondary education and not present at all in tertiary education. Figure 66: Gender Equality in Education by Wealth Quintile, 2012 20 Percentage points (difference 10 0 Lowest quintile from 50%) -10 Second quintile -20 Third quintile -30 Fouth quintile -40 Richest Above zero : female advantage -50 Primary Secondary Higher Source: Authors’ calculations based on DHS 2012. 214. Boys and girls benefit almost equally from government education spending, except for in higher education. Based on their participation levels, 80 girls from the richest quintile received about FCFA 36.6 billion of government spending on education in 2015, while boys only received FCFA 24.2 billion (Figure 67). For other quintiles, both genders received roughly equal 80 Spending disaggregation by gender is not available. 66 benefits. Overall, girls benefited on average from 56 percent of recurrent government education expenditures. Figure 67: Government Subsidy to Education from Current Expenditure by Gender and Wealth Quintile, 2015 In billion 2015 CFA Total subsidy to girls Total subsidy to boys 36.6 24.2 13.9 13.2 11.6 11.0 5.3 3.7 1.2 1.9 Lowest quintile Second quintile Third quintile Fourth quintile Richest Source: Authors’ calculations based on DHS 2012 and 2015 current expenditure (excluding permanent salary, administrative units, and investment). Analysis of Expenditures: Cross-cutting Issues Public Financial Management Issues 215. Budget allocations do not always reflect program objectives, pointing to inefficiencies in planning and budgeting. For example, even though the aim of the 2015 PAP was to increase access to education, the government focused on allocating funding to improving the quality of education. Moreover, the 2016 PAP’s objective was also increase access to education, but government actions focused on resource management. 216. Execution rates vary across education levels. Public expenditures in secondary education, higher education, and vocational training all exceeded their initial budget allocations in 2015 (Table 10). In secondary education, excess expenditures were the results of budgetary extensions to finance goods and services for examinations and competitions, school fees allocated to students, and school aid. 81 Initial results are not very different for 2016. Table 10: Execution Rates by Education Level and Economic Category, 2015 National Education Higher Expenditure Technical Vocational Primary Secondary Pilotage/Support education Prof. Personnel 0.2% Salary of non- 94% 99% 445% 75% 100% permanent staff Good and services 71% 157% 86% 71% 72% 87% Transfer 165% 90% 139% 116% Capital 62% 208% 34% 50% 32% Source: Loi de règlement 2015. 81 RAP 2015, Mission Education Nationale. 67 217. Several factors must be considered when interpreting spending execution rates. The discrepancy between budget allocation and actual expenditures can measure the planning capacity of line ministries and the overall quality of governance at central government ministries and departments. However, there are several bottlenecks in the budget that may prevent departments from using budgeted funds. Even though these blockages may not necessarily be due to a lack of capacity to execute budgets, they bring attention to the overall management of the budget, particularly cash flow. Moreover, budget-execution rules imposed by the Ministry of Finance may prevent the full implementation of departmental programs, especially during the compulsory reserve (a 15 percent freeze in allocations) or early closure of the commitment period. At the policy level, discrepancies start with the US$3 billion “2010 Etats Généraux” strategy that defines the development of Gabon’s education system between 2010 and 2020, but which was not implemented due to financial and capacity constraints. 218. The compulsory reserve and procurement delays have constrained the execution of capital expenditure in primary education. The National Agency for Major Investment Projects executed the sector’s investment program between 2009 and 2015, with MENEC as the authorizing officer. The slow pace of public procurements resulted in the execution of only 40 percent of the contracts during the 2015 financial year. Investment documents are prepared by the Department of Patrimony and the Department of Investment Planning which are also the contracting authorities. However, the departments of DAPs and the departments of zone academies are not involved in the process, and allocations do not include investment credits for the entity in charge of BOP and the OU. 82 219. Insufficient funding for goods and services generates huge amount of arrears at the MTEFTPIJ. The ministry suffers from recurrent budget shortages and lacks the resources to properly manage examinations and tests. For instance, test and examination supervisors have not been paid for years. Moreover, three vocational training centers in Libreville and the National Training Agency did not receive any budget lines for goods and services in 2016, resulting in strikes by trainees due to a lack of training materials. As a result, capital expenditure is mainly financed by external aid and incorporated into the PAPs. 220. Budget cuts are also common at the MESRSFC. In 2016, the ministry received 5 percent less than what was allocated for goods and services, 10 percent less for transfers, 15 percent less for investments, and appropriations were only allowed to be 50 percent of the budget. On the other hand, higher education institutions received full allocations. Human Resource Management Issues 221. Despite specific laws covering civil servants in the education sector, 83 Gabon does not have a coherent system to track the number of teachers in the country. In 2016, the number 82 According to interviewed departments, some investments are decided at ministry level and implemented directly by the minister’s cabinet and the Department of Budget and Public Finance. 83 Law No. 20/92 on March 8, 1993 deals with the status of civil servants that work in the education sector. Decree No.000904 on June 6, 1996 fixes the general system of remuneration of personnel performing the services of examinations and tests. 68 of school teachers was estimated at 8,355 for primary education and about 3,076 84 for secondary (general and technical) education. The Finance Law sets maximum ceilings on the number of staff allowed to be employed in each ministry, including teachers in public institutions, seconded teachers in private RUP schools, and administrative staff at the local and central government level, which is counted against the wage bill for education. The 2017 Finance Law set the limit at about 24,000 staff, including teachers for MENEC. 85 However, line ministries do not know the exact number of public or private RUP teachers they are paying for, hampering the analysis of teacher allocation, distribution, and qualification. The central government’s Department of Human Resources is temporary responsible for the management of teachers, as the country-wide Human Resource Management System 86 is not yet operational. In higher education, teacher recruitment is managed directly by the autonomous education institution after receiving approval from the Department of Human Resources. 222. Qualified teachers are recruited after they successfully complete the relevant training programs and receive their certificates, although unqualified teachers are also recruited. Primary school teachers with Baccalaureat degrees are recruited once they have attended ENI (Ecole Normale des Instituteurs), which trains both public and private teachers. About 1,000 teachers graduate from ENI each year, of which about 35 percent teach in the public sector. However, local collectivities also recruit primary school teachers without any minimum qualifications or guidance from MENEC. Collectivity teachers represent about 9 percent of the teacher workforce in primary education. Secondary general education teachers need to graduate from ENS (Ecole normale supérieure), and secondary technical education teachers need to graduate from ENSET (Ecole normale supérieure de l’enseignement technique). Teachers are hired as either deputy professors (category A2 in the public servant ranking) or professors (category A1 in the public servant ranking), depending on the diploma they pursue: Bachelors or CAPES (Certificat d’aptitude au professorat de l’enseignement secondaire). About 1,000 teachers, of which about 48 percent will work in the public sector, graduate from ENS or ENSET each year. University graduates can also be recruited as secondary teachers without any initial training. After recruitment, civil servants (including teachers) are assigned to a provincial DAP by the Department of Human Resources before being assigned to a school by the DAP. The transfer of teachers and civil servants under MENEC is based on the minister’s decision. Due to budget constraints, there has been no in-service teacher training since 2013 for secondary education. 223. The supervision of primary schools is well established, although it is less clear for secondary schools. Inspectors for primary education are the head of the school districts and are supported by pedagogy advisors. Each inspector must visit at least 21 schools per year, and reports on the visits are prepared by the pedagogy advisors and submitted directly to both the DAP and the General Inspection of Services. Primary schools also have school supervisors and administrative staff. Inspectors in charge of secondary education are mostly former pedagogy advisors. Their number has decreased over the time, and they are mostly located in the capital, greatly constraining their effectiveness. 84 Data derived with the support of donors to justify the additional recruitment of secondary teachers. 85 The maximum number of staff allowed for education in the 2017 Finance Law is 24,173 for MENEC, 3,059 for MESRSFC, and 1,503 for MTEFTPIJ. 86 An application called XGEST is being installed in all DAP to manage the school map and human resource management. 69 224. There is a deficit of mathematics teachers in secondary education, and their distribution varies greatly across provinces. The average pupil-teacher ratio in secondary education is over 60:1, suggesting an overall shortage of teachers across public and private schools. Since there is a limited number of graduates each year from ENS in many core subjects such as mathematics, science, philosophy, and languages, Gabon imports many secondary education teachers. In 2016, foreign teachers at Gabon’s secondary schools were about 1,500, 87 or about half of all teachers. However, the shortage of teachers is not evenly distributed across subjects and provinces. 88 Despite a large number of foreign teachers in mathematics, there is a high deficit of mathematics teachers (-191) compared to other subjects such as biology (0). Beyond an overall deficit, mathematics teachers are also unevenly distributed around the country, with a relatively high deficit in the Estuaire province (including Libreville) compared to other provinces (Figure 68). Figure 68: Deficit of Mathematics Teachers in Secondary Education by Provinces, 2013 0 Ogooué-Lolo Moyen-Ogooué Nyanga Woleu-Ntem Estuaire Maritime Ngounié Ogooué-Ivindo Haut-Ogooué Ogooué -10 -20 -30 -40 -50 -60 -70 -80 Source: Authors’ calculation based on dynamic school map, 2013. 225. Teacher deficits in the public sector are compounded by high attrition rates and recurrent strikes. Gabon’s attrition rate for teachers in public schools is estimated to be high,89 as many teachers quit for other jobs, including in private schools. This is a particular challenge as many public schools already face a shortage of teachers, especially in secondary education. In addition, there are recurrent strikes among public-sector teachers at all education levels, which has contributed to schools often being open only 12 weeks each school year (instead of the standard 35 weeks) over the past 15 years. Infrastructure Utilization Issues 226. Gabon is suffering from a shortage and misallocation of classrooms. Almost 40 percent of primary schools in the Estuaire province had pupil-to-classroom ratios above 45:1 in 2013, while 25 percent had 55:1 or above, and 12 percent had between 46:1 and 55:1 (Figure 69). Among schools with ratios above 55:1, 38 percent were in the Estuaire province. Moreover, 25 out of the country’s 858 secondary schools had pupil-to-classroom ratios above 70:1, and some schools in the Estuaire province had ratios as high as 176:1. Based on a target of a pupil-to-classroom ratio of 35:1, about 700 and 1,295 additional classrooms are needed in primary and secondary education, 87 Most of foreign teachers are from West Africa (63 percent), other African countries (5 percent), and Europe (2 percent). They often teach mathematics, physics, chemistry, and languages. 88 According to the 2013 school map showing the needs for teachers. 89 This is based on MENEC’s estimates, as official data on teacher attrition rates does not exist. 70 respectively. The Estuaire province represents 59 percent of the classroom deficits, although shortages are also a challenge in the Ogooue Maritime and Moyen-Ogooue provinces. 227. Government efforts to reduce the classroom deficit are hampered by procurement delays. MENEC planned to start constructing new schools and classrooms in 2015 to reduce the pupil-to-classroom ratio to 35:1 by 2017, although execution was delayed because of procurement issues. The ministry’s plan only called for new classrooms in the Estuaire province, which would benefit from three new schools among the eleven already planned, suggesting a concerted effort by the government to target the province with the highest classroom deficits. 90 Figure 69: Distribution of Student-to-classroom Ratios in Public Primary Schools by Provinces, 2013 100 90 >55 80 70 46-55 60 Percent 36-45 50 40 26-35 30 <25 20 10 0 Source: Authors’ calculation based on dynamic school map, 2013. Recommendations Key policy options to improve the adequacy, efficiency, and equity of public education spending in Gabon: 228. Increase public education spending. Gabon spends less on education than many comparable upper middle-income countries with similar income levels. If well managed, an increase in public spending allocated to education could have a positive impact on the country’s education outcomes. The government should consider creating fiscal space to increase spending on education by increasing the mobilization of non-oil revenue, improving the overall efficiency of spending, and agreeing on minimum fiscal targets. 91 Moreover the governments needs to assess the level of private household spending in the education sector to better leverage private-sector instruments. For instance, a clear picture of private spending in education could help authorities evaluate the economic constraints an increase in school fees may represent. 229. Increase the share of public spending for primary and secondary education. At present, higher and upper secondary education receive most of the public financing for education at the expense of primary and lower secondary education and TVET. As a result, the budget constraints facing the primary education system have led to a shortage of both teachers and classrooms and a lower teaching quality. Considering the high private returns for higher and upper 90 RAP 2015, Mission Education Nationale. 91 This is currently being implemented through the IMF program. 71 secondary education, the government could increase private sector involvement at this level and reallocate scarce public resources to basic primary and lower secondary education. These funds could then be used to make quality improvements and improve access to lower secondary education in under-served areas, which could raise transition rates. Moreover, the government should consider allocating more public resources to upper secondary education and TVET in the Estuaire province and difficult-to-reach areas where there are supply deficits and/or a shortage of private schools. 230. Reform the bursary scheme to make the system more efficient and more equitable. The high proportion of spending to higher education due to non-means-tested scholarships raises a concern of allocative efficiency and equity. The scholarship program should better target youth from poor households. Gabon should consider introducing means-tested scholarships in upper secondary and higher education, particularly for girls 92, while building better linkages between safety-net programs and school attendance. The back-to-school allowance program could also be expanded for all levels of education. Affordable student loans should be introduced at tertiary education level to expand participation in tertiary education. 231. Reduce repetition and improve overall internal efficiency. The school repetition rate is not an indicator of quality, rather it correlates with a lack of quality. 93 With its current high percentage of repeaters, Gabon is wasting resources with no gain in learning outcomes. It is therefore important that the country reduces its repetition and dropout rates to improve the internal efficiency of the entire education system. The government should aim to decrease the repetition rate to at least 10 percent, and it can do this through: 1) increasing learning time, 2) improving the teaching and learning process, and/or 3) strengthening school governance. Expanding the school year to the standard 35 weeks can give students more time to learn the required skills, although this would require a reallocation of resources to build more classrooms and train additional teachers in basic education. Moreover, deploying more adequate teachers could help identify students at risk of dropping out and provide an opportunity to institute proper remedies. Better teacher training could also increase the quality of basic education and improve the coordination of curricula between education cycles, improving the internal efficiency in upper secondary and higher education. Finally, authorities can strengthen governance by involving parents and communities in the monitoring of education outcomes and making schools accountable for results. Many SSA and Latin American and Caribbean countries have successfully implemented school- based management systems to increase the accountability of teachers, reducing teacher absenteeism. Additional efforts to reduce repetition rates involve understanding the causes of the low transition between lower and upper secondary education and identify policies to address them. 232. Invest more in teacher training. The government needs to invest in the right training programs and approaches to equip teachers with the adequate subject-level and pedagogical skills and ensure they are applied in the classroom. This would allow teachers to deliver quality teaching and provide more skills to students. A tutoring program could be particularly useful for newly appointed teachers that not trained in the conventional teachers’ training institutes. The coaching/tutoring and supervision chain should include the school principals, inspectors, and education officials from the regional and central level. Moreover, the government should abandon 92 The scholarship program in tertiary education needs to be carefully reviewed. 93 See Cf. RESEN 2012 for a detailed analysis on the correlation between repetition and quality.. 72 the current administrative supervision system and provide more technical assistance to teachers. Authorities should also focus on improving the teaching of mathematics and science, ensuring more graduates from secondary schools have a strong foundation in these subjects. This would allow students to pursue post-secondary studies in science, technology, engineering, and mathematics and gain employment in technologically advanced fields, or become mathematics or science teachers in secondary schools. 233. Improve the governance of public TVET and higher education institutions. While rates of return on TVET and higher education are higher than for other education levels, returns and employability could be higher if the quality was higher. To increase quality would require an increased role of the private sector, including in: (i) the identification of training needs; (ii) the design of the curricula and training approaches, including identifying adequate teaching and learning materials, equipment, and assessment packages; (iii) the development of internship programs, including dual-training options; and (iv) the evaluation and assessment of programs. The government should also explore providing more financial and decisional autonomy to TVET institutions to allow them to provide skills that are more relevant in the labor market (and improve their own efficiency). This could be done through performance-based contracting and/or financing approaches. 94 234. Leverage the private sector in the financing and/or provision of secondary, TVET, and higher education while strengthening the quality assurance framework. The government must revisit its current partnership with the private education providers to make sure it effectively helps in expanding access and in leveraging quality. Private education providers (with or without formal PPPs) should be encouraged to increase their investments in upper secondary education, TVET, and higher education, as it would increase access to higher education levels and free up public resources (which could be reoriented to more equity-enhancing measures). Examples of incentives for the private sector in the provision of secondary, TVET and higher education could be (i) provision of land, (ii) provision of one additional classroom to one classroom built where the demand is higher than the supply, (iii) provision of guarantee fund to allow the private sector borrow money from commercial bank, (iv) provision of unused public infrastructure, (v) education voucher linked with a conditional cash transfer. The government may also impose a certain number of youth from poor households without paying school fees to schools benefiting from the government support. However, this would need to take place within a strengthened quality assurance framework. Moreover, authorities would need to assess the quality of existing private primary schools and PPPs in secondary education. The government would also need to investigate why secondary education enrollment is unevenly distributed among different income quintiles despite public support to many private schools. 