GUINEA-BISSAU PUBLIC EXPENDITURE REVIEW Managing Public Finance for Development Document of the World Bank Macroeconomics, Trade and Investment Global Practice AFCF1 Country Management Unit Africa Region The World Bank Group This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. i Acknowledgments The Guinea-Bissau Public Expenditure Review was prepared by a team led by Fiseha Haile Gebregziabher (TTL and Economist, GMTA2) and Kjetil Hansen (co-TTL and Sr. Public Sector Specialist, GGO). The team also included Bernard Harborne (Lead Technical Specialist, GSUGL), Edson Correia Araujo (Senior Economist, GHN04), Gabriel Yorio Gonzalez (Public Spector Specialist, GGO), Susana Sanchez (Senior Economist, GMTA2); Sylke von Thadden (Consultant, GMTA2), Jorge Retana de la Peza (Consultant, GMTA2); Nicolas Drossos (Consultant, GGO); Nelly Rakoto-Tiana (Consultant, GED); Andry Ralijaona (Consultant, GSUGL); Paul Maximilian Bisca (Consultant, GTFMR); Melanie Marie Laloum (Consultant, GMTA2); Isolina Rossi (Consultant, GMTA2). The team has received very useful inputs form Bernardo da Cruz Vasconcellos (Education Specialist, GED07); Kebede Feda (Senior Economist, GED07); and Patricia De Sousa Robalo P Rosa Merle (Consultant, GED07). Overall guidance was provided by Louise Cord (Country Director, AFCF1); Lars Christian Moller (Practice Manager, GMTA2); Alexandre Arrobbio (Practice Manager, GGOAW); Christine M. Richaud (Lead Economist, GMTA2); Paolo Zacchia (Program Leader, AFCF1); and Kristina Svensson (Country Representative). Maude Jean-Baptiste and Micky O. Ananth provided an excellent administrative and operational assistance, respectively. Sally Hinchcliffe (Consultant, GMTA2) edited the report. The peer reviewers are Maimouna Mbow Fam (Lead Financial Management Specialist, GGO), Jeremy Henri Maurice Veillard (Senior Health Specialist, GHN), and Rick Emery Tsouck Ibounde (Senior Economist, MTI). The team gratefully acknowledges the collaboration with the Government of Guinea Bissau. Cover picture taken from Hirschhofer (2017), OECD Development Matters. Available at https://oecd- development-matters.org/2017/06/20/development-finance-2-0-improving-conditions-for-local-currency- financing Regional Vice President: Makhtar Diop Country Director: Louise Cord Senior Global Practice Director: Carlos Felipe Jaramillo Practice Manager: Lars Christian Moller Task Team Leaders: Fiseha Haile Gebregziabher (TTL) and Kjetil Hansen (Co-TTL) ii Abbreviations and Acronyms AENF Alfabetização e educação não formal (Literacy MICS Multi-Indicator Cluster Survey and Non-Formal Education) MINSAP Ministerio da Saude Publica (Ministry of Public ARCP Public Procurement Regulatory Authority Health) CAR Central African Republic MMR Maternal Mortality Ratio CIT Corporate Income Tax MOE Ministry of Education CITES Convention on International Trade in MOH Ministry of Health Endangered Species of Wild Fauna and Flora MPA Ministry of Public Administration CMNND Communicable, Maternal, Neonatal, and NBE State Budgetary Nomenclature Nutritional Disease NCD Non-Communicable Diseases CPIA Country Policy and Institutional Assessment NMR Neonatal Mortality Rate CPLP Community of Portuguese Speaking Countries OOP Out of Pocket DGA Direcção-Geral da Alfândega (Directorate PALOP Portuguese-speaking African countries General for Customs) PEFA Public Expenditure and Financial Accountability DGASS Direção Geral da Administração da Saúde PBM Performance-Based Management (Directorate General for Health Administration) PFM Public Financial Management DGCI Directorate-General of Taxes and Duties PHC Primary Health Care DAF Administrative and Financial Directorate PIP Public Investment Program DALY Disability-Adjusted Life Year PIT Personal Income Tax DNT Non-Titled Expenditures PNDS Plano Nacional do Desenvolvimento Sanitário DRE Regional Directorate of Education (National Plan for Health Improvement) DRS Direção Regional de Saude (Regional Health PPP Purchasing Power Parity Directorates) ppts Percentage Points DSA Debt Sustainability Analysis RESEN 2015 Education Status Report EAGB Electricidade e Aguas da Guine-Bissau SAB Autonomous Sector of Bissau ECOWAS Economic Community of West African States SIGFIP Sistema Integrado de Gestao Financeira EMI Entraide Médicale Internationale Publica (Integrated Financial Management EMIS Education Management Information System System) ENA Escola Nacional de Administração (National SIGRHAP Sistema Integrado de Gestão de Recursos Public Administration School) Humanos (Integrated