235. Explore innovative delivery mechanisms in secondary education and TVET. The relative spatial concentration and adequate telecommunications and Internet coverage in Gabon could make it a good candidate to experiment with distance and semi-distance secondary education in rural areas. In TVET and skills development, the country is already experimenting with dual apprenticeships and other short-term skills-development programs aimed largely at out-of-school 94 Performance-based financing could include transferring resources to institutions on the basis of indicators that measure the relevance of programs and results, such as enrollment, girls’ enrollment, examination results, number of graduates absorbed by the labor market, and number of publications in relevant journals. 73 vulnerable youth. If well implemented, these programs have the potential to improve the access to technical skills by poor households. To address the immediate lack of classroom, double shift mechanisms could be considered. 236. Enhance overall PFM and improve the planning, budgeting, execution, and reporting of education spending. The government needs to fully disaggregate executed spending data across sectors, including the wage bill, to allow for meaningful spending analyses. 95 Within each social sector, authorities need to be able to disaggregate budgets and actual spending, including salaries, across programs. This implies being able to monitor spending data for the wage bill by education level/cycle. 96 Moreover, PAPs and PARs should be detailed at the level of OUs, and organigrams should be aligned with programs. Key stakeholders should also be trained on the preparation and reporting of program budgets to avoid discrepancies and mismatch between program objectives and actual spending. 97 Finally, the government needs to speed up the execution of public funds through urgently dealing with procurement delays and releasing funds for goods and services (notably in TVET). Authorities also need to clarity and improve the allocation criteria for teachers and classrooms. 237. Modernize human resource management and attach staff to programs. In Gabon’s education system, teachers are unevenly distributed across provinces, and there is a serious deficit of mathematics teachers in secondary education. As a result, the government should attach staff to specific programs, which would help teachers focus on the right activities, improving the overall accounting and deployment of teachers. Creating predictability in the deployment of teachers and resources would allow authorities to assess skills needs in order to develop training plans, redeploy resources, optimize career follow-ups, and anticipate employment changes. 238. Finally, Gabon should develop a reliable education management information system (EMIS) and improve the availability of baseline diagnostics. The lack of statistical data is an important constraint to the efficient management of the country’s education system. Without a reliable EMIS, adequate tools to measures and monitor learning and labor market outcomes, and efficient mechanisms to share information with schools, education authorities are unable to monitor progress and take informed decisions on policy reforms. A well-developed EMIS will help the government improve budget allocations and produce quality PAPs with better data on performance and actual needs. 98 The development of a reliable EMIS and assessment system should include: i) strengthening the capacity of the departments in charge of planning and assessment (including at the provincial level); ii) reviewing data-collection formats and mechanisms; iii) developing assessment tools; iv) acquiring relevant applications for data analysis; v) reviewing dissemination mechanisms (including an updated website for MENEC); and vi) defining the roles and responsibilities of each entity in the system. Finally, Gabon should consider undertaking a service-delivery assessment, such as the System Approach for a Better Education 95 The government started doing this in its 2017 Finance Law. 96 This is being done in the 2017 Finance Law and the 2017 PAPs and PARs. 97 This will be important in the context of the 2017 program budgets, as they will be presented with a multi-year perspective over three years. 98 MENEC does have records of enrollment at secular private education providers in addition to its overall teacher force. 74 Results – Service Delivery Indicator survey, to create an up-to-date baseline on quality and service delivery in the country. 75 5. SOCIAL PROTECTION Introduction 239. This chapter examines public expenditures for social protection mechanisms in Gabon in light of ongoing efforts to improve social protection programs for the country’s most vulnerable population. The vast majority of these programs are safety nets targeting the “economically weak Gabonese” (Gabonais économiquement faible, GEF) 99 and are designed to support consumption, increase access to basic social services, and promote equitable and inclusive economic growth. 100 There are five main types of social protection mechanisms in Gabon: cash transfers, in-kind transfers, labor-intensive works or cash-for-work, fee exemptions and subsidies, and special assistance to specific vulnerable groups (Table 11). First, this chapter presents an overview of the country’s social protection sector, with a focus on its evolving safety net system, followed by a review of expenditures across programs to identify the structure, sources, and trends of public spending in the sector. These public expenditures are then analyzed to evaluate if they meet the needs of the people and are in-line with the government’s strategy. The efficiency of public spending is also evaluated along with its effects on equity. The chapter concludes with recommendations for how to direct social protection expenditures to increase poverty reduction and accelerate inclusive growth. Table 11: Social Protection Mechanisms in Gabon Type Description Cash Transfers These transfers provide beneficiaries with the financial resources necessary to maintain a minimum level of consumption, and they can be coupons or cash and conditional or unconditional. In-kind Transfers These transfers distribute food or food supplements (e.g., in school feeding programs) or non-food items, such as school kits, household items, seeds, or work tools. Labor-intensive Works / Beneficiaries receive cash or in-kind transfers in exchange for participating in Cash-for-Work public works programs, which are sometimes combined with training sessions. Public works create common goods that can improve community living standards. Fee Exemptions and Beneficiaries are exempt from school or health fees to facilitate access to Subsidies essential services for those who are unable to pay. They can also receive subsidies to pay for certain items or services, such as water, electricity, or food. Special Assistance to This support targets extremely vulnerable groups, such as street children, covers Specific Vulnerable Groups a wide range of needs, and is often provided as a comprehensive package (e.g., psychosocial support, housing support, and provision of health and education services). 99 A GEF is defined by national legislation as “a physical person who: (i) is of Gabonese nationality, (ii) is at least 16 years of age, and (iii) earns less than FCFA 80,000 per month).” 100 The terminology commonly used by the government distinguishes between two main groups of non-contributory programs: social benefits transfers (prestations sociales), which are destined to all poor families, and safety nets, which is a new name given to the old “social aids” (aides sociaux) and target specific vulnerable groups such as persons with disabilities and street children. In this chapter, the term “social safety net” is used to cover all social transfers in accordance with the internationally accepted definition. 76 240. The analysis in this chapter focuses on the 2014-16 period, but the insufficient availability of data makes over-time comparisons difficult. The PAP/RAP 101 documents, which provide detailed programming by objectives and expected results, are the main sources of information for the analysis of social protection expenditures. Although the Finance Laws are available for several years and contain aggregated data, PAP documents are only available for 2015 102 and 2016, making it impossible to perform detailed trend analyses of social protection expenditures and programs. In addition, year-on-year comparisons are difficult to make from the information provided by the main implementing institutions because the roles and responsibilities of the different entities as well as the structure and nature of some programs were redefined after 2014. Also, the transition to budgeting by program objectives in 2015 further complicates the comparison of data. As a result, a substantial data collection effort was carried out with various institutions to gather additional information on allocated and spent amounts, average subsidies, beneficiaries, and other financial and non-financial data by program. Complementary data were also available in Gabon’s Human Investment Strategy (Strategie d’Investissement Human du Gabon, SIHG). 241. Only information from publicly funded interventions was used to analyze social protection spending in Gabon, which is different from most methods used in other African countries. In most Sub-Saharan African (SSA) countries, especially those with lower levels of economic development, over 70 percent of resources for social safety nets come from development partners and humanitarian organizations. However, Gabon has a well-developed social assistance system that is financed entirely by the government. Since development partners offer mainly technical support, and the role of non-governmental organizations (NGOs) in the country’s social protection system remains marginal, they are not included in the analysis. 101 A PAP is a programmatic government plan for Results and Action Planning applied to public institutions. 102 The 2015 PAP also contains some incomplete data for 2014. 77 Overview of the Social Protection Sector Safety-net System Box 4: Restructuring the Social Protection System in Gabon Health Insurance 2007: The National Health Insurance and Social Guarantee Fund (CNAMGS) was created with a mandate to implement compulsory health insurance. 2008: The health insurance fund for the GEF was created. The fund was managed by CNAMGS. 2011: The health insurance fund for the public sector was created. The fund was managed by CNAMGS. 2013: The health insurance fund for the formal private sector was created. The fund was managed by CNAMGS. Social Safety Nets 2012: The National Social Assistance Fund (FNAS) was created with a mandate to support vulnerable groups through “social aids.” 2014: A national strategy for social safety nets and economic self-sufficiency (SIHG) was developed. 2015: The implementation of the national strategy started through CNAMGS and FNAS. Contributory Social Benefits 2014: The Pension and Family Benefits Fund (CPPF) was created to manage pension and social benefits for public employees. Social Protection 2016: The National Social Protection Policy (PNPS), which defines the country’s national framework for social protection, was validated. 2017: The new social protection code (Law no. 28/2016) was adopted. Source: Information provided by the Government of Gabon. 242. Gabon’s social protection system has steadily evolved over the past ten years. The government started an ambitious process to restructure the national social protection system in 2007 (Box 4), resulting in a set of reforms aimed at achieving universal health insurance and improving the country’s social safety nets. A concern for the country’s most deprived households has guided the transformation, leading to the identification and operationalization of various non- contributory schemes, including the creation of the country’s first health insurance fund aimed at GEF. However, the economic difficulties of recent years slowed the restructuring process, and a significant number of reforms – related to system coverage, funding, and targeting – have yet to be fully implemented. 243. Social protection programs are managed primarily by four institutions. These are the CNAMGS, the National Social Assistance Fund (Fonds National d'Aide Sociale, FNAS), the National Social Security Fund (Caisse Nationale de Sécurité Sociale, CNSS), and the Pension and Family Benefits Fund (Caisse des Pensions et des Prestations Familiales, CPPF). Benefits and services differ according to the professional status of beneficiaries (Table 12). • Public-sector workers. CNAMGS provides social protection services for public-sector workers and is responsible for compulsory health insurance, and the CPPF is responsible for managing the following risks: (a) work accidents and occupational diseases; and (b) old 78 age, disability, and death. 103 The financing of social insurance benefits for public-sector workers is guaranteed by employee contributions (2.5 percent for active staff and 1 percent for pensioners) and employer contributions (5 percent). 104 • Formal private-sector workers. Social protection benefits for workers in the formal private sector are managed by CNAMGS, which is in charge of the compulsory health insurance fund for the private sector, and the CNSS, which is responsible for managing the following risks: (a) work accidents and occupational diseases; and (b) old age, disability, and death. Benefits are financed by employee contributions (2 percent for active workers and 1 percent for retirees) and employer contributions (4.1 percent). • Informal sector workers. There is no public mechanism for managing the social risks of informal sector workers who account for more than half of the workforce (57 percent of jobs are in the informal economy). Individuals with incomes below the minimum wage (FCFA 80,000 per month) are eligible to participate in non-contributory GEF programs, but a specific health or social insurance scheme has not yet been established for other households. The recently approved social protection code has provisions for the creation of a new health insurance fund for the non-poor informal sector, and actuarial studies are envisaged to this end. • GEF. Programs available to GEF are all non-contributory and managed by CNAMGS, and FNAS is responsible for economic self-sufficiency programs that support income- generating activities (IGA) for GEF cooperatives or associations. Table 12: Entities Responsible for Social Protection Programs by Household Type Contributory Social Non-contributory Household Type Health Insurance Transfers Social Transfers Public-sector workers CNAMGS (contributory) CPPF (not eligible) Formal private-sector CNAMGS (contributory) CNSS (not eligible) workers Informal private-sector no system in place no system in place (not eligible) workers CNAMGS (non- GEF -- CNAMGS and FNAS contributory) 244. The national strategy to assist deprived households is defined in the SIHG. The strategy outlines the main social protection programs for the GEF and contains a detailed action 103 The CPPF was established in 2014 but only became operational in 2016. Previously, the General Directorate of the Budget (for family benefits) and the General Directorate of Public Treasury Services (for pensions) were responsible for providing social benefits to civil servants. 104 The following categories are exempt from personal contributions: veterans, public servants receiving an invalidity pension, public servants and retirees with incomes below the minimum wage, and retirees residing abroad. 79 plan for their implementation, including an estimated yearly budget totaling FCAF 76-103 billion. The strategy is based on four mechanisms: 105 • Safety nets for GEF households (FCAF 44-55 billion); • A “graduation” program to help deprived households become self-reliant with the help of IGAs (FCAF 2-3 billion); • Improved access to social services (e.g., in the education, health, water, and electricity sectors) through targeted sectoral investments in the most disadvantaged areas (FCAF 20- 25 billion); and • Opening up rural and remote areas and integrating poor peri-urban areas (FCAF 10-20 billion). 245. The SIHG includes seven safety-net programs with a total of 19 different instruments. Each of the seven programs targets a specific vulnerable group: P1) GEF households with dependent children; P2) poor single women with dependent children; 106 P3) the deprived elderly; P4) deprived widows; 107 P5) deprived disabled persons; P6) deprived young people aged 18 to 25; and P7) street children. Target groups are defined using two criteria: household income, which must be below a given threshold, and the type of vulnerability. Both criteria must be met. For each of the seven vulnerable groups, the strategy proposes a tailored “package” of interventions belonging to four categories: cash transfers, subsidies, free goods and services, and solidarity income/cash-for-work (Table 12). 246. The SIHG defines poor households with children as the main beneficiaries and cash transfers (conditional and unconditional) as the primary social protection instruments (Figure 70). These households (two-parent or single-parent) are expected to account for about 54 percent of all beneficiaries and represent about 55 percent of the government’s annual projected safety-net spending. 108 The main objective of the strategy is to support poor households, including traditional vulnerable groups that need specialized assistance (e.g., the disabled, the elderly, and orphans), by stabilizing their consumption and investing in their human capital - especially children’s. As a result, cash transfers account for 45 percent of the total budget allocated to safety nets, while subsidies and free goods and services account for 23 percent each, and cash-for-work programs account for only 9 percent. Free goods and services (including in-kind transfers) are envisaged mainly for particularly vulnerable groups such as street children. 105 Only actions related to the first two mechanisms (i.e., social-safety nets and graduation programs) are analyzed in this PER, as the other two are multi-sectoral investment activities that go beyond social protection measures. 106 Compared to P1, P2 offers some extra services (i.e., childcare, health subsidies, and vocational training) that are particularly aimed at single women with children. 107 The SIHG only identifies widows (women), but this has been interpreted to mean any surviving spouse living alone, male or female. 108 This includes the beneficiaries of P1 “(families with children”) and P2 (single mothers). 80 Figure 70: Safety-net Beneficiaries and Budget Allocations, SIHG Transfert 2% monétaires 3% Famille et l’enfance 7% 9% Mères célibataires Subventions 12% Personnes âgées 48% 23% 45% Veuves Gratuité de bien Handicapés 22% et services Jeunes 6% EDR 23% Revenus solidaires/Argent Contre Travail Source: Estimates based on 2014 SIHG – Technical Report. 247. The government validated the National Social Protection Policy (Politique Nationale de Protection Sociale, PNPS) in 2016, and a new social protection code was adopted in 2017. The objective of the PNPS is to reduce social inequality, poverty, and social exclusion through: (a) improved access to social services by extending and strengthening social security coverage (e.g., create unemployment insurance and extend health insurance to the informal sector); (b) targeted actions in favor of vulnerable groups (e.g., support consumption for the GEF and develop IGAs); and (c) a stronger and consolidated legislative and regulatory framework (e.g., reform the CNSS and adopt a new social protection code). Outcomes 248. Gabon had an opportunity to make substantial progress in poverty reduction during the years of high oil prices, but available data point to a modest, if any, decline in monetary poverty since 2005. 109 The volatility of the country’s economic growth rate as well as the lack of economic diversification and strong social policies (e.g., in health, education, and social protection) have likely hampered the government’s ability to create better lives for the poor. In the absence of recent household consumption and income data, different estimates suggest only minor improvements in the poverty rate over the last ten years. 110 The more optimistic estimates place monetary poverty at around 22-25 percent in 2015, whereas others place it at 32 percent, close to the 2005 level. A qualitative study also reveals that people’s perceptions tend to confirm these estimates; if anything, people feel that the situation in the country is worsening. 249. Structural poverty indicators point to a growing urban-rural divide. The 2012 DHS (EDSG-II) allows for a number of comparisons with the 2005 data in terms of households’ access to basic services and their general living conditions. For instance, there was a clear improvement 109 Gabon has a notable lack of data to track poverty despite its middle-income status (it is one of three countries in SSA (along with Liberia and Sudan) to only have one household survey with income or expenditure data between 1991 and 2012). Thus, it is impossible to create an accurate picture of the evolution of the income distribution since the last income and expenditure household survey in 2005. 110 See: (a) Wouters, Olivier, 2017. “Gabon vulnerability note: Asset indices as proxies for living standards,” unpublished; (b) World Bank, POV team, 2017. “Poverty Projections in Gabon 2005-2015,” unpublished; (c) Ngoua Nguema, Ignace, 2017. “Enquête qualitative sur la pauvreté au Gabon,” unpublished. 81 in the access to improved water and sanitation between 2005 and 2012, although it was more significant for quintiles 2 through 4 than for the poorest quintile. As a result, six in ten of the poorest Gabonese households did not have improved access to water in 2012. Also, the Multi- Dimensional Poverty Index (MPI) estimates that 11 percent of the urban population and 47 percent of the rural population were classified as being multi-dimensionally poor in 2012, with rural poverty being more severe (MPI score was 0.209) than urban (0.044). The Ogooué-Ivindo, Ogooué-lolo, Nyanga, and Ngounié provinces (as well as Woleu-N’tem in some categories) consistently demonstrated the worst outcomes and spatial disparities. 250. Disparities appear to be particularly striking for Gabon’s most vulnerable population. Orphans and abandoned children; widows, especially women who lived in traditional customary marriages or unions; people living with disability; and indigenous groups (also known as Pygmies) are small in numbers but at a high risk of poverty and violence. Indigenous people are especially vulnerable, as they are not officially counted in the national census or by other data- collection efforts, are under-identified in the civil registry since many do not have identifications, live in extreme poverty, and have dismal human development outcomes. Moreover, their small and scattered presence in the country makes them difficult to target by social policies. Overview of Social Protection Expenditures: Structure, Sources, and Trends 251. The social protection system is financed by contributions from workers, contributions from private employers, the national budget, and earmarked taxes. Pensions, health insurance contributions paid to public workers, the overall administration of the national social protection system, and most non-contributory instruments are funded by the national budget. 111 Earmarked taxes finance health insurance for the GEF and came from two sources until 2016: a 10 percent levy on the turnover of mobile-phone companies (i.e., the obligatory fee for health insurance (redevance obligatoire à l'assurance maladie, ROAM)) and a 1.5 percent levy on the transfer of money abroad (i.e., TTS). The ROAM was expected to provide the government with about FCAF 20 billion per year, but only two of the country’s four mobile-phone companies paid regularly, resulting in significantly lower revenues (about FCAF 13-15 billion per year). In addition, there was often a delay in transferring resources from the Treasury to CNAMGS, which tended to hinder operations on the ground. As of March 2017, the ROAM has been replaced by the Special Solidarity Contribution, a new tax equivalent to 1 percent of the VAT that is expected to provide CNAMGS with in an estimated FCAF 25 billion per year. 252. Expenditures for non-contributory social protection programs, other than health insurance and subsidies, are considered a part of social security (mission 22). In Gabon, social security activities are organized around three major programs: 111 Family allowances for civil servants, which are paid by the government, are not considered part of social protection but a part of overall salary expenditures. Family allowances for public-sector employees include: (a) child allowances: FCFA 8000/month per child; (b) new baby allowances: FCFA 60,000 per baby; (b) pre-natal allowances: FCFA 80,000 per pregnancy; (c) education allowances: FCFA 62,500/year per child 3-21 years old; and (d) unique salary allowances: FCFA 2,200/month per active worker or FCFA 1,500/month per retired worker when there is only one salary from the formal sector in the household. 82 • The Family Protection and Promotion program aims to promote women’s rights, protect surviving spouses and orphans, assist the elderly, and protect families and children. In 2016, this program mainly supported the development of an effective safety-net system (e.g., the implementation of an identification system for GEF) and no longer included direct transfers to vulnerable individuals (this was the case until 2014 and would appear to be the case again in 2017). 112 • The National Solidarity program focuses on social exclusion by addressing social problems, such as statelessness and young people “in situations of social distress and exclusion,” improving the living conditions of persons living with disabilities, and supporting civil society organizations. Activities financed under this program include social assistance to specific marginalized groups (e.g., street children) as well as the poverty graduation program that is under the responsibility of FNAS. • The Social Protection program focuses on social and economic risk management through social insurance and welfare programs. It includes non-contributory instruments (i.e. social transfers and health insurance for GEF) and employer insurance contributions for civil servants. Funding under this program is transferred to CNAMGS. 253. Budget allocations for social protection increased during the last three years mainly due to a more than two-fold increase in pension expenditures. The budget for social security (i.e., mission 22), pensions (i.e., mission 20), and health insurance for the GEF (i.e., ROAM and TTS tax revenues) represented about 2 percent of the total national budget 113 in 2014, before increasing to almost 6 percent in 2016 (Table 13). The budget for each of the two missions is augmented each year, but pensions received the largest increase in funding from FCAF 26 billion in 2014 to FCAF 61.3 billion in 2016, compared to an increase for social security of 15 and 19 percent in 2014-15 and 2015-16, respectively. 114 Even though there was little change in the funding for health insurance for GEF over the 2014-16 period, the new Special Solidarity Contribution tax, which went into effect in March 2017, is expected to increase funding considerably. Finally, targeted subsidies for water and electricity provision to GEF households have remained stable. 115 Table 13: Social Protection Expenditures, 2014-2016 (FCAF billion) Social Protection Instrument Finance Law 2014 Finance Law 2015 Finance Law 2016 Pensions (mission 20) 26 35 61.3 112 According to the information contained in the PAP 2016, the activities budgeted as “transfers” under this program were in fact activities to strengthen the targeting of vulnerable households (i.e., the development of a file identifying different vulnerabilities and processing data). These are not direct transfers to households but activities that should support the development of a more effective safety-net system. 113 According to the Finance Law. 114 The budget for CNAMGS was mistakenly included twice in the 2015 Finance Law but only disbursed once. 115 Subsidies for petroleum products and flour are not included because these types of subsidies do not specifically target poor households. They are also considered regressive since they only marginally reach the most vulnerable groups. According to the SIHG, the government spent about FCAF 200 billion annually on subsidies for refined petroleum products and flour (the so-called SMAG) until 2013, with less than 10 percent benefitting the most deprived households. The government has made a considerable effort to reduce this type of expenditure from FCAF 119.7 billion in 2014 to FCAF 30.5 billion in 2015, before reaching FCAF 88.4 billion in 2016. The increase in 2016 was due to a temporary reinstatement of the petrol subsidy during the election. 83 Social Protection Instrument Finance Law 2014 Finance Law 2015 Finance Law 2016 Social Security (mission 22) 35.9 66.7* (41.3) 49.1 GEF Health Insurance (ROAM)** 12.7 13.8 12.9 Electricity/Water Subsidies ** 1.2 0.99 n.a. Total Social Protection 75.8 116.5* (91.1) 123.3 Total Budget Voted 2,955 2,068 2,152 % of Total Budget for Social Protection 4.6% 2.8% (4.4%) 0.0% Note: *FCFA 25.4 billion were subsequently excluded since a part of the allocations for the CNAMGS was mistakenly included twice. The figures in parentheses are the figures after correcting for this mistake. **Data on health insurance for the GEF and for electricity and water subsidies represent the actual amount spent rather than the budget voted. 254. Contributory social protection mechanisms receive more funding than non- contributory instruments. In 2014, contributory and non-contributory programs appeared to receive roughly equivalent funding (Table 14). However, 38 percent of the total social protection budget was allocated to pensions in 2015, contributing to an increase in the share of contributory programs to 52 percent of all funding allocated for social protection. 116 In 2016, pensions represented almost 50 percent of the budget for social protection, and funding for contributory mechanisms increased to about 59 of the total social protection budget. 255. Social security expenditures for personnel has increased steadily while transfers to the poor has fluctuated. While allocations to the three social security programs are not comparable since a portion of the funding for social safety nets was managed by CNAMGS under the social protection program in 2016, a comparison of overall expenditures by function (Table 14) shows an ascending trend for Title-2 personnel. There are three different types of public expenses under the social protection program: i) insurance contributions for civil servants (FCAF 11 billion in 2015), ii) operating expenses for CNAMGS (FCAF 2 billion in 2015), and iii) cash transfers for GEF households with children (FCAF 12 billion in 2015 117). 118 The overall trend would not change if the budget for cash transfers to GEF families were included in the Title-4 transfers, but the allocations to transfers would be higher than those to personnel. Table 14: Social Security Budget, 2015-2017 (FCFA million) Type of Expenditure 2015 2016 Voted for 2017 Total mission 22: Social Security 66,745* (41,354) 49,157 51,264 116 In 2015, government contribution to the health insurance of civil servants totaled FCFA 11 billion, and half of the funds provided to CNAMGS for its own functioning totaled FCFA 1 billion. CNAMGS manages contributory and non-contributory programs, and it is not possible to disentangle the administrative costs of one type of program from another. For this estimate, the FCFA 2 billion/year for CNAMGS’ operating costs has been split in half. 117 These were not disbursed. 118 Cash transfers for GEF families with children are called family benefits (prestations sociales) because they mirror the family benefits included in the salary packages of formal-sector workers. 84 Type of Expenditure 2015 2016 Voted for 2017 BY ECONOMIC CATEGORY Title 2 - Personnel (salaries, bonuses and 52,273* (26,873) 27,060 34,770 benefits) Title 3 – Goods and Services 741 430 503 Title 4 – Social Transfers 13,532 21,667 15,990 Title 5 - Investments 200 BY PROGRAM Family Protection and Promotion 32,693* (7,293) 2,402 5,985 Title 2 - Personnel (salaries, bonuses et 26,510* (1,110) 1,028 5,010 benefits) Title 3 - Goods and Services 293 180 171 Title 4 - Transfers 5,690 1,194 804 Title 5 - Investments 200 National Solidarity 8,330 3,792 8,082 Title 2 - Personnel (salaries, bonuses et 503 772 4,454 benefits) Title 3 - Goods and Services 362 200 190 Title 4 - Transfers 7,464 2,820 3,438 Title 5 - Investments -- -- Social Protection 25,723 42,963 37,197 Title 2 - Personnel (salaries, bonuses et 25,260 25,260 25,307 benefits) Title 3 - Goods and Services 85 50 47 Title 4 - Transfers 378 17,653 11,748 Title 5 - Investments -- -- -- Piloting and Support Title 3 - Goods and Services -- -- 95 Source: Ministry in charge of Social Security: Loi de Finance 2014 and 2015/ PAPs 2015 and 2016 and Loi de Finance 2017. Note: *FCFA 25.4 billion was excluded because part of the allocations for the CNAMGS was mistakenly included twice. The figures in parentheses are the figures after correcting for this mistake. 85 256. In 2016, social transfers (FCFA 21.7 billion) accounted for about 1 percent of the total budget voted in the Finance Law, or about 0.3 percent of the national GDP. 119 The available data make it impossible to differentiate the operating costs for social transfers from other costs incurred by the ministry in charge of social affairs and CNAMGS. However, funds earmarked for social safety in Gabon appear low after assuming operating costs are twice as high as international averages (i.e., 8-10 percent for middle-income countries and 15-18 percent for low-income countries) and after including the social transfers that are “hidden” as personnel expenditures. The level of spending would be even lower if only actual spending on social transfers was considered. As a result, spending on non-contributory social protection in the entire social security budget (mission 22) only represents 0.5 percent of the country’s total GDP, lower than international averages. Spending on social safety nets averages 1.6 and 1.5 percent of GDP in upper middle- income and lower-income countries, respectively. 120 Gabon also spends less on safety nets than many of its comparator countries such as Chad, Malaysia, and Ecuador (Figure 71). Figure 71: Financing of Social Safety Nets as a % of GDP in Comparator Countries 3.5 3 2.5 2 1.5 1 0.5 0 Source: State of Safety Nets, 2017, World Bank (draft). Analysis of Expenditures: Adequacy Issues 257. Gabon’ national social protection system covers about half of its estimated 1.8 million people. Total coverage is calculated by using data from the CNAMGS database, which contains all beneficiaries of the country’s contributory and non-contributory social protection programs. In particular, the GEF database functions as a unique social register, as it is shared among the various programs for vulnerable groups. In 2016, 934,917 persons were enrolled in one of CNAMGS’ three health insurance funds: 239,351 public-sector workers and their dependents, 177,663 private- sector workers and their dependents, and 517,903 GEF (Table 15). The absence of a health insurance fund dedicated to the informal sector partly explains the system’s relatively low 119 GDP US$ 14.3 billion (2015). Official exchange rate: 591XAF = 1US $ (2015). Source: data.worldbank.org/country/gabon 120 World Bank, 2015. 86 coverage rate (the average for upper middle-income countries is 67 percent), 121 as workers in the informal sector account for 57 percent of all workers in Gabon. Table 15: Coverage of Gabon’s Social Protection System (Health Insurance), 2016 Number of Individuals % of Gabon’s Population Public Fund 239,351 13.3 % Private Sector Fund 177,663 9.9 % GEF Fund 122 517,903 28.8 % Informal Sector (other than GEF) 0 0% Total 934,918 52 % Source: Data provided by the CNAMGS, 2016. 258. There is a greater risk of errors in the health insurance funds for informal workers and GEF than for public- and private-sector workers. 123 In particular, it is estimated that a large number of households working in the informal economy that cannot obtain medical insurance are enrolled in the GEF fund even though their incomes are above the threshold. This is possible because registrations are often based on self-declaration instead of social surveys or other checks. However, it is also likely that many GEF are not enrolled in the social protection system, as the poor tend to have limited access to information 124 and/or lack the required identification documents. 125 The 2017 household survey, whose results will be available in early 2018, will be able to provide a precise estimate of inclusion and exclusion errors. Even though the number of GEF included in the CNAMGS database is more or less equivalent to the estimated number of poor people (517,903 persons enrolled in the GEF fund versus approximately 595,000 poor people estimated by extrapolating the EGEP 2005 data), it is unlikely that the coverage rate of vulnerable households is close to 100 percent. 259. Gabon’s safety-net system is estimated to cover less than one in ten GEF, or approximately 3 percent of the country’s population. Excluding non-contributory health insurance for the poor, the total number of individuals who benefitted from any type of social transfer in 2015 did not reach 50,000. Even assuming that all the beneficiaries of the various programs were GEF (i.e., there were no inclusion errors) and no one benefitted from more than one program, less than 10 percent of the 30 percent Gabonese believed to be GEF was covered by 121 The ASPIRE database, available at http://datatopics.worldbank.org/aspire/income-group-comparison/ 122 Potential beneficiaries of the FNAS and SEEG programs are included in the CNAMGS database. 123 In the past, some public or private sector workers have registered as GEF because the GEF fund was the first fund to be created under the universal health insurance scheme. This type of error is easily identifiable and the CNAMGS has already addressed it. 124 For example, it has been reported that some people fear they will be automatically registered for a political party if they sign up for insurance. 125 This is particularly a problem for the Pygmy population. 87 the safety-net system in 2015. This compares unfavorably to the 14.2, 16.4, and 32.8 percent averages for SSA, low-income, and upper middle-income countries, respectively. 126 260. The social protection system is also unable to clearly define who is a GEF, and its targeting process suffers from several key weaknesses. The government’s definition of a GEF (and the related threshold), which is the basis for identifying beneficiaries, presents a number of shortcomings. First, the income threshold is set as the minimum wage but is not linked to a measure of poverty, making it likely that the threshold is much higher than the absolute poverty threshold. Second, the definition does not take into account total household income since only the primary income of the household head is considered. Finally, the cost of living, including household composition and place of residence, is not considered. These shortcomings generate considerable risks of inclusion and exclusion errors, which are exacerbated by weaknesses in the targeting process. The system relies on self-declaration, as independent verification (enquêtes sociales) is not thorough or systematic. Moreover, the GEF database does not track household socio-economic indicators, which prevents meaningful monitoring, and there is no clear definition of a household as a consumption unit,127 making it difficult to assess the adequacy of social transfers and other social protection programs to address vulnerability. 261. Estimating the adequacy of social transfers is important to create a national safety- net system that supports households without discouraging efforts to achieve self-reliance. If the benefits that households receive are insufficient to meet their daily needs, they will inevitably engage in negative coping strategies (e.g., reduce food intake, pull children out of school, or sell productive assets) that will make it harder to break the cycle of poverty. However, benefits will encourage dependency if they are too generous. To evaluate the adequacy of the benefits in the present system, all transfers available to households need to be compared. Since there are no recent household-budget data, a preliminary analysis can be done by using the 2005 data with an updated the inflation rate. Nevertheless, policy decisions concerning the parameters of each program should only be made once the results of the next household consumption survey are available. 262. A medium-sized poor household in Gabon is estimated to require a transfer equivalent to about FCAF 211,000 a year to reach the poverty line. On the basis of the 2005 EGEPE survey, the national poverty line is calculated at about FCAF 528,700 a year per person (discounted by inflation), resulting in a 10 percent poverty gap, 128 which is approximately FCAF 53,000 a year per person. This means that a medium-sized household (i.e., two adults and two children) would require a transfer equivalent to FCFA 211,000 a year to meet their basic needs and reach the poverty line. Table 16 shows the benefits of different non-contributory programs for a household of this type, comparing implementation in 2015 with what was envisaged in the SIHG. 126 ASPIRE data, available at http://datatopics.worldbank.org/aspire/region/sub-saharan-africa and athttp://datatopics.worldbank.org/aspire/income-group-comparison/ 127 The GEF database contains 294,102 primary beneficiaries, 222,554 children, and only 1,246 spouses. This means that in most cases the husband and the wife are registered separately and both as primary beneficiaries. At the same time, it appears that many children are excluded since the average household size is 4.1. 128 The poverty gap measures the average distance that separates the poor from the poverty line. It is used to estimate the resources that would be needed to bring everyone up to the poverty line and eliminate poverty through perfectly targeted cash transfers. While it is understood that perfect targeting cannot be achieved, the poverty gap provides an indication of the resources needed to potentially eliminate poverty. 88 Table 16: Benefits of Different Non-Contributory Programs for an Average Household (FCFA/year) Cash Transfers Subsidies In-kind Social Transfers as % Uncond. Cond. Health Water & (Newborn Total of the Poverty electricity kits) Gap Envisaged in 100,000 80,000 52,000* 65,000* (50,000) 297,000 140 % the SIHG Implemented 0 0 114,800** 219,335 (50,000) 334,135 158 % in 2015 Source: Estimate based on information provided by the ministry in charge of social security. Note: *Estimate: total cost of subsidies/number of beneficiary households. **Average unit cost x 4. Figure 72: Safety Nets for an Average Household (FCFA/year) ENVISAGED IN THE SIHG IMPLEMENTED IN 2015 TOTAL: 297,000 TOTAL: 334,135 Newborn Newborn Kits Unconditi Kits 50,000 onal Cash 50,000 Health Water & Transfers Subsidies Electricity 10,000 114,800 Subsidies 65,000 Water & Condition Electrity Health al Cash Subsidies Subsidies Transfers 219,335 52,000 80,000 Source: CNAMGS, Ministry in charge of social protection. 263. Estimates indicate that the planned and actual benefits of Gabon’s various social protection programs targeting GEF households are generous, which is in sharp contrast to their limited coverage. According to the SIHG, each “regular” poor household (i.e., those that do not belong to a special vulnerable group) should receive systematic support in the form of cash transfers and subsidies, with the former accounting for least 60 percent of the total support (free goods and services should be offered only in special cases). For a medium-sized household with two children enrolled in primary education, combined transfers and subsidies would represent about 140 percent of the estimated poverty gap. When considering the service-delivery outcomes for 2015, the national system appears to have been even more generous, having provided a combined value of transfers equivalent to 158 percent of the poverty gap, though mainly in the form of subsidies rather than direct transfers. The value of most cash-transfer programs in SSA is about 20 percent of the poverty line, equivalent to about FCFA 106,000 a year per person in Gabon. The generosity of these social transfers combined with their limited coverage points to a key flaw in the country’s social protection system. 