Human Resource ESP Education Sector Plan Management System) FCS Fragile and Conflict Affected States SIPRI Stockholm International Peace Research FDI Foreign Direct Investment Institute GDP Gross Domestic Product SOE State-Owned Enterprise GPE Global Partnership for Education SSA Sub-Saharan Africa GST Goods and Services Tax SSR Security Sector Reform HRM Human Resources Management SYGADE Système de Gestion et d'Analyse de la Dette IEC Código do Imposto Especial de Consumo (Debt Management and Analysis System) (Consumer Excise Code) TdC Tribunal de Contas (Court of Auditors) IISS International Institute for Strategic Studies TFP Total Factor Productivity IGAS Inspeção-Geral para Assuntos da Saúde TIN Taxpayer Identification Number (General Inspection for Health Affairs) TOFE Tableau des Operations Financières de l’Etat IMF International Monetary Fund (Table of State Financial Operations) INAFOR Institute for Technical and Vocational Training U5MR Under-Five Mortality Rate INDE National Institute for the Development of UAC University of Amilcar Cabral Education UCCP Central Public Procurement Unit INEP National Institute of Studies and Scientific UNCTAD United Nations Conference on Trade and Research Development LAN Local Area Network UNDP United Nations Development Programme LIC Low Income Country UNESCO United Nations Educational, Scientific and MDG Millennium Development Goal Cultural Organisation MEF Ministry of Economy and Finance WAEMU West African Economic and Monetary Union MFDC Movement of Democratic Forces of Casamance WFP World Food Program iii Executive Summary 1. Guinea-Bissau is a small state in West Africa with a population of around 1.8 million .1 It is rich in natural resources (fisheries, forestry and agriculture) and biodiversity. Economic activity is, however, dominated by the production and sale of unprocessed cashew, which is also the main source of income for more than two thirds of households. Guinea-Bissau is host to a large variety of ethnic groups, languages and religions, with communal and ethnic-based violence remaining low. The country has a history of political and institutional fragility dating back to its independence from Portugal in 1973. Since independence, four successful coups have been recorded, with another 16 coups attempted, plotted, or alleged. Political fragility has been manifesting itself in frequent government turnover. Political instability has been responsible for large drops in output and government expenditure.2 After almost three years of political gridlock, a new consensus government came to power in April 2018. In sum, Guinea-Bissau has been caught in a vicious cycle of poor governance, fragmented elites, weak public sector capacity, and a poorly diversified economy. 2. Guinea-Bissau continues to face formidable development challenges, but resources remain very limited. Government spending on the social sectors is low, which, combined with low efficiency, translated into weak outcomes. In health and education, not only does the country spend less than most Sub-Saharan African countries but also achieves poor outcomes for every dollar spent. In addition, line ministries lack the most basic functions of a responsible and accountable public financial management (PFM) system. The private sector is embryonic, mainly focused around the processing of cashew, while foreign direct investment remains very low given the uncertain policy and regulatory environment and the high cost of doing business in Guinea-Bissau.3 The quantity and quality of infrastructure remains weak and inadequate, particularly in the energy sector. Access to electricity is extremely limited and very costly, at US$0.50 per kWh, making it among the most expensive in the continent (World Bank, 2017). 3. The objective of the Guinea-Bissau Public Expenditure Review (PER) is to analyze government expenditure, fiscal revenue, and public financial management in selected sectors (education, health, and security). The PER is a follow-up to the World Bank’s (2017) Public Expenditure Analysis that provided an overall review of public finances in Guinea-Bissau (see Annex I). It contains a wide range of analyses, with some chapters examining public spending trends and outcomes, while others are more process oriented and place a strong emphasis on PFM systems, at macro- and micro-levels. The education and health chapters go beyond the confines of traditional World Bank PERs—namely the efficiency, effectiveness, and equity of spending. Both of these chapters also review the PFM systems in the respective line ministries with a view to identifying options for reform. Further, the PER analyzes the fiscal implications of continuing to spend over 15 percent of the budget on the security sector and nearly 9 percent of GDP on wage and nonwage compensation. 