89 Analysis of Expenditures: Efficiency and Equity Issues 264. The composition of social protection programs in Gabon points to significant inefficiencies. Executed spending on social safety nets in 2015 was grossly misaligned with envisaged budget allocations, suggesting there exist significant inefficiencies in the country’s safety-net system. Subsidies accounted for 89 percent of the expenditures for GEF in 2015, while cash transfers represented only 10 percent, contrary to what is envisaged in the national strategy (Figure 73). Figure 73: Safety Nets Implemented in 2015 by Type of Intervention 2% 10% Transe. monétaires Subventions Gratuité 89% bien/services Source: Estimate is based on information provided by CNAMGS, the Ministry of, SEEG, and FNAS). Cash Transfers 265. Several cash-transfer programs are unconditional, except for education-related transfers that are conditional on school enrollment (Table 17). In 2015, CNAMGS assumed responsibility of all cash transfers. Previously, transfers that targeted “traditional” vulnerable groups such as the disabled widows, and the elderly had been the responsibility of FNAS. Table 17: Cash-transfer Programs Number of Number of Amount of Total Yearly Target Reference Program Beneficiary Beneficiary Transfer Allocation Group Year Households Individuals (FCFA) (FCFA) GEF hh with 2014 (not Family 5,000/child/ children - 75,512 172,706 10.3 billion implemented Allowance month P1/P2 since) 2013-2014 Back-to- GEF hh with 5.000/child/ (not school children - 60,000 121,941 609.7 million year implemented Allowance P1/P2 since) GEF children in private Tuition 300,000 – primary/lowe 84 84 73.5 million 2015 Assistance 2,000,000/year r secondary - P1/P2 90 Number of Number of Amount of Total Yearly Target Reference Program Beneficiary Beneficiary Transfer Allocation Group Year Households Individuals (FCFA) (FCFA) Allowance to All persons Individuals living with 5,815 5,815 75,000/year 436.1 million 2015 with disabilities - Disabilities P5 Allowance to All surviving Surviving 15,530 15,530 75,000/year 1.3 billion 2015 spouses - P4 Spouses Elderly 2013 (not Allowance Elderly GEF 386,383 35 35 13.5 million implemented - P3 (average)/year since) Emergency GEF with Support children 885,000 5 5 4.4 million 2015 (social cases) (average)/year - P1/P2 Aide to 2014 (not Professional GEF 18-25 69,375 40 40 2.8 million implemented Training year old - P6 (average)/year since) Source: Data provided by CNAMGS and FNAS. 266. The country’s two largest cash-transfer programs have been suspended for a lack of funds.129 These transfers constituted unconditional family allowances and conditional back-to- school allowances for low-income households (P1 and P2). Even though these programs received about FCFA 10-11 billion per year between 2011 and 2014 and are described in the national strategy as playing a central role in the country’s social protection, they have not been financed since 2014. Even when they were funded, their payments were irregular. For instance, part of the October 2014 tranche was only paid in 2015. The predictability of cash transfers, which allows households to stabilize consumption and implement medium-term investment strategies, is a necessary and fundamental condition for transfers to drive poverty reduction (Box 5). As a result of their inefficiencies, Gabon’s two largest programs were unable to live up to their potential and were eventually suspended. If they had been implemented today, the number of beneficiaries would have been at least four times larger than in 2015, equivalent to close to 30 percent of the poor population. 129 There were many specific programs in 2013 and 2014, such as housing assistance, funeral aid, citizenship aid, and disaster relief, that were not continued in 2015 and are not included in the SIHG. The budget for these transfers was FCFA 85 million in 2014 (FNAS data), which is a relatively limited amount considering the total envelope of social safety nets. 91 Box 5: Cash Transfers as a Long-term Poverty Reduction Tool Cash transfers are increasingly used as a social protection tool in many African countries. Available evidence suggests that well-structured cash-transfer programs, beyond providing direct consumption support to poor households, contribute to beneficiary households transitioning out of poverty as well as to local development. A recent report* summarizes the impact of different unconditional cash-transfer programs in Zambia, Kenya, Lesotho, and Ghana. These programs spay US$4-20 a month to households with children and are therefore comparable to Gabon’s family allowance program. Findings from the report suggest that: • cash transfers often have a positive impact on the productive activities of participating households (e.g., funds are often used to purchase agricultural inputs, invest in livestock, or start-up non-agricultural enterprises); • members of beneficiary households tend to work more in productive family businesses rather than spending their working time as low-paid salaried laborers; • the level of indebtedness of beneficiaries decreases and their savings increase; • there is a significant reduction in the implementation of negative response strategies to shocks such as begging or taking children out of school; • in beneficiary households, enrollment in secondary education increases and the rate of repetition in primary and secondary education decreases (in Ghana);** and • transfers to households, which are often spent locally, have a significant positive impact on the surrounding economy (the multiplier effect varies between 1.34 in Kenya and 2.5 in Ghana, which translates into US$1.34 and US$2.5 generated in the local economy for every US$ transferred to households in Kenya and Ghana, respectively). Positive impacts, however, are only possible if certain conditions are met, including: (i) ensuring the predictability of transfers; (ii) making sure transfers are adjusted for consumption and poverty levels (i.e., that benefits are neither too low nor too high); and (iii) creating measures to raise awareness on the use of transfers. Note: *Is Graduation from Social Safety Nets Possible? Evidence from Sub-Saharan Africa, IDS Bulletin Volume 46 Number 2 March 2015. **Livelihood empowerment against poverty program impact evaluation, 3IE Grantee Final Report, March 2014. 267. The Tuition Assistance and Emergency Support programs are two small-scale cash- transfer programs that in principle target the same beneficiaries as the previous two large- scale programs. However, both programs operate in an ad-hoc manner, and they provide relatively large amounts of financing to a tiny number of beneficiaries that are selected with the sole criteria of being a GEF. In 2015, the Tuition Assistance program offered 84 students, which were selected by order of arrival, support for attending a private primary or secondary school with unit costs ranging from FCFA 300,000 to FCFA 2 million per year. The FCFA 73 million used to support these 84 children could have provided back-to-school allowances to 14,700 children. The Emergency Support program is reserved for “social cases” that require shelter or money to meet immediate needs and benefitted only five households in 2015 at an average cost of almost FCFA 900,000 each, which could have financed the family allowances of 73 households for an entire year. Neither program is included in the SIHG. 268. Cash-transfer programs in 2015 favored groups considered traditionally vulnerable. While cash transfers to poor households with children have been suspended (except for the small- scale Tuition Assistance and Emergency Support programs), most current transfers are used to target specific segments of the population. 95 percent of cash transfers in 2015 went to widows/widowers and individuals living with disabilities regardless of their income, which is 92 contrary to what is stipulated in the national strategy (programs P3 and P5) 130 (Figure 74). As a result, the elderly and youth in GEF households (programs P4 and P6) have not received any transfers since 2013 and 2014, respectively. While it is reasonable to offer financial assistance to persons suffering from a specific vulnerability, this type of support tends to make it harder to generate long-term poverty reduction and economic growth (Box 5). For this reason, it is not recommended to restrict cash transfers to specific groups to the detriment of “regular” poor households. Figure 74: Cash Transfers Implemented in 2015 by Type of Beneficiary 4% Famille et 24% Enfance Veuves Handicapés 72% Source: Calculation based on information provided by the different Safety Net institutions (CNAMGS, Ministère de la prévoyance sociale, SEEG, FNAS) Subsidies 269. Targeted health subsidies remain the government’s favorite social protection tool. Despite budgetary difficulties, the main subsidy programs were maintained in 2015 at a total estimated cost of FCFA 17.3 billion (Table 18). 270. Spending on health subsidies for GEF increased from about FCFA 1-2 billon in 2008 to FCFA 5 billion in 2011, before reaching FCFA 12 and FCFA 16 billion in 2014 and 2015, respectively. There are three types of health subsidies for the GEF: (i) medical insurance that covers 80 percent of health service costs; (ii) payment of user fees that covers the remaining 20 percent of health service costs; and (iii) medical evacuations. While the medical insurance subsidy is intended for all GEF, the payment of user fees, which is normally the responsibility of the beneficiary, and medical evacuations are only considered in exceptional cases. Therefore, user fees should only be paid when they clearly exceed the financial capacity of a poor household, and medical evacuations should only be authorized for certain procedures that require skills unavailable in Gabon. Given the high average unit cost of these two subsidies (about FCFA 1 and FCFA 10 million, respectively), the eligibility criteria should be strict. However, this does not seem to be the case, as information provided by CNAMGS shows that user fees are paid simply by order of arrival until the funds run out. 130 In addition, the more than FCFA 1.3 billion given to widows and widowers in 2015 went beyond the needs identified in the national strategy, estimated at FCFA 0.7 billion per year. 93 271. Electricity and water subsidies have greatly diminished and have currently a limited coverage in Gabon. Between 2013 and 2015, the cost of these subsidies decreased from about FCFA 5.9 billion to FCFA 987 million, and the number of targeted households fell from 52,000 to 4,500. This dramatic drop was likely attributable to a change in the eligibility criteria, as the new policy mandates that the consumption of GEF is only subsidized below a given threshold (all GEF were eligible under the previous policy). Since lower-income households tend to share— often illegally—the same meter, it is likely that many GEF consume above the allowed limit. Furthermore, considering the small number of beneficiary households, it is also likely that this program benefits disproportionately urban dwellers as well as the more well-off GEF who can afford an individual meter. Even though the Water and Energy Company of Gabon (Société d'Énergie et d'Eau du Gabon, SEEG) has lower tariffs for the so-called social connections to encourage poor people to connect, subsidizing water and electricity consumption may be an inequitable way to support low-income households. Table 18: Targeted Subsidies in 2015 Number of Number of Total Subsidy Service Target Subsidy Amount Beneficiary Individual Allocation Group (FCFA) Households Beneficiaries (FCFA) Health GEF 294,102 517,902* 28,800 14,910,502,120 Insurance (average/pers/yr) Health User GEF 43 43 978,000 42,074,834 Fees (average/pers/yr) Medical GEF 132 132 10,682,457 1,410,084,395 Evacuation (average/pers) Electricity and GEF 4,500 22,850** 219,335 987,007,677 Water Subsidies (average/hh/yr) Source: data provided by the CNAMGS and the SEEG. Note: *The total number of persons registered in the GEF fund and therefore eligible for health subsidies. The average subsidy per person is calculated on this basis. However, using the number of potential beneficiaries rather than the actual users of the subsidies greatly underestimates the true average costs (the actual number of GEF that receive medical care each year is unknown). **Average household size = 5. Free Goods and Services 272. Two in-kind transfer programs that targeted mothers and persons with disabilities were implemented in 2015 (Table 19). Both were under the responsibility of CNAMGS and limited in terms of budget and the number of beneficiaries. Moreover, they were among nine programs envisaged in the SIHG to address the specific needs of particularly vulnerable groups, but they appear to have been the only programs of this type implemented in 2015. 131 131 During previous year when the programs were part of an overall social aids program administered by the FNAS, other in-kind transfers were included, such as food and school kits. 94 273. The first program distributed newborn kits to GEF mothers, mirroring the benefit included in employment packages available in the formal sector. Approximately 1,000 mothers have received a kit, suggesting a very large coverage deficit in relation to the estimated 15,000 eligible women. 132 Targeting seem to have been more or less random, but there are concerns of a strong urban bias, as kits in 2014 were only distributed in Libreville. 133 The value of each kit was estimated at about FCFA 50,000, but the total cost of the program cannot be determined because the costs for acquisition, storage, and distribution are not available. The program’s small size, urban bias, lax targeting, and high administrative costs compared to cash transfers question its suitability in Gabon. 274. The second program financed the purchase or repair of wheelchairs or other equipment for people with disabilities. The amount allocated per person varied considerably and depended on the need. For example, repairing a wheelchair could cost FCFA 250,000, but other interventions could be cheaper. These free goods and services complemented the monetary transfers allocated to persons with disabilities regardless of their income level. The option to receive in-kind goods instead of cash can be justified by the difficulties faced by many beneficiaries (especially in remote areas) to access the market for specialized equipment or by the absence of a competitive market for producers and repairers. Table 19: In-kind Transfer Programs in 2015 Number of Number of Allocation Total Program Target Group Beneficiary Beneficiary Amount Allocation Households Individuals (FCFA) (FCFA) Newborn kits GEF mothers 1,015 1,015 50,000 50,750,000 (per kit) Specialized Persons living 5,039 5,039 55,000 276,905,000 Equipment for the with disability (average) Disabled Source: Data provided by the CNAMGS. 275. The coverage of Gabon’s social assistance programs appears biased in favor of the capital area. Data collected by FNAS, which was responsible for many social assistance initiatives between 2013 and 2014 (Table 20), suggest that the vast majority of safety-net programs only reached people in Libreville and the surrounding Estuaire province. The cash-transfer and school- kit programs that cover the entire country were suspended in 2015. Assuming a similar geographic coverage after 2014, the scholarship program, which represents only 9 percent of total expenditures on safety nets, is the only remaining social assistance program with coverage beyond the Estuaire province. 132 With approximately 500,000 GEF and an estimated birth rate of 30.17 children per 1000 people, 15,000 births per year in GEF households are expected, which is much more than the 1,050 beneficiaries of newborn kits. 133 FNAS. “Bilan d’activités 2013-2015.” 95 Table 20: Geographic Coverage of Social Assistance, October 2013 to December 2014 Type of Assistance Geographic Coverage (2013-2014) Medical assistance Estuaire: Libreville and surroundings Assistance to disaster victims Estuaire: Libreville and surroundings Newborn kits Libreville Cash transfers Entire national territory Funeral assistance Estuaire: Libreville and surroundings School kits Entire national territory Gabon (Estuaire, Haut-Ogooué, Woleu-Ntem, Ngounié), France, Ukraine, School fees Cameroon Assistance for vocational training Libreville Food assistance - kits Estuaire: Libreville and surroundings Food assistance – fees Estuaire: Libreville and surroundings Housing assistance Libreville Mobility assistance Estuaire: Libreville and surroundings Birth certificates Libreville Material assistance to micro- International Widows Day: Haut-Ogooué, Moyen-Ogooué, Nyanga, projects Ogooué Ivindo, Woleu Ntem TOTAL Source: FNAS, Rapport d’activités du premier semestre 2016. The “Graduation” Program 276. To complement the assistance provided through the country’s safety nets, the government has created a program to help the GEF become financially self-sufficient and eventually “graduate” out of poverty. The program is under the responsibility of FNAS and aims to help low-income families develop IGAs by facilitating access to funding, offering technical assistance (e.g., for the formulation of business plans), and building capacity through training. Registration in the GEF database is a prerequisite for participating in the program, and beneficiaries are required to be grouped into legally constituted associations or cooperatives. The maximum cost of a project is FCFA 5 million, of which 30 percent is paid by FNAS in the form of a grant, and the remaining 70 percent is covered with a micro-credit with an average interest rate of 8 percent, excluding taxes (according to an agreement with the Gabonese Development Bank, the program’s sole financial partner). 277. However, there have been questions regarding the program’s financial sustainability since it was launched in 2015. During the first year of the program, implementation mechanisms were put in place and expenditures were essentially limited to operating costs. Implementation started in 2016 with the training of 1,198 GEF women and the issuance of grants to 73 groups at an average cost of FCFA 3.7 million per microenterprise, or an average per capita cost slightly over FCFA 1 million (Table 21). Given the early stages of the program, it is difficult to evaluate its adequacy or efficiency. Nevertheless, operating costs appear to have been very high considering the total support provided to GEF was FCFA 270 million by mid-2016 out of FNAS’s total budget of about FCFA 2 billion. 96 278. A comparison to similar programs in other developing countries suggest that Gabon should be able to obtain good results with the same or lower budget but with a wider range of services. For example, successful graduation programs in Pakistan and Ghana offer comprehensive services, including regular cash transfers and frequent home visits, for $1,987 and $1,800 per capita per year, respectively. While the pilot nature of Gabon’s program can help to explain its high administrative costs and the limited services and number of beneficiaries, there appears to be considerable room to increase its efficiency. However, FNAS is to be commended for its successful efforts to reach beneficiaries throughout the country. Table 21: Economic Self-Sufficiency Program in 2016 Number of Number of Allocation Total Service Target Group Beneficiary Beneficiary Amount Allocation Households Individuals (FCFA) (FCFA) Training GEF (women) 1,198 1,198 IGA funding GEF 249 249 1,084,769 270,107,594 cooperatives or (average) associations Source: Data provided by the FNAS. Pensions and Social Insurance 279. Gabon’s contributory social protection programs are designed to address the main risks associated with illness and invalidity, old age, and death. They are financed by contributions from employees and employers (private sector) or subsidies (public sector) and are complemented by health insurance and a number of family benefits that are available to all formal sector workers. The CPPF is responsible for almost 19,000 public-sector workers, and the CNSS is responsible for almost 24,000 formal private-sector workers (Table 22). The services provided by the two institutions are essentially the same, but family benefits in the private sector tend to be more generous. 280. Public-sector pensions are distributed regularly to retirees, but they are financed by yearly subsidies rather than regular public contributions. As a result, only a fraction of the CPPF’s estimated costs for goods, services, and investments is financed, which prevents the CPPF from extending its presence throughout the country134 and establishing the required reserves to prevent potential risks. An actuarial study has been commissioned to inform policy discussions on how to increase pensions in a sustainable way. Table 22: Pensions, 2015 Employee Employer Total Paid for Number of Operating Sector Contribution( Contribution Pensions Beneficiaries Costs (%) FCFA) (FCFA) (FCFA) Public sector 18,964 20.2 billion 32.5 billion 7% 52.7 billion 134 The CPPF has only offices in the capital. 97 Employee Employer Total Paid for Number of Operating Sector Contribution( Contribution Pensions Beneficiaries Costs (%) FCFA) (FCFA) (FCFA) Private sector 23,795 3.1 billion 119.4 billion 26%* 69 billion* Source: Data provided by the CPPF and CNSS. Note: * Includes family benefits. Recommendations 281. The following recommendations aim to increase the contribution of social protection policies to poverty reduction and inclusive growth. Recognizing the difficult macroeconomic context, they focus on measures that will make better use of available funds through more efficient and equitable approaches that will benefit Gabon’s most vulnerable populations. 282. Improve the targeting of non-contributory services to increase their impact on poverty reduction. The national social protection system suffers from significant weaknesses in its targeting of beneficiaries, which has significant implications for the overall adequacy and effectiveness of the system. First, the definition of GEF, which is the basis for beneficiary identification, presents several shortcomings that need to be addressed. Second, the targeting process relies on self-declaration, and independent verification (enquêtes sociales) is not thorough or systematic. Third, the GEF database does not contain socio-economic indicators for households, which prevents meaningful monitoring. Finally, there is no clear definition of a household as a consumption unit.135 As a result, the system is exposed to important exclusion and inclusion errors. Steps to improve targeting in the social protection system includes: • updating the poverty line based on data from the 2017 household budget survey, taking into account household composition and place of residence; • creating a new methodology to identify vulnerable households linked to overall household consumption and standard of living, using a proxy means test or poverty scorecard; 136 • determining the instruments to be used for classifying households and verifying their situation (e.g., questionnaires and scorecards); • producing an implementation manual that clearly defines the steps, roles, and responsibilities, as well as grievance and redress mechanisms; • restructuring the GEF database to include key socio-economic indicators, the vulnerability score, and indicators linked to specific vulnerabilities (e.g., disability and old age); and • improving the existing GEF database by applying the new methodology and ensuring single and complete entries for each household. 283. Consolidate and modernize social transfer programs to enhance both efficiency and adequacy through lower fragmentation. The main features of the SIHG identifies a social 135 The GEF database contains 294,102 primary beneficiaries, 222,554 children, and only 1,246 spouses. This means that husbands and wives are registered separately as primary beneficiaries. It also appears that many children are excluded since the average household size is 4.1. 136 General principles and a preliminary poverty scorecard have already been created. 98 protection system that: (i) targets economically weak households with children (rather than traditional “social cases”); (ii) prioritizes cash transfers to support consumption; (iii) provides services adapted to people with specific vulnerabilities; and (iv) promotes self-reliance through IGAs as an exit strategy. The strategy’s weakness is its administrative and operational complexity due to the large number of instruments. This excessive fragmentation keeps administrative costs unnecessarily high and complicates coordination and implementation. It is therefore recommended that the government greatly reduces the number of social protection instruments, starting with the consolidation of programs of the same type–a process that has already started with a decree that specifies priority groups for monetary transfers and exemptions. For example, the eight cash- transfer programs could be merged into one, reducing management costs and reaching more poor households with the same budget envelop. The distribution of funds could vary according to key household characteristics, which should be determined based on the results of the next household budget survey. However, distribution mechanisms, monitoring and evaluation systems, and grievance and redress instruments would remain the same. In addition, the newborn kit should be monetized and merged with the cash-transfer program, as in-kind transfers tend to be considerably more expensive to administer than monetary transfers, and this program in particular is small and inequitable. The process of merging programs should also be an opportunity to create distribution mechanisms that take advantage of modern technologies to transfer money, such as cell phones and ATM cards. This would reduce costs, discourage fraud, simplify monitoring, and facilitate the effective extension of safety-net services to the entire country. 284. Promote self-reliance as an integral part of the safety-net system to help poor households graduate out of poverty. There are no sufficient exit strategies in the current system, as there are no expectations for households to leave the GEF database once they have been registered. Moreover, the database does not contain information that would make it possible to reclassify households. The only link between the program run by FNAS to promote economic self- sufficiency for the GEF and the country’s various social transfer programs is the use of the same database to determine eligibility. The international experience of successful “graduation” programs offers important lessons for how to effectively reduce long-term poverty by offering a comprehensive package of interventions that support IGAs (Box 6). Adapting these lessons to the conditions in Gabon could help the country: • determine when GEF should be eligible to participate in “graduation” programs to avoid efforts that will have counterproductive effects for households still vulnerable to indebtedness; • ensure that households that are eligible for support to IGAs receive regular cash transfers that allow them to stabilize their consumption and concentrate on developing microenterprises rather than day-to-day survival; • integrate frequent and regular technical and psychological assistance to beneficiaries; and • develop a robust monitoring system based on objective indicators that can be used to adapt the approach to new results and track households in their path to economic self-sufficiency. 99 Box 6: Ghana’s Graduation Program In 2006, the Consultative Group to Assist the Poor launched the pilot of a “graduation” program in eight countries: Ethiopia, Ghana, Haiti, Honduras, India, Pakistan, Peru, and Yemen. The approach, which is inspired by the experience of the Bangladesh Rural Advancement Committee in Bangladesh, provides a coherent set of interventions to very poor households that are implemented in a sequential and complementary way to help beneficiaries achieve economic self-sufficiency. The program is structured around three stages: 1. Regular cash transfers are initially provided to stabilize household consumption, which is necessary for households to meet their basic needs that, if not satisfied, would prevent them from investing in productive activities. 2. Beneficiaries receive support to develop a livelihood, which takes the form of in-kind transfers (i.e., tools) or productive assets (e.g., animals) associated with specific training (e.g., in a technical field or how to manage savings) and advisory support. Assistance is provided at no cost to beneficiaries since the program targets the most vulnerable households who need to strengthen their productive capacity before they can gain access to credit. During this stage, households receive regular (often weekly) visits from the organization that runs the program to offer encouragement, management assistance for IGAs, and health training. 3. The final stage consists of linking households to microfinance services to enable them to secure the savings generated by their productive activities and/or obtain credits for possible expansions. A recent study* analyzed the results of this model in six out of the eight pilot countries. In all six countries, the program had a remarkable and lasting impact on the living standards of beneficiary households. For example, the following effects were observed in Ghana: the consumption of beneficiary households increased by 11 percent, their non-farm income increased by 91 percent, and their savings level was three times greater than that of non- beneficiary households. In addition, beneficiaries had gained a more positive view of the future and of others. The program was also economically efficient. While it was fairly expensive (about US$1,800 per beneficiary household per year), the return on investment was calculated at 133 percent, with profits in terms of increased consumption and estimated productive activities totaling more than US$7,000 per household. Source: *Banerjee, Abhiji and al. 2015. "A Multi-faceted Program Causes Lasting Progress for the Very Poor. Evidence from Six Countries." Science 348(6236). 285. Secure funding for safety nets and non-contributory health insurance to achieve sustainable results in poverty reduction. Funding for social protection programs should ideally be accumulated during economic expansions to be used during economic downturns when the need for safety nets is often greater (i.e., countercyclical). Unfortunately, many governments cut funding for social transfer programs during periods of budgetary constraints. This was the case in Gabon when the government’s two largest cash-transfer programs were suspended in 2015. As a result, the country spends less on social transfers to its most vulnerable people (as a percentage of public expenditures) than the average of both low- and middle-income countries. The GOG should therefore guarantee a minimum funding level each year, such as the FCFA 50 billion envisaged in the national SIHG, for its social transfer budget. This would make it possible to implement a nationwide long-term cash-transfer program for poor households that could be complemented by a “graduation” program as well as specialized services for particularly vulnerable groups such as street children. Guaranteed funding is a precondition to achieve predictability, a key attribute of successful safety-net programs. Social transfers need to be both regular and predictable to smooth household consumption and encourage productive investments. However, securing the funding for safety nets needs to be combined with other reforms, such as improving the targeting and efficiency of social protection instruments, to have a long-term impact on poverty alleviation. 286. Improve PFM processes, data availability, and institutional capacity to support other reform efforts. Many of the challenges in PFM facing Gabon’s social protection sector are the 100 same as in its education sector, including the need to improve budget and spending data, better align program objectives and actions in the PAPs, and ensure more consistent reporting on implementation through the RAPs. For social protection, there needs to be more clarity and accuracy in the budget nomenclature to enable the government to properly evaluate and plan expenditures. Moreover, authorities need to undertake household surveys on a regular basis to improve the quality of the PAPs. Finally, CNAMGS, which has become the main institution responsible for delivering social protection services (including health insurance), needs to be both strengthened–through adequate capacity building–and made more accountable for service delivery. Currently, the government is using a performance contract to hold CNAMGS accountable to the ministry in charge of social protection. To increase the effectiveness of services, the government should consider splitting the institution in two and separate the health insurance management (CNAM) from the management of the safety nets (GS). 101 6. HEALTH Introduction 287. This chapter examines recent trends in health spending. It reviews the key characteristics of Gabon’s health sector, including the status of key health indicators and the composition and distribution of health expenditures across agencies at different levels of the public administration. The analysis concludes with recommendations designed to enhance the efficiency of public health spending, which is especially critical in the context of the government’s ongoing fiscal consolidation. 288. The analysis focuses on the years 2010-2015/16. It draws primarily on the 2016 National Health Accounts (NHA), which provide highly detailed information on the health sector over the 1995-2015 period, which encompasses the 2007 reform of Gabon's health financing and social protection systems. Additional data are drawn from the Finance Laws and BOOST dataset, as well as the National Health Development Plan (Plan National de Développement Sanitaire, PNDS). Context The Structure of the Gabonese Health Sector 289. The Gabonese health sector consists of three parallel health systems. The first is the large public health system managed by the Ministry of Health (MOH) and other ministries. The second is the “para-public” system governed by the CNAMGS. The third is the private-sector health system, which consists of various private health facilities and healthcare providers. In total, Gabon has an estimated 959 health facilities, ranging from small rural health posts to urban university teaching hospitals. 137 290. The MOH is both a key healthcare provider and the health sector’s chief regulator. In addition to managing the public health system, the MOH defines standards of care and promulgates guidelines for healthcare providers in the public and private sectors. MOH facilities are divided into three groups: (a) Primary facilities: These include dispensaries and health posts (cases de santé) at the community level, health centers (centres de santé) in urban areas, and medical centers (centres médicaux) at the departmental level; (b) Secondary facilities: These include regional health directorates (directions régionales de la santé) and specialized agencies (e.g., for epidemiology, sanitation, and maternal and child health services), which coordinate primary care at the regional or provincial level, as well as regional hospitals (centres hospitaliers régionaux), which provide referral and hospital services; 137 Gabon Statistical Yearbook 2011. 102 (c) Tertiary facilities: These include centrally managed referral diagnostic services, national health programs, national administrative and technical offices, and research institutions, as well as three tertiary-level hospitals located in Libreville. 291. Though extensive by regional standards, Gabon’s primary healthcare system is underdeveloped. The geographical coverage of public primary health facilities is relatively equitable, though some remote areas, such as the Northern Health Region, have too few facilities to serve the local population. Moreover, even in areas where health centers are widespread, they are often not functional. Rural facilities face especially serious challenges in retaining health staff and accessing pharmaceuticals and other supplies. Many rural facilities also lack basic equipment, including an adequate cold chain for vaccinations. In addition, public health outreach programs are not sufficient to address the country’s high rates of communicable disease. 292. The health sector has an estimated total of 4,000 beds, or 25 beds per 10,000 inhabitants. The bed occupancy rate is 44 percent, 138 below the average for regional comparators such as Cameroon (49 percent) and Congo-Brazzaville (60 percent). However, many of these beds are located in non-functional primary care facilities, which skews the occupancy ratio downward. The ratio of occupied beds to beds in fully operational facilities is likely much higher. 293. At the national level, the total number of health staff is broadly adequate. The health sector’s workforce is estimated at about 12,000, of which the overwhelming majority (11,385) are employed in the public sector. Gabon has 0.3 public-sector medical doctors per 1,000 inhabitants, or 0.5 per 1,000 when the military, para-public, and private health systems are included. Gabon’s doctor-to-population ratio is within the range recommended by the World Health Organization and Joint Learning Network (WHO-JLN) and in line with the ratios of peer countries such as Botswana (0.3) but lower than those of wealthier comparators such as South Africa (0.8), Ecuador (1.7), and Malaysia (1.2). The national ratio of midwives to women of childbearing age is one per 935, more than four times the WHO standard of one per 4,000. The population-to-state-nurse ratio is one per 2,190 inhabitants, also well above the WHO standard of one per 4,000 inhabitants. 294. However, the distribution of health personnel is skewed toward urban areas, especially the country’s largest cities, but is generally consistent with the distribution of the population. In 2014, 44 percent of health personnel were concentrated in Libreville-Owendo, while the shares in all other health regions ranged from 9 percent to just 3 percent (Figure 75). This distribution is largely aligned with the relative population sizes of different regions. However, Libreville-Owendo, the Center-South, East, and West regions have somewhat larger proportions of health staff than their respective population shares, while the South-East Maritime, and North regions have fewer health staff per capita. 138 Gabon Statistical Yearbook 2011. 103 Figure 75: The Distribution of Healthcare Workers and the Distribution of the Population, 2014 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% % of population % of health workforce Source: 2014 National Health Accounts. 295. The distribution of health services continues to favor hospital-based curative care at the expense of preventative care delivered by primary health facilities. Although preventive care, community-based care, and public outreach programs can significantly improve health indicators at a relatively modest cost, Gabon’s health system continues to devote a large share of resources to expensive curative care delivered by hospitals. Nevertheless, treatment options for non-communicable diseases, particularly high blood pressure, diabetes, and cancer, remain limited, particularly in the interior of the country. 296. Under the new 2017-2021 PNDS, the MOH plans to substantially increase investment in operationalizing the primary health care system at the local level. The strategy could shift resources toward cost-effective preventive care, community-based health services, and public outreach. The PNDS is designed around nine strategic axes: (a) Promoting good health at all stages of life; (b) Strengthening the management of infectious and parasitic communicable diseases; (c) Intensifying efforts to combat non-communicable diseases; (d) Supporting epidemic preparedness, monitoring, and response capabilities and building capacity to manage other public health emergencies and disasters, including efforts undertaken at the national level in the context of International Health Regulations; 139 (e) Strengthening the governance of the health sector and the leadership of the MoH; (f) Establishing a well-functioning national health information and public health surveillance system; (g) Operationalizing health departments; (h) Achieving progress toward universal health coverage by: (a) optimizing the management of human resources; (b) providing facilities with adequate equipment and infrastructure; (c) increasing the availability of essential generic medicines, medical devices, vaccines, 139 (2005). 104 blood products, and diagnostic services; (d) improving patient reception, quality of care, and safety; (e) improving the management of financial resources; and (f) reducing the cost of care; and (i) Fostering community ownership of health systems and encouraging public participation in the activities of the health sector. 297. The reform of Gabon's health financing and social protection system that began in 2007 with the creation of CNAMGS has enabled the government to implement a national compulsory health insurance system. CNAMGS operates a third-party-payer health insurance scheme under the supervision of the Ministry of Social Welfare. Its Board of Directors includes representatives of the government, private-sector employers, and employees. It is headed by a director general, who is assisted by two deputy general managers. 140 CNAMGS has decentralized structures (branches) in each of Gabon’s nine provinces. Its objectives are to: (i) expand access to healthcare by reducing health costs, pooling financial resources, and avoiding the use of temporary or informal medicine; (ii) improve the quality of care by signing service contracts with health facilities; (iii) enhance the health status of vulnerable populations and contribute to social protection; and (iv) support efforts to combat poverty and social exclusion. The reform created a fund exclusively dedicated to supporting GEF, who represent about 30 percent of the total population. The GEF fund is financed by a tax known as the obligatory fee for health insurance (redevance obligatoire à l'assurance maladie, ROAM). CNMAGS administers two additional funds for workers in the public sector and the formal private sector, which were created in 2011 and 2013, respectively. 298. CNAMGS is responsible for the accreditation and monitoring of primary, secondary, and tertiary service providers, both for outpatient care and hospitalization. The CNAMGS Department of Medical Control and the Fight against Fraud accredits health facilities according to a set of well-defined criteria. Each health facility signs a partnership agreement with CNAMGS, and at end-2014 the health sector encompassed 325 accredited providers, including 99 pharmacies and dispensaries, 99 private clinics and medical centers, 97 public hospitals and medical centers, and 30 military hospitals and medical centers. The MoH defines the medical nomenclature used by CNAMGS, the list of essential drugs for health facilities at each level, and the regulations for opening private health facilities. Health Indicators, Access to Healthcare, and the Quality of Health Services 299. Gabon has a relatively high fertility rate, and a growing number of youth will continue to increase demand for healthcare and other social services. In 2016, Gabon’s population of 1.8 million was growing at a rate of 2.9 percent, and in 2013 the total fertility rate was estimated at 4.1 percent. 141 The share of the population under age 15 is sizeable at 37 percent. At its current growth rate, Gabon’s population will double by 2045, while the dependency ratio142 will decline from 78 percent in 2011 to 69 percent in 2025. 140 The director general and deputy general managers are nominated by the Board of Directors and appointed by the President of Gabon. 141 WHO, 2016. The total fertility rate is the number of children expected to be born to women of child-bearing age. 142 The youth under-14 years of age as a proportion of population in the working age group (15-64 years). 105 300. Gabon’s health indicators are relatively low by the standards of peer countries and those with similar levels of health spending. In 2011, average life expectancy at birth was 63 years (60 for men and 64 for women). The under-five mortality rate is 65 deaths per 1,000 births, well above the level of comparator countries such as Malaysia, Cameroon, and the Republic of Congo (Figure 76a). Although Gabon’s infant mortality rate (43 deaths per 1,000 live births) and under-five mortality rate have declined since 1992-1996, both have remained broadly stable since 2013. 143 Infant and under-five mortality rates are highest in the North region and lowest in the East. Figure 76: Infant and Maternal Mortality Rates, Gabon and International Comparators (a) Infant Mortality Rate (b) Maternal Mortality Rate Worse than average Performance relative to health spending per capita Performance relative to health spending Worse than average Gabon Gabon Congo, Rep. Swaziland Swaziland Cameroon Namibia Cameroon Namibia Congo, Rep. Algeria Algeria Tunisia Mauritius Tunisia Better than average Mauritius Better than average Malaysia Malaysia Better than average Worse than average Better than average Worse than average Performance relative to income Performance relative to income per capita Source: World Bank staff calculations based on data from the IMF, WHO, and the World Bank’s WDIs. Note: The comparator countries shown in this figure have levels of per capita GDP and public health spending similar to those of Gabon. The same international comparator group is used throughout this chapter. 301. Gabon has made similarly limited progress in reducing maternal mortality rates over the past decade. The maternal mortality rate is estimated at 277 deaths per 100,000 live births,144 far above the levels of comparable countries (Figure 76b). Hemorrhage and hypertension are among the chief causes of maternal mortality. High maternal mortality rates reflect the limited and inequitable coverage of comprehensive, high-quality maternal health services, such as antenatal care, institutional deliveries, postnatal care, and emergency referrals. The rate of contraceptive use is low at 31 percent, and the unmet need for contraceptives is substantial at 26 percent and highest among lower-income households. HIV, tuberculosis, malaria, and respiratory diseases are the main causes of mortality across all age groups (Figure 77). 143 GDHS, 2013. 144 GHDS, 2013. Other estimates range from 230 (WHO, 2010) to 520 (PNDS, 2010). 106 Figure 77: Disability-Adjusted Life Years (DALYs) and Causes of Mortality, 2012 (a) (DALYs***), all ages, per 1,000 (b) Causes of Mortality, all ages, % HIV, TB, malaria Maternal, neonatal,… HIV/AIDS Other infectious diseases* Lower respiratory… Other NCDs** Malaria Neuro-psychiatric… Stroke Cardiovascular… Ischaemic heart disease Acute respiratory… Diarrhoeal diseases Unintentional injuries Tuberculosis Chronic respiratory… Cancers Preterm birth… Musculoskeletal diseases Protein-energy… Suicide, homicide and… Birth asphyxia and birth… 0 50 100 150 200 250 0 5 10 15 20 all ages DALYs (thousands Percentage (%) Source: WHO, 2015. Note: *Infectious diseases other than acute respiratory diseases, HIV, TB and malaria **Other non-communicable diseases include non-malignant neoplasms; endocrine, blood and immune disorders; sense organ, digestive, genitourinary, and skin diseases; oral conditions; and congenital anomalies. 302. Gabon is undergoing a nutritional transition. In 2012, about 6.5 percent of children under the age of five were underweight for their age (i.e., malnourished), and about 7.7 percent were overweight. Boys are more likely than girls to be either under- or overweight. Approximately 17.5 percent of children were significantly below the appropriate height for their age (i.e., stunted), and 3.4 percent were underweight for their height (i.e., wasted). Only 32 percent of infants were breastfed within the first hour of birth, and just 6 percent of infants were fully breastfed until six months of age. 303. Overall health-service utilization rates have increased in recent years, but wide disparities persist across regions. While most key health indicators have remained stable or improved since 2001 (Table 23), access to and utilization of health services is highly unequal. The rates of medical consultations and hospital admissions per capita show especially large variations between regions and across departments within regions. Table 23: Key Health Indicators, 2001-2013 Indicators GDHS, 2001 GDHS, 2013 Infant mortality rate (deaths per 1,000 live births) 57 43 Under-five mortality rate (deaths per 1,000 births) 84 65 316 (GDHS, 2013) 519 (GDHS, 2001) Maternal mortality rate (deaths per 100,000 live births) 230 (WHO, 2010) 420 (WHO, 2000) 520 (PNDS, 2010) Contraceptive use 33% (12% modern) 31% (19% modern) Pre-natal care visits (%) 94% 95% 107 Postnatal care visits (%) 66% 90% Institutional deliveries (%) 85% Urban 94%; rural 70% 17% 32% Child (12-23) with full immunization coverage (%) (measles=55%) (measles = 74%) 48% 68% Children under five with ARI receiving treatment (%) (urban 52%, rural 34%) (urban 71%, rural 52%) Children under five with diarrhea receiving ORT (%) 33% 37% Children under five with fever seeking treatment (%) 62% 67% Children under five with fever taking anti-malarials (%) -- 26% Children under five sleeping under a mosquito net (%) -- 51% Source: Gabon Statistical Service. GDHS, 2001, 2013, WHO, 2010. 304. Coverage rates are relatively high for certain maternal care services, but coverage rates for other essential services and indicators of service quality remain low and variable across regions. During the past decade, the coverage of prenatal care has remained consistently above 90 percent, and institutional delivery rates exceeded 90 percent in 2012. However, access to postnatal care and family planning services remains limited nationwide, and the quality of maternal health services is in urgent need of improvement, especially at the primary care level. Substantial economic and geographic disparities in utilizations rates reflect constraints on both the supply and demand for health services. Coverage rates for maternal health services are far lower in rural areas (70.3 percent) than in urban centers (93.9 percent), and institutional delivery rates exceed 90 percent among urban women, compared to just 70 percent among rural women. 145 305. While the coverage of child health services has steadily increased, indicators of preventive care are still relatively poor. Just 51 percent of children sleep under bed nets, and although the share of children with measles vaccinations increased from 55 percent in 2000 to 74 percent in 2012, this rate remains dangerously low. Important deficiencies also persist in the treatment of certain diseases. For example, only 37 percent children with diarrhea receive oral rehydration therapy. 146 306. Households at all income levels rely on public healthcare facilities. Most Gabonese women give birth in public facilities, regardless of income level. 147 However, at least one-third of women in the highest consumption quintile report using private facilities, and at least one-fifth of women in the lowest consumption quintile deliver at home. Across all income levels, patients seeking consultations often bypass lower-level facilities and go directly to public hospitals. Public dispensaries are more popular among patients from lower consumption quintiles, whereas patients from higher consumption quintiles tend to prefer private clinics. 148 307. In the 2012 GDHS, respondents who had given birth in public healthcare facilities and were dissatisfied with the experience most often cited poor quality of care—especially 145 GDHS, 2013. 146 Ibid. 147 Ibid. 148 EGEP, 2009. 108 respondents outside Libreville. 54 percent of respondents reported being dissatisfied with the quality of services, and 30 percent cited the attitude of health personnel as the reason for their dissatisfaction. Although fewer people reported negative attitudes among health personnel in 2012 than in 2000, it was still the number one reason for patient dissatisfaction. While the lack of personnel was mentioned by only 2 percent of respondents, 18 percent reported inadequate health personnel, likely referring to an insufficient number doctors and trained specialists. Many respondents noted that facilities lacked equipment or drugs, and while these reports varied across areas, lack of equipment and drugs was among the top three concerns among respondents from rural areas and cities other than Libreville. Only 10 percent of respondents reported excessive wait times, and just 4 percent of respondents were dissatisfied with the cost of care. However, the cost of care was a much more significant concern in rural areas than in cities. In rural areas, 17.9 percent of respondents cited the high cost of care as the reason they did not deliver in a health facility, compared to just 5.8 percent of respondents in urban areas. The distance to a health facility was also much more serious obstacle in rural areas, where 38.9 percent of respondents reported excessive distance as the reason for not delivering in a health facility, versus 20.5 percent in urban areas. 308. Disaggregating respondents by income level reveals stark disparities in health-service delivery indicators, with respondents from the poorest quintile almost always reporting the worst outcomes. Among maternal health indicators, the rate of teen childbirth was highest among the poorest households, while the prevalence of family planning techniques (including condoms) was lowest. Only 67 percent of respondents in the lowest consumption quintile reported giving birth while attended by a skilled healthcare worker, compared to 97 percent of respondents in the high consumption quintile. Financial barriers to accessing health services were also most common among households in the poorest quintile. Similar disparities were evident for some, but not all, child health indicators: respondents from wealthier households were much more likely to report having a child with fever treated by a professional healthcare worker, either at a health facility or pharmacy, but rates of vaccination coverage were higher among the poor than the wealthy. This may reflect the fact that vaccination services are free in Gabon, removing the financial barrier that discourages poor households from obtaining other types of care. 109 Table 24: Health-Service Indicators by Consumption Quintile, 2012 Health indicator (%) Financial Professional Modern barrier as Child Childbirth Delivery treatment Quintile family reason for not completely among in a sought for planning, seeking care vaccinated girls age health children including among women (including 15-19 facility under five condoms of reproductive yellow fever) with fever age 1st (poorest) 40.6 11.9 75.5 82.4 30.0 51.0 2nd 30.5 18.8 92.4 73.1 42.0 63.0 3rd 25.8 20.2 94.1 77.3 18.2 73.3 4th 15.5 22.1 94.7 69.4 36.4 69.3 5th (wealthiest) 9.2 21.9 96.3 55.2 16.1 84.0 Total 22.5 19.4 90.2 70.4 29.1 67.2 Source: Gabon Demographic Health Survey, 2013. Health Expenditures Expenditure Levels and Trends 309. Total health spending in Gabon is below the average for SSA as a share of GDP, but above the average in per capita terms. In 2015, health spending equaled 3.3 percent of Gabon’s GDP, well below the SSA average of 6.2 percent. However, in nominal terms health spending was equivalent to US$599 per capita, among the highest levels in the region (Figure 78). Relative to SSA comparators, Gabon’s public-sector accounts for an especially large share of total health spending (Figure 79). Figure 78: Total Health Spending per Capita, Gabon and SSA Comparators, 2014 1,400 1,200 1,000 USD, PPP 800 Gabon, 599 600 400 200 Central African… Sao Tome and… 0 Botswana Senegal Congo Burkina Faso Seychelles Niger Zambia Ethiopia Chad Cameroon Uganda Sierra Leone Swaziland Guinea-Bissau South Africa Lesotho Madagascar Guinea Liberia Zimbabwe Ghana Kenya Source: WHO, Health Accounts, 2016. Note: Figures expressed in constant international US dollars in purchasing-power-parity terms. 110 Figure 79: General Government Health Spending as a Share of Total Health Spending, Gabon and SSA Comparators, 2014 100 90 80 Gabon, 68 Percent (%) 70 60 50 40 30 20 10 Democratic Republic… Cabo Verde… 0 Botswana Nigeria Benin Burkina Faso Congo Eritrea Mauritania Niger Sierra Leone South Sudan Namibia Sudan Cameroon South Africa Burundi Djibouti Lesotho Liberia Guinea Zimbabwe Mozambique Gabon Source: WHO, Health Accounts, 2016. 310. Gabon’s expenditure levels are even lower by the standards of its global peers. Gabon spends significantly less on health than the average of countries with similar income levels, both as a share of GDP and in per capita terms (Figure 80). Moreover, Gabon compares unfavorably to most peer countries in terms of both total health spending and public health spending. Figure 80: Total Health Spending per Capita and GDP per Capita, Gabon and International Comparators, 2014 Total Health Expenditure (THE) per Capita versus Income per capita in current US$ (2014) 1,000 Total Health Expenditure (THE) per 900 800 700 600 Capita in US$ 500 400 300 Gabon 200 100 0 - 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 GDP per capita US 111 Total Health Expenditure (THE) as share of GDP versus Income per capita (2014) 20 % Gross Domestic Product (GDP) Total Health Expenditure (THE) 15 10 5 Gabon 0 - 5,000 10,000 15,000 20,000 25,000 GDP per capita US Source: WDI, 2014. 311. Like total health spending, Gabon’s level of public health spending is lower than those of comparable countries. In 2015, Gabon’s public sector spent the equivalent of 1.5 percent of GDP. When social security spending is included, this share rises to 2.2 percent. Both figures are well below the shares spent by most countries with similar income levels (Figure 81), as well as regional and global peers such as South Africa (4.3 percent), Botswana (3.1 percent), and Ecuador (4.5 percent). Public health spending in Gabon is also relatively low on a per capita basis, and as a share of general government expenditures. In 2015, Gabon’s government allocated about 8.9 percent of the budget to the health sector, in line with the regional average (9 percent), but below the average for countries with similar income levels (11 percent) and far below the Abuja Declaration target (15 percent). 149 Figure 81: Public Health Spending and Income per Capita, Gabon and Comparators, Latest Available Data 18 16 Public Health Spending 14 12 (% of GDP) 10 8 6 4 2 0 Gabon 0 5000 10000 15000 20000 25000 30000 GDP per Capita (2014 US$) Source: WDI, 2014. Health Spending by Source 312. While Gabon’s health expenditures remain below both the levels of global peers and the Abuja target, the government has steadily increased public health spending over time. General government health expenditures increased from about 5.7 percent of total general 149 In 2001, representatives of the African Union’s member states met in Abuja, where they pledged to increase health spending to at least 15% of their annual budgets. 112 government spending in 2008 to 8.9 percent in 2015, more than doubling in real terms over the same period (Figure 82). Meanwhile, the public sector’s share in total health spending grew from 39 percent in 2008 to 66 percent in 2015 (Table 25), bringing Gabon above the SSA average. Table 25: Health Spending in Gabon, 1995-2015 General Total Total General Social Out-of- Government Private Health Health Total Government Security Pocket Health Health Spending Spending Health Health Spending (% Spending Spending Spending per per Year Spending Spending of General (% of (% of (% of Total Capita Capita (% of (% of Total Government Total General Health (US$ at (US$ in GDP) Health Health Health Government Spending) Exchange PPP Spending) Spending) Spending) Budget) Rate) Terms) 1995 3.4 5.2 36.4 14.5 63.6 54.0 158.0 418.8 2000 2.9 5.3 40.3 14.2 59.7 50.7 118.1 338.5 2005 3.0 5.2 40.7 14.4 59.3 50.3 187.2 385.3 2006 3.2 5.6 41.0 13.5 59.0 50.1 216.0 421.6 2007 3.3 6.5 41.9 12.3 58.1 49.3 261.6 457.5 2008 2.7 5.7 39.7 13.5 60.3 51.2 283.4 376.3 2009 3.7 6.9 42.5 24.9 57.5 48.8 291.3 493.6 2010 3.5 10.7 47.4 27.1 52.6 44.6 322.4 490.7 2011 3.4 7.8 52.9 27.1 47.1 40.0 401.4 516.4 2012 3.5 8.2 51.2 27.1 48.8 41.4 396.7 558.2 2013 4.0 10.1 74.0 20.0 26.0 20.5 380.0 494.0 2014 3.0 7.4 68.0 26.0 32.0 22.1 321.0 494.0 2015 3.3 8.9 65.6 30.9 34.4 21.3 245.9 Source: National Health Accounts, World Health Organization. Figure 82: Health Spending in Gabon, 1995-2014 400 Health Spending (US$ in PPP 350 335.9 300 250 200 terms) 150 158.1 100 109.2 50 0 1995 2000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 General government expenditure on health (GGHE) Private expenditure on health (PvtHE) Out of pocket expenditure Source: WHO, Health Accounts, 2016. 113 313. The launch of the CNAMGS national health insurance program has dramatically altered the composition of health financing sources over the past decade. Health-sector funding comes from three main sources: (i) the government budget; (ii) contributions from employers and employees and non-contributory insurance systems (including CNAMGS, social security, and some private insurers); and (iii) out-of-pocket (OOP) expenditures by households and other forms of private spending. Public spending has reached 66 percent of total health spending, driven by rising CNAMGS and social security expenditures. 150 The MOH’s share in total health spending declined from 36 percent in 2010 to 28 percent in 2015, while CNAMGS’s share increased from 3 percent to 23 percent. 314. Prior to the establishment of CNAMGS, health spending in Gabon was mainly financed by taxes, via the national budget, and by OOP payments by households. Additional funding was provided by employer and employee contributions to social security, private insurance payments, tontine funds, charitable donations, and external assistance. The establishment of CNAMGS has contributed significantly to both the changing composition of health financing and the significant increase in total general government health spending. As noted in the previous chapter, spending on GEF health subsidies rose steadily from 2008 to 2014, with a more modest increase in 2015. Figure 83: Public Health Financing by Source, 2010-2015 40 35 30 Ministry of Health Percent (%) 25 Other ministries (Defense, 20 Education) 15 National health insurance (CNAMGS) 10 Social Security (CNSS) 5 0 2010 2011 2012 2013 2014 2015 Source: MOH. Note: This disaggregation is not entirely aligned with the national accounts’ data, but the magnitude and trends of the large aggregates are similar. 315. Funding for CNAMGS and other social health insurance systems comes from a combination of general and earmarked taxes, contributions from public-sector workers, and contributions from employers and employees in the parastatal and private sectors. Until 2016, the GEF was financed by the ROAM and a 1.5 percent levy on money transfers outside of the West African Economic and Monetary Community. In 2017, the ROAM was replaced by a new tax known as the Social Solidarity Contribution (Contribution Spéciale de Solidarité, CSS), which has a wider tax base. The GEF’s funding sources and accounting nomenclature are different 150 Public spending via the National Social Security Fund (Caisse Nationale de Sécurité Sociale) represented about 30 percent of all general government health spending in 2015. 114 from those of other funds, which are financed through more traditional employers and employees’ contributions. 316. The share of household OOP payments in total health spending has declined substantially over time, falling from 51 percent in 2008 to 21 percent in 2015. OOP spending was almost halved in real terms over the same period (Figure 81). The WHO benchmark for OOP health spending is no more than 15–20 percent total health expenditures. Gabon’s share of OOP health spending has fallen below this benchmark over the past seven years, as the CNAMGS program has expanded and public spending has increased. 317. Though rising overall, the trajectory of public health spending has been volatile, which could threaten its long-term sustainability. Large annual fluctuations in health spending as a share of total spending reflect the lack of a consistent strategy for health expenditures combined with revenue volatility. This same pattern is also evident in the sharp fluctuations in CNAMGS financing (Figure 83), which were exacerbated by the narrow ROAM tax base and the inconsistency of budgetary transfers. Moreover, the income from each CNAMGS funding source is dedicated to a specific fund, and CNAMGS cannot use cross-subsidies to equalize spending. As discussed in the social protection chapter, if the CSS is implemented effectively, it should help to increase and stabilize CNAMGS revenues. However, further measures will be necessary to enhance the sustainability of non-insurance-related public health financing, especially funding for primary care facilities. Expenditure Analysis: Efficiency Issues 318. Rising levels of public health spending have not generated commensurate improvements in health indicators, suggesting both technical and allocative inefficiencies in the use of resources. While public spending per capita in 2013 was 2.46 times higher than it was in 2013 ($335.9 vs $136.4), maternal mortality only decreased by about 39 percent, and under-five mortality by about 25 percent. Moreover, health spending in Gabon appears to generate less value for money than in comparable countries. For example, maternal and child mortality rates are significantly higher in Gabon than in most countries with similar levels of per capita health spending (Figure 84). In a context of fiscal consolidation, enhancing the allocative and technical efficiency of health spending will be especially important. 115 Figure 84: Health Spending and Health Indicators, Gabon and Upper-Middle-Income Comparators, 2014 600 180 Maternal mortality ratio (per Mortality rate, under-5 (per 500 160 100,000 live births) 140 400 120 100 1,000) 300 Gabon 80 200 60 Gabon 40 100 20 0 0 0 1000 2000 3000 0 1000 2000 3000 Health expenditure per capita, PPP Health expenditure per capita, PPP Source: WDI, 2016. Note: The mortality rates shown are slightly lower than those reported in other databases. 319. The MOH budget prioritizes hospital care. According to 2011-2015 PNDS data, hospitals receive 58 percent of the MOH budget. Most of these funds go to regional secondary hospitals, but 20 percent of MOH funding goes to tertiary hospitals. Tertiary hospital funding is likely to increase in the medium term, as the authorities plan to upgrade several regional hospitals to tertiary-level care. Primary care facilities receive only 16 percent of the MOH budget, and public health programs receive just 13 percent. These shares are low by regional standards. In SSA, spending on primary care averages 34 percent of the health budget, with shares rising as high as 56 percent in Côte d’Ivoire and 47 percent in Ghana. 151 Hospitals also receive more funding than primary care facilities when controlling for the relative number of beds or health staff. 