4. The sectors for analysis were selected based on a combination of factors, including spending trends, budget size, the country’s development program, and World Bank engagement . The PER has been initiated by the World Bank and the team worked closely with the Ministry of Economy and Finance (MEF), and line ministries throughout the process. The PER is well aligned with the FY2018–21 Country Partnership Framework (CPF). The first two objectives of the CPF are “increased access to and quality of 1 See World Bank (2017) Guinea-Bissau Country partnership Framework (CPF) and World Bank (2016) Systematic Country Diagnostic (SCD) for a more detailed discussion on country context and development challenges. 2 Political instability is associated with most periods of recession in Guinea-Bissau. Growth rates systematically fall following instances of political instability. IMF (2015) estimates that GDP per capita could have been 65-90 percent higher had it not been for political fragility. The uncertainty created by political turmoil increases the risk aversion of local entrepreneurs and foreign investors alike. Donors also withdraw their support, as in 2012 and 2016. Since donor support has been vital to finance both public investment, the effect of donor disengagement is considerable. 3 World Bank (2018) Doing Business. 1 primary education” and “increased access to and quality of maternal and health services.” The CPF identifies limited fiscal space and weak PFM systems as being among the key challenges to improving access to quality social services. The PER also supports the fourth pillar of the Government’s National Development Plan, Terra Ranka (Fresh Start), which aims to develop the human capital potential of the population. The three sectors under consideration together account for over 30 percent of government expenditure; thus, any efficiency improvements and/or expenditure savings in those sectors are likely to be substantial. 5. Public spending in Guinea-Bissau is low and volatile, indicating the need for enhancing domestic revenue mobilization through tax policy reforms and further revenue administration measures. Total public spending averaged 19 percent of GDP over 2010-17, although this masks substantial volatility, the bulk of which represents recurrent spending. At below 4 and 2 percent of GDP, Guinea-Bissau’s total public and private investment, respectively, are among the lowest in Africa. Low and volatile government spending partly reflects the fact that tax revenue remains low, at 9-10 percent of GDP, despite improvements in 2017 supported by revenue administration measures. Guinea-Bissau’s tax policy framework has been largely unchanged for decades, with the personal income tax rate, at 12 percent, the lowest among peers. In the medium-term, Guinea-Bissau will have to establish a simple and transparent tax regime, while strengthening tax and customs administration. It is also important to minimize and better target tax exemptions, and recover outstanding tax arrears through stricter enforcement mechanisms. 6. The efficiency of public spending could be enhanced by improving the allocation of limited resources, among others. Social spending in Guinea-Bissau is among the lowest, while security spending is among the highest compared to its economic and regional peers. Security spending averaged 2.5 percent of GDP over 2015-17, while health and education spending stood at 2.2 and 2.5 percent, respectively. The security sector (military and internal) accounts for 37 percent of total public servants, 35 percent of the wage bill, and 15 percent of the national budget; however, actual policing and justice continue to be left to non- state and customary authorities. In addition, while the official wage bill is relatively small (averaging 4.9 percent of GDP during 2010-17), the total compensation of government employees (wage and nonwage benefits) is estimated at nearly 9 percent of GDP. It will thus be essential to rethink the existing allocation of public resources to better align it with national priorities, which includes increased spending on the social sectors and key infrastructure (notably energy). It is also important to break the security sector reform deadlock, and work toward rebalancing security and development priorities. 