320. In addition, 80 percent of the budget for the 2011-2015 PNDS was allocated to curative care, with only 20 percent allocated to preventive care. The funding allocation in the new 2017-2021 PNDS is similarly unbalanced, with 75 percent going to curative care and 23 percent to preventive care. A heavy focus on hospital-based curative care at the expense of primary and preventive care represents a major source of allocative inefficiency, as the international evidence clearly shows that primary and preventative health spending generates a much larger impact on health outcomes. 321. Divergent trends in the amount of funding allocated to recurrent and capital expenditures are another key source of inefficiency. Between 2010 and 2015, recurrent spending rose by 29 percent, while spending on investment and maintenance fell by 28 percent (Figure 85). As a result, recurrent spending now represents over 80 percent of total health expenditures. 151 Primary Health Care Performance Initiative, 2017. 116 Figure 85: Trends in Recurrent and Capital Spending, 2010-2015 400 350 339 300 293 Billions FCFA 283 275 250 263 247 244 220 212 228 226 200 176 150 100 92 68 71 65 50 43 49 0 2010 2011 2012 2013 2014 2015 Recurring Investment Total Source: NHA 2010-2015. 322. Non-salary recurrent spending on preventive health care is very low, and actual spending is even lower due to below-average execution rates. 152 Non-salary administrative costs consume an especially large share of the budget, and only a very small share is spent on training. The BOOST dataset for the 2015 MOH non-salary budget is organized into four programmatic areas: health-service provision and access; prevention and health safety; 153 combatting HIV/AIDS; and sectoral coordination. The total MOH non-salary budget for 2015 was about FCFA 46 billion. 154 86 percent of the budget was dedicated to health-service provision and access, 3 percent to prevention and health safety, 2 percent to HIV/AIDS, and 9 percent to sectoral coordination (Figure 86). Staff training received just 0.6 percent of the budget for health-service provision and access, or 0.5 percent of overall non-salary spending. The overall execution rate was 99.4 percent, but the prevention and health safety category had the lowest execution rate at 89 percent, while execution rates for all other categories were between 99 and 100 percent. This clearly suggests a suboptimal distribution of resources, which negatively affects the quality and efficiency of health services. 152 Unlike the PNDS, the BOOST expenditure data do not disaggregate by level of care and only include salary data for 2015. Despite these limitations, they provide a useful breakdown of MOH non-salary spending. 153 Prevention and health safety includes vaccinations, health and hygiene outreach, and disease prevention, inter alia. 154 This figure does not include the CNAMGS and CSS budgets, which appear under another sector of the budget. 117 Figure 86: Ministry of Health Non-salary Budget, 2015, Planned and Executed, FCFA 45 Billions 40 35 30 25 20 15 10 5 - Prevention and Supply and Fight against Coordination for Health Security access to health HIV/AIDS health policy and services social welfare Approved Executed Source: BOOST, 2017. 323. A recent evaluation of the financing of the 2011-2015 PNDS found that while the execution rate for the investment budget reached 91 percent, most financing went to tertiary care and curative services rather than to primary and preventive care. This finding is consistent with other reports on the desultory state of primary care facilities. The first strategic axis of the PNDS investment budget is to “ensure geographic access to quality health services,” and 65 percent of the original budget was allocated to the primary and secondary levels of the health system—via the operationalization of regional and departmental health authorities—and only 35 percent was allocated to tertiary hospitals. However, a full 89 percent of the executed investment budget was spent on tertiary care facilities, while just 11 percent was spent on both the primary and secondary levels (Figure 87). Figure 87: The Planned and Executed Investment Budgets of the 2011-2015 PNDS Planned Investment Budget Executed Investment Budget Reinforce tertiary care Reinforce tertiary care Make Health Regions Make Health Regions operational operational Make Health Departments Make Health Departments operational operational Availability of essential drugs Availability of essential drugs and medical productions and medical productions Information system/public Information system/public health surveillance health surveillance Governance/leadership Governance/leadership Source: Evaluation Finale du Plan National De Développement Sanitaire 2011 – 2015 du Gabon, 2016. 324. The overwhelming majority of planned PNDS investments in primary care were not implemented. The plan called for constructing 105 new primary facilities and rehabilitating another 284. Yet by the time the plan was completed, only 8 facilities had been built and another 8 rehabilitated. Because cost-effective preventative health services and community outreach 118 efforts are typically provided at the primary level, the underdevelopment of primary facilities significantly reduces the overall efficiency of health spending. Table 26: 2011-2015 PNDS for Investing in Periphery and Primary Care Services Type of Facility Built Rehabilitated Planned Executed Planned Executed Departmental Hospitals (Medical Centers) 11 0 18 3 Health Centers 10 6 20 5 Dispensaries 84 2 246 0 Total 105 8 284 8 Execution rate 8% 3% Source: Evaluation Finale du Plan National De Développement Sanitaire 2011 – 2015 du Gabon, 2016. 325. The distribution of expenditures confirms the health system’s overreliance on hospitals. Most health spending by the government, social insurance systems, and households went to public hospitals (31.7 percent), pharmacies and other providers of medical goods (27.6 percent), private hospitals (8.3 percent), public walk-in care providers (6.7 percent), and private walk-in care providers (3.6 percent). 155 An apparent demand-side preference for hospitals over primary care facilities compounds the low levels of government spending on primary health care. In this context, the absence of well-functioning referral and counter-referral mechanisms are likely leading to cost inflation. 156 Figure 88: The Distribution of Health Spending (public and private) by Provider Type, 2014 Public hospitals and nursing homes Private hospitals and 22.2% nursing homes 31.7% Public ambulatory care providers Private ambulatory care providers 27.5% 8.3% Pharmacies, other medical product 6.7% providers Other 3.6% Source: 2014 National Health Accounts. 326. Finally, the apparent misalignment of health spending by disease group with Gabon’s epidemiological profile indicates further allocative inefficiencies. The prevention and treatment of communicable diseases is highly cost-effective in Gabon, where deadly communicable diseases 155 National Health Accounts, 2014. 156 Poor referral systems are particularly problematic for university hospitals, which have difficulty carrying out their primary training and research tasks because they are overwhelmed by patients who should be receiving care at dispensaries and health centers. 119 such as malaria and typhoid are endemic. However, the 2014 NHAs show that non-communicable diseases drive health spending, even though they represent only the fourth-leading cause of mortality and impose the fourth-largest disease burden in terms of disability-adjusted life years. Non-communicable diseases account for 34.3 percent of health expenditures, followed by communicable diseases at 32.3 percent, reproductive health at 17.1 percent, trauma at 5.7 percent, and nutritional deficiencies at 0.2 percent (Figure 89). Malaria prevention and treatment is the largest single expenditure item (9.5 percent), followed by trauma (8 percent), sensory organ diseases (4.6 percent), respiratory infections (4.6 percent), and tumors (3.9 percent). The prevention and treatment of HIV/AIDS is the tenth-largest expenditure item, accounting for 3.4 percent of health spending, yet it is the leading cause of death in Gabon, responsible for 15 percent of all deaths (Figure 90). The misalignment of health spending with the disease profile is consistent with an excessive emphasis on hospital care and a limited focus on preventive care. Figure 89: The Distribution of Current Health Expenditures by Disease Group, 2014 Infectious and parasitic 10.4% diseases 5.7% Reproductive health 32.3% Nutritional deficiencies Non-communicable diseases 34.3% Trauma 17.1% Other diseases/conditions not specified 0.2% Source: 2014 National Health Accounts. 120 Figure 90: Current Health Expenditures by Disease or Type of Care, 2014 20 17.13 15 Percentage 9.5 10 7.99 7.49 5.65 4.6 4.58 3.873.863.85 5 3.53 3.37 3.03 2.53 2.25 1.7 1.641.57 0.63 0.41 0.260.18 0 0 Reproductive health Malaria Other diseases and diseases not specified, non-communicable Other diseases not specified, infectious and parasitic Trauma Disorders of the sensory organs Respiratory tract infections Tumors Digestive tract diseases Vaccine-preventable diseases Diseases of the urogenital apparatus HIV/AIDS and other Sexually Transmitted Diseases Cardiovascular diseases Diarrheal diseases Endocrine and metabolic diseases Mental and behavioral disorders and neurological conditions Oral diseases Respiratory tract diseases Tuberculosis Neglected Tropical Diseases Drepanocytosis Nutritional deficiencies Other diseases/conditions not specified Source: National Health Accounts, 2014. Expenditure Analysis: Equity Issues 327. In addition to reducing allocative efficiency, prioritizing hospitals at the expense of primary care distorts the equity of health spending. The poorest households are least able to access hospitals and most likely to use primary clinics and public dispensaries, while the opposite is true for the wealthiest households. GEF health subsidies improve distributional equity by attempting to cover at least the insurance portion of the health costs borne by poor households. However, as described in the previous chapter, inclusion and exclusion errors diminish the targeting accuracy of GEF subsidies, while funding volatility for the GEF and reliance of facilities on user fees leads to a “first-come, first-served” approach to service delivery limits their equity impact. Equity concerns may explain why the cost of care was as much or even more of a concern among respondents in the 2012 GDHS than it had been in the 2000 GDHS. 157 328. Reflecting the uneven allocation of hospitals and doctors across the country, public health spending is inequitably distributed and does not reflect the health needs of local populations. Resources are concentrated in urban areas and wealthier regions. Some concentration of heath resources is inevitable in a country with a low population density, a poor road network, and a population that is 86 percent urban, but the inequalities observed in Gabon go beyond the constraints of geography and demographics. Per capita public health funding varies substantially across regions, with the highest levels in Libreville-Owendo and other more heavily urbanized regions and the lowest levels in the country’s poorest and most rural regions (Figure 89 and Figure 91). Libreville-Owendo is home to only 43 percent of the population, yet it receives 70 percent of public health expenditures (Figure 92). In the country’s other nine regions, the relative size of the 157 It should be noted that these surveys do not capture the significant decrease in private spending observed in 2013 and 2014. See Table 25. 121 population is consistently larger than the share of health spending, with particularly significant disparities in the Southeast and North. 329. Because most of Gabon’s hospitals are located in Libreville, and close to half of all health expenditures occur at the hospital level, the geographic distribution of resources is again consistent with a systemic emphasis on hospital-based care. Health resources are distributed to reflect the disposition of hospitals, not the health needs of the population. There is no apparent relation between health spending and poverty, and indeed per capita spending is generally lower in the poorest regions. Consequently, the lack of clear criteria for distributing public health resources across regions exacerbates geographic inequalities and ultimately reduces the allocative and technical efficiency of health spending. Figure 91: The Public Health Budget per Capita, Figure 92: Population Distribution versus Public 2015 (US$) Health Expenditure Distribution (%) 120.00 80% 100.00 70% 60% USD $ 80.00 50% 60.00 40% 30% 40.00 20% 20.00 10% 0.00 0% Hospitals Primary/ambulatory care % of population % of total health expenditures Source: NHA 2015. Source: NHA 2014. Recommendations 330. Reorienting expenditures toward preventive care and public outreach efforts executed through primary and secondary care facilities could greatly enhance the allocative efficiency of health spending. The distribution of health spending does not reflect either Gabon’s epidemiological profile or the demographic distribution of its population, which reduces both the efficiency and the equity of the health budget. Resources are not allocated to either the regions or disease groups where they would have the greatest impact, and tertiary hospitals provide services that could be supplied at a much lower cost by primary and secondary care facilities. The investments in primary care planned for the 2011-2015 period could have restructured the health sector, easing its emphasis on hospitals, but these investments were largely unrealized. 331. As Gabon strives to meet its objectives for the health sector in a context of fiscal consolidation, improving the efficiency and distributional equity of sectoral resources will be especially critical. To achieve its goals, the government must (i) operationalize health districts, (ii) improve referral systems, (iii) strengthen pharmaceutical management, (iv) develop transparent criteria for allocating resources, (v) establish oversight mechanisms to ensure investment plans are 122 fully implemented, and (vi) create more efficient budget-management and incentive systems, such as input-based payments. 332. Even in the current period of fiscal consolidation, Gabon has several options to improve the mobilization of health resources. As discussed in the social protection chapter, the proper implementation of the CSS could both increase and stabilize CNAMGS revenues. Other measures to enhance public health financing, especially for primary care facilities, include greater financial decentralization, expanded copayments for preventive and outreach services, and the creation of new budgetary space through broad fiscal reform. Allowing health facilities to deposit user fees into dedicated bank accounts and increasing copayments for primary care could help match the supply of health services with demand. The authorities should also develop a comprehensive financing strategy for the healthcare that reflects the objectives of the new PNDS. 333. To enhance the allocative efficiency of health spending, the government must rebalance both capital and current expenditures away from hospitals and toward primary care facilities. The primary care subsector lacks adequate staff and funding, and its coverage is limited. Effective primary care can reduce overall health costs by providing locally appropriate prevention and treatment options. Investing in primary care is critical to manage communicable diseases, improve maternal and child health indicators, and address the rising incidence of non- communicable diseases and lifestyle-related health conditions. Like the previous strategy, the 2017-2021 PNDS allocates a large share of expenditures (over 80 percent) to the regional, departmental, and community levels (Table 27). However, the highly uneven execution of investment projects severely undermined the previous strategy’s efforts to rapidly expand primary care. Successfully executing the regional-, departmental-, and community-level expenditures outlined in the 2017-2021 PNDS would substantially reorient health-sector financing away from tertiary services and toward more efficient and cost-effective primary care. Table 27: 2017-2021 PNDS Public Health Budget by Expenditure Type, Level, and Year Year Recurrent (FCFA) Investment (FCFA) 2017 36,298,869,388 43,675,000,000 2018 39,755,736,439 76,355,000,000 2019 37,369,094,033 47,305,000,000 2020 39,097,226,989 42,580,000,000 2021 40,419,234,941 40,930,000,000 Proportion 43% 57% Level of care Budget (FCFA) Proportion Community 11,646,280,554 2.6% Department 283,964,662,966 64.0% Region 64,492,549,690 14.5% Central 83,681,668,580 18.9% Total 443,785,161,790 100.0% Source: PNDS 2017-2021. 334. Improving the quality of primary care facilities could reduce costs by encouraging patients to seek treatment at the appropriate level of the health system. In an effort to obtain quality health services, patients routinely bypass underfunded and understaffed primary care 123 clinics in favor of hospitals. Strengthening the primary care system could alleviate the burden on hospitals and lower overall costs, both for the health system and for patients. While CNAMGS reimbursements to providers have the potential to improve primary care services, quality issues reduce demand, reducing the benefits of demand-side financing. Directly investing funds from the central government budget in primary care facilities could significantly increase the quality of primary care in the short term, strengthening public confidence in primary care facilities and leveraging CNAMGS reimbursements to catalyze a virtuous cycle of demand-driven growth at the primary level. 335. Consistent with the reorientation of spending toward primary care, the authorities should shift focus from curative care and treatment to preventive care and public outreach. Immunization coverage rates and the use of bed nets are both rising, and accelerating progress in these areas could significantly improve key health indicators at a modest fiscal cost. Additional interventions to control malaria are not widespread and should be expanded. Overall, the government devotes insufficient resources to public health programs and tends to favor facility- based treatment over community-based prevention. 336. Creating an effective referral system would complement investments in primary care and further reduce overall health costs. Well-functioning referral networks are especially important for maternal and child care and emergency services. Referring patients who do not require expensive hospital-based treatments to less costly primary care facilities and community- level programs could significantly reduce the marginal cost of care. 337. Redeploying staff to better reflect the disease profile could improve care and reduce costs. Currently, a large share of health staff is concentrated in hospitals and urban areas. However, certain health workers, such as nurses and midwives, could be more effective in primary care facilities at the community level, given the appropriate training. Community-based workers are also needed to implement local programs. 338. Investments in family planning can provide an especially cost-effective means of improving maternal and child health indicators. Rates of contraceptive use in Gabon are low, and though contraceptives are relatively inexpensive, there is substantial unmet demand. Meanwhile, abortion rates are high, and unsafe abortions are a major cause of maternal death. Providing subsidized contraceptives as part of the basic benefits package covered by CNAMGS could significantly reduce abortion rates and promote better overall maternal and child health outcomes. 339. Adopting performance-based financing (PBF) approach by the government could help improve the quality of care, especially at the primary level. The international experience has demonstrated the value of PBF in strengthening both preventive and curative health services, enhancing the managerial and supervisory activities of health district teams, and encouraging citizen engagement in the delivery of health services. PBF can facilitate the transfer of resources to the point of service delivery and improve accountability for the utilization of these resources. Gabon recently launched a PBF pilot in four health regions and two health departments, and the authorities should closely monitor its results. 124 340. The government should increase funding for health-facility maintenance and public outreach and convert standard budget lines to performance-based payments. The current health budget is based on distributing the available resources, and it does not adequately reflect demand for services or health outcomes. Consequently, operations, maintenance, and public outreach (including HIV/AIDS programs) are systematically underfunded, and there are significant disparities in per capita health spending across regions. Introducing PBF principles, including direct transfers to service providers based on results achieved, will help increase value for money in health spending. 341. Increasing the execution rate for the health budget could improve expenditure efficiency. In recent years, execution rates have fallen to as low as 70 percent in some categories of health spending, due in part to deficiencies in the budget formulation and preparation processes, a lack of oversight mechanisms for budget execution, and procurement delays. Increasing budget- execution rates in the health sector will require a comprehensive effort to strengthen PFM. These issues are explored in greater detail in Chapter 1. 342. Boosting investment in staff training and strengthening performance incentives for healthcare workers could enhance the quality of health services. The government invests very little in training health staff, with negative consequences for patient care. Moreover, performance incentives appear to be weak. Although a bonus system exists, the high rate at which survey respondents cite the poor attitude of personnel as a reason for dissatisfaction with health services suggests that these bonuses are not an effective motivational tool. The authorities should review the bonus system and explore opportunity to introduce new performance-based incentives. In other countries, PBF mechanisms have been shown to effectively spur behavioral changes in staff. 343. Reorganizing tertiary facilities could enhance the efficiency of the hospital system. Large-scale investments in hospital construction have resulted in an excess of hospital beds, and many hospitals have occupancy rates as low as 40 percent. The authorities could increase efficiency by merging hospitals and converting underutilized facilities to primary care centers for maternal health, child health, and emergency services. In addition, many regional hospitals lack appropriate specialists, who could be redeployed from the capital. 