7. The Government may wish to improve PFM systems, including through further strengthening the operations of the Treasury Committee and the integrated financial management system, SIGFIP. Since late 2016, the Treasury Committee (TC) has been providing a measure of control over cash management and public expenditure. However, overall PFM systems (budget preparation, execution, and control) remain weak while the quality of expenditure is low, with expenditures being driven by immediate needs rather than prioritization of social or economic returns. Limited use of the Government’s financial management information system and a lack of external audits reduce the efficiency of government expenditure and lead to unreliable data. Short-term priorities include preparing a credible budget to guide expenditures and consolidating recent gains in expenditure management and controls through the TC while building formal PFM institutions in the medium-term. Guinea-Bissau should also consider improving the reliability of public finance data by strengthening and expanding the coverage of SIGFIP. 8. Improving education and health outcomes will require increasing the budget share of non- salary spending, in line with the new World Bank Human Capital Project 4, and strengthening PFM 4 The new World Bank Human Capital Project proposes that governments should invest more to deliver better health and education outcomes. The Human Capital Project constitutes ‘a new effort to better understand the link between investing in people and economic gr owth, and to accelerating financing for human capital investments.’ 2 systems in these sectors. Despite increased spending and considerable progress in improving access to basic education, outcomes remain poor with alarmingly low levels of learning and completion rate, and the distribution of resources remaining regressive. In health, Guinea-Bissau’s infant, under-five, and maternal mortality rates are among the highest in the world. Poor education and health outcomes partly reflect the low level of public spending and high out-of-pocket payments. Guinea-Bissau needs to build budgeting and planning capacity, improve budget transparency by making financial information (including the amount and usage of fees collected from health centers and schools) available to the public, and implementing a standard financial management system in the line ministries. It is also imperative that Guinea-Bissau improves expenditure prioritization, while also strengthening formal PFM institutions in the sectors over the medium term gradually allowing them to play their role in budget preparation, execution, and control. 9. The PER addresses the following questions with the Government of Guinea-Bissau and the World Bank as the primary audiences. Chapter 1 analyzes fiscal revenue and public spending performance over 2010–17. Chapter 2 examines the fiscal implications of wage and non-wage compensation. Chapter 3 reviews Guinea-Bissau’s macro PFM systems by focusing on the operations of the Treasury Committee and the Government’s financial management information system, SIGFIP. Chapter 4 assesses PFM systems in the education sector, as well as the efficiency and equity of education spending. Chapter 5 analyzes public spending trends and outcomes in the health sector, and links them with existing PFM systems and practices in the sector. Finally, Chapter 6 provides a rapid assessment of the fiscal affordability of the security sector and sets up a macroeconomic foundation for potential security sector reforms (SSRs). Overview of Guinea-Bissau’s Fiscal Position 10. Guinea-Bissau has recently registered high growth driven by favorable international cashew prices, but tax revenue remains low, despite some improvements in 2017. Real gross domestic product (GDP) growth averaged 6 percent over 2015–17, up from 3.1 percent during 2010–14. At about 10 percent of GDP, tax revenue in Guinea-Bissau is below potential and those for its peers. Tax revenue is dominated by indirect taxes, with sales tax accounting for the largest share of tax revenue, at 3.7 percent of GDP over 2010–17. While trade tax and corporate income tax revenues have increased in recent years, that from personal income tax and excise taxes has remained low and stagnant. Guinea-Bissau’s potential tax collection rate is estimated to be around 18 percent of GDP, indicating that it has scope to further increase tax revenue collection. Its