344. Higher-quality data would support more informed decision-making. The government relies on irregular household surveys to monitor progress in the health sector, and while institutional information systems exist, they produce limited data. The MOH has established an information management unit (Cellule d´Observation de la Santé Publique), and efforts are underway to expand population-based data and create geographical information systems. Greater investment in IT will be necessary to enhance sectoral oversight, and facility-based information systems will be crucial to monitor performance and assess local needs for pharmaceuticals and medical supplies. While CNAMGS has an online platform to register beneficiaries, an online system for managing claims could improve the efficiency of claims management and promote timely reimbursement. The World Bank’s e-Gabon project, launched in 2016, is expected to substantially improve Gabon’s health-information systems. 125 ANNEX A: EDUCATION Details on Education system and service delivery characteristics The education sector of Gabon is governed by Law No.21/2011 of February 14, 2012. This law: (i) defines the fundamental principles, missions and academic organization of education, training and research; (ii) determines the terms and conditions for teachers, trainers and researchers, as well as other personnel in the education, training and research sectors; (iii) defines the general framework of curricula, training supplies, school and academic calendars 158, guidelines and social care for learners; and (iv) establishes a general framework for quality assurance in education, training and research, and for PPPs. Preschool is essentially private and is not compulsory. It concerns the educational population whose age range is between three and five years. The general organization of Gabonese preschool was established in 1984 159. At present, most preschool institutions are private. In 2015-2016, only 35% of preschool schools were public (412 out of 1177 schools). Enrollments in preschool have been growing. The number of children enrolled in preschool was around 67,400 in 2015-2016 according to the Ministry of Education (Table A.1), which represents an estimated 50 percent gross enrollment rate based on the population estimates provided by United Nations Educational, Scientific and Cultural Organization (UNESCO) school statistics. Table A.1. Enrollment at pre-primary, primary and secondary, 2001 to 2015 2001 2002 2003 2011 2015 Pre-primary 15,568 na na 45,225 67,400 Primary 265,714 281,871 279,816 317,946 295,870 Lower secondary 75,673 78,546 na na 139,227* Upper secondary 25,045 26,645 na na 46,090* Secondary 100,718 105,191 na na 185,317* Sources: UNESCO database, 2001 to 2011; MENEC for 2015. Note: *enrollment in secular private secondary unknown. Primary education includes five years of study since the beginning of 2005 school year. Before 2005, primary education was 6 years but after the introduction of the competency-based approach (approche par les compétence de base), primary education has been reduced to five years instead of six, organized into three phases: exploratory, basic, and advanced. The official primary school age is 6 to 10 years. At Grade 5, students must sit for the CEP, the primary exit examination. There are 1,666 primary schools in the country, of which 776 are public and 890 private (about 54 percent). From 2001 to 2011, enrollment increased from 265,700 students to 317,900, an increase of 20 percent over 10 years (Table A.1), which follows a period of more significant growth in the 1990s. In the academic year 2015-2016 about 296,000 students were enrolled in primary of which about 172,000 in the public sector (meaning that about 42 percent of students were enrolled in the 158 School year in Gabon lays between the beginning of October to end June of the following year. Official exams are in July. 159 Law No.10/84 of July 30, 1984 defines the general organization of pre-school. 126 private sector). At the same time, in the same academic year, there were 11,001 (primary and secondary) teachers of which 6,456 working in the public sector and 4,550 working in the private sector (about 41 percent). General secondary education spans a seven-year period divided into two cycles. The first cycle lasts four years, while the second cycle is three years. Since 2014, access to general secondary education is not competitive anymore. However, it is conditioned by the average of scores at Grade 5 of primary education and the average scores obtained at the CEP examination. At the end of the first cycle of general secondary education, students must sit for the first-cycle BEPC examination. At the end of the second cycle of general secondary education, students sit for the Baccalauréat which gives access to higher education. High school is for the age group aged 11 to 17 years. Technical and vocational education is provided in public and private institutions offering short courses and long courses. Admission to TVET is competitive. The system awards certificates of vocational aptitude (Certificat d’Aptitude Professionnelle, CAP) after two years of education, or certificates for professional studies (Brevets d’Etudes Professionnelles - BEP) in three years. Approximately 13 public schools offer a long technical education: 11 technical high schools (LT) leading to Technical and Vocational Baccalauréat, a national commercial school in Port-Gentil and an institute of technology at Owendo that trains high level technicians (Brevet de Techniciens Supérieurs, BTS). The technical and vocational education program is run by the Direction Générale de l’Enseignement Technique (DGET) within MENEC. In 2015-2016, there were respectively about 8,600 and 807 students and teachers in technical and vocational education. Vocational training offers five types of training. Vocational training is provided through: (i) training by apprenticeship for handicrafts by master craftsmen; (ii) training by ten public vocational training and professional development training centers; (iii) training by 51 private institutions; (iv) training by six enterprise-based training institutions; and (v) training by institutions such as the Petroleum Institute, the Mining School, and the Wood School, which attract students from the sub-region and beyond. The vocational training program is run by the department of vocational training (Direction Générale de la Formation Professionnelle, DGFP) within the MTEFTPIJ and is operated by the national training and professional development agency (Agence Nationale de Formation et de Perfectionnement Professionnels, ANFPP). Training centers award vocational training certificates (Certificats de Formation Professionnelle, PSCs). The length and the duration of the training depends on the type of subject and profession. The TVET subsector represents a limited enrollment share of overall enrollment in secondary education. Of the total number of youth enrolled in post-primary education, only about 8 percent are enrolled in TVET institutions. This compares unfavorably with other emerging economies where the TVET enrollment rates often reach 30 to 40 percent. Vocational education is constrained by capacity and spatial considerations. Most institutions are concentrated in urban areas and regional administrative centers. Private technical education providers are very limited. Each year, over 5,000 applicants apply to vocational training institutions but only approximately 1,700 gain admission. Figure A.1. shows the distribution of enrollment per province in technical and vocation education institutions. 127 Figure A.1 Enrollment (stock) in TVET (2015-2016) Figure A.2 Enrollment in tertiary education (without ENS) (2015-2016-2017) 160 4000 Enrollment (Public) Enrollement (Private RUP) 507 ENSET 3500 3000 2062 INSG 2301 534 2500 2000 IST 3531 1270 1500 IUSO 1000 500 UOB 0 8836 USS USTM Source: MENEC - DGETP 2016. Note: ENSET: Ecole normale supérieure de l’ enseignement technique; INSG: Institut national des sciences de gestion; IST: Institut supérieur de technologie; IUSO: Institut universitaire des sciences de l’organisation; UOB: University Omar Bongo; USS: Université des sciences et de la santé; USTM: Université des science et techniques de Massuku. The Ministry of Higher Education provides three programs: tertiary education courses, scientific research and innovation, support to students. There are three public universities (university of Umar Bongo, university of science and technique of Masuku) and four graduate schools (normal school 161, technical normal school 162, management school, institute of technology). Higher education institutions (HEIs) receive subsidies from the State for public service charges. HEIs are autonomously managed but are accountable to the Ministry of Higher Education through performance-based contracts. The contract specifies objectives and reciprocal commitments and indicators with the program manager. No data is available for private HEIs although they are expected to have developed significantly during the last decade as per the MHE. Scientific Research and Innovation is the responsibility of the National Center for Scientific Research, which is composed of five institutes. The support to student life program includes financial support to ensure living conditions conducive to academic success (bursaries distributed according to per capita grants, canteen, etc.). Migration to Licence-Maîtrise-Doctorat 163 (LMD) system was enacted in 2007 but implementation started only in 2016. About 20,000 students are enrolled in tertiary education with a high concentration in one university. The University of Omar Bongo enrolls almost half of university students in Gabon 160 For UNS Gestion and USTM, data are only available for 2015-16 and data for ENS are not available. 161 The two normal schools for primary teachers (Ecole normale d’instituteurs) train teachers for one year after the Baccalaureat. Normal school (Ecole normale supérieure - ENS) trains pedagogical advisors, inspectors, and secondary teachers in three years for bachelor diploma holders, in two years for masters’ holders to get the certificat d’aptitude au professorat des collèges (CAPC) for middle school teachers, and certificat d’aptitude au professorat de l’enseignement secondaire (CAPES) for senior secondary teachers. 162 Technical normal school or ENSET trains teachers: three years for Baccalaureat + 2 years holders to get certificat d’aptitude au professorat des collèges d’enseignement technique (CAPCET) for junior technical secondary education teachers, two years for CAPCET holders or engineers to get certificat d’aptitude au professorat des LT (CAPLT) for senior technical secondary. 163 Bachelor-Masters-PhD 128 (Figure A.2.). The technical normal school INSET and the college of technology IST with a total enrollment of about 500 students each represent only 6 percent of the total enrollment. Only 25 percent of youth enrolled in technical schools are girls (with 68 percent of them enrolled in the health school). Although data is missing for ENS and ENSET, evidence shows that there are few candidates in science, languages, and philosophy in both institutions as the country must import important number of secondary school teachers from abroad every year. 129 ANNEX B: SOCIAL PROTECTION Details on Safety Nets Programs Table B.1: Safety net programs in the SIHG Families w/ P2 – Single Handicapp P7 – Street P6 - Youth children children mothers Widows Elderly Mechanism P1 – P3 – P4 – P5 - ed Cash Transfers: 20 billion /year Food transferts X X X X Health transfers X X X Back-to-school allowances X X X Youth transferts X Extra transferts X Back-to-school supplies X Subsidies: 10 billion / year Water and electricity X X X X Health (including user fees) X X X X X X X Free goods and services: 10 billion/year Food kits X Hygienic kits X Newborn kits X Medical devices for people with disabilities X Professional training X X Childcare X X Legal assistance X X Institutional care X X X Foster families X Solidarity income: 4 billion/year Payment for services rendered to the State X X X X Guaranteed day of work X X X X Source: SIHG 2014 –Technical report. 130 ANNEX C: EXPENDITURE TRENDS Expenditure Data use and Constraints Expenditure data quality is mixed in Gabon. The country does not yet have a fully functional Integrated Financial Information Management System (IFMIS) and expenditure and budget data are instead scattered across multiple separate Treasury IT platforms. Each of these platforms only records data for a portion of the expenditure execution chain and does not include local government and special account expenditures. The social sector expenditure analyses conducted for this Public Expenditure Review relied mainly on official figures from initial and revised budget laws (lois des finances) and their annexes 164, budget execution reports 165 (lois de règlement) and their annexes, and Court of Account budget execution auditing reports. These analyses of education, health and social protection somewhat aggregate official figures were complemented with analyses of more granular data extracted directly from Gabon’s Treasury systems for 2014 and 2015, and reorganized into a (more user-friendly) BOOST database. 166 Official reports’ figures were used wherever possible, particularly when expenditure totals from the BOOST did not match official numbers, or when reports’ numbers included local government expenditures. 167 BOOST data analysis results were used wherever other data sources did not include information detailed enough to help explain trends. The introductory macro and the overall expenditure chapters used both Gabonese Treasury tables of government financial operations and IMF staff estimates and projections. Both Treasury tables and IMF estimates are for Central Government only, and chapters 1, 2 and 3’s overall expenditure figures may therefore be slightly lower than wage bill and social sector chapters’ figures. Gabonese authorities’ data rather than IMF data was used wherever both were available, but they differed. Cross-country fiscal data comparisons used IMF Government Financial Statistics and World Economic Outlook data. Gabon’s budget classification was significantly modified in 2015 with the notable introduction of program budgeting (budgétisation par objectifs de programme) and the coverage and the very definition of government expenditures were also altered. For instance, 2010-2014 debt service–related expenditure official numbers include both expenditure on interests and amortization. 2015 debt service expenditure data does however follow international practice and only includes interest spending while amortization is below (rather than above) the line. Similarly, spending for corporate restructuring and social plans is recorded within the debt service or within the ‘other expenses’ totals depending on the year. Robust analyses of medium- and long- term trends are therefore particularly difficult. 164 Since 2015 annexes to the budget laws include the PAPs that describe the strategy, objectives and milestones of each program. 165 Since 2015, annexes to the budget execution laws include the RAPs. These performance reports detail the budget of each mission and program, its execution and the associated results. 166 See http://wbi.worldbank.org/boost/boost-initiative for more details about BOOST. 167 Though efforts were made to reconcile all discrepancies between BOOST and budget execution report data, differences of the order of -/+ 1 to 5 percent remained for selected economic categories of spending. 131 From 2015 onward, the focus shifted from budgeting (and execution) by economic line item and specific administrative unit to budgeting by ministry/section and mission (program and action), as well as providing a global budget envelope for each broad economic category (wages, goods and services, subventions and transfers, and investments). The level of disaggregation of the data available in most official documents decreased and analysis of specific line item spending (e.g. maintenance spending) can now only be performed with detailed Treasury data. Finally, there is no real functional classification in use Gabon. The functional breakdown of expenditures has to (and can roughly) be approximated from the breakdown by ministry (for 2010- 14) and mission, program, action (for 2015 onward), but, for instance, a significant portion of (investment) expenditures is recorded as cross-sectional and not disaggregated either by ministry or by mission. In addition, permanent staff wages and salaries (i.e. approximately 80 percent of the total wage bill) are not broken down by administrative unit or mission within Treasury systems, except for 2017 budget data, but rather included in Ministry of Budget or mission 14 (public finance management) expenditures. Official data was corrected (to the extent possible) to reflect proper ventilation of payroll spending by mission in all functional expenditure breakdowns presented in this report. The wage bill analysis used payroll system data instead of Treasury expenditure data and includes local government spending; Treasury and BOOST data do not. 132 ANNEX D: THE PERFORMANCE BASED FINANCING MECHANISM IN PUBLIC HEALTHCARE Key features The PBF program is a systemic reform of the health System designed to enhance the coverage and quality of health services and improve the efficiency and equity of service provision. Implementation of PBF best practices is necessary to achieve the expected results. These Best practices include: (i) Autonomous health providers; (ii) Separation of functions; (iii) Regulator not interfering in health facility management and concentrating on regulatory issues (indicators, quality reviews, costing); (iv) Public and Private Partnership, with equal treatment of public, private, religious providers; (v) Importance of cash payments of providers and their access to accredited distributors of inputs operating in competition; (vi) Importance of verification, coaching, action research done by Contract Development & Verification agencies; (vii) Community voice empowerment is made through local NGOs (verification, patient satisfaction); (viii) Balancing revenues and expenditures by health providers; (ix) Stimulating economic multiplier effects; (x) Competition between health provider for contract, and between accredited wholesalers of inputs ; (xi) Applying multi-sectorial PBF. The program gives more autonomy to health facilities to find local solutions to challenges faced to deliver quality health care to the population. Financial resources are allocated to health facilities in cash based on their performance. In addition to PBF resources, the health facilities collected money from patient directly or indirectly (through insurance mechanism) for the health care delivered to the patients. Health facility managers and their staff meet each quarter to develop their business plan, and each month to decided how to use their resources in the most efficient way to achieve their objectives defined in their business plan, and share performance bonuses coming from the benefit made by the health facility. PBF uses systematic and detailed registry data from health facilities, as well as community-level verification, to ensure that patients for whom PBF subsides are paid to the health facility, have in fact used the health services reported by providers. Service quality is measured by a checklist, which includes indicators of hygiene and waste- management practices, such as the availability of incinerators, latrines, and a water source, as well as the absence of organic waste or used syringes in the facility. The confidentiality of consultations and the availability of functional medical equipment are also assessed. In addition, the quality review teams analyze a sample of patient records for certain services, such as prenatal care, to assess adherence to the recommended protocols. Managers’ role The program also allows facility managers greater autonomy to run facility operations and patient care. Managers can choose input suppliers, open a bank account, and hire and fire personnel, inter alia. The PBF program also provides tools and mechanisms to improve the management of health facilities, including a template for a quarterly business plan, which establishes quarterly objectives and describes strategies for achieve them. The program also offers different verification and evaluation tools that allow facilities to self-assess the quality of the services they provide. In addition, an index tool helps facilities manage revenues and expenses and plan its budget. The index tool disaggregates income (i.e., user fees, PBF subsidies, government financing, support 133 from nongovernmental partners, etc.) and expenses (i.e., drugs, equipment, recurrent costs, etc.) from the previous quarter and relates those data to activities in the current quarter. Income of the health facilities come from different sources: i) Internally generated revenue (directly paid by patient through user fees or indirectly through insurance) and ii) Externally generated revenue (government budget, PBF subsidies, contribution from partners). PBF subsides represent only a small proportion of health income which varies according to the performance of health facilities and whether the health facility is based in urban or rural area. Apart from purchasing pharmaceuticals, equipment and other necessary inputs from suppliers, health facilities continue to receive input related to vertical programs such as immunization, HIV program, etc... from other donors. Oversight and regulation District authorities and regional health delegations regulate the program, and they also sign their own performance contracts. Regulatory activities include supervising and monitoring facilities, assuring the quality of essential medicines by approving suppliers, and evaluating the technical quality of facilities. In addition, Directorates at the Ministry of health have their own performance contract to develop and update norms and standard procedures of health care to be applied at different level of health system, develop training modules, design continue cascade training and integrated supervision of health personnel at regional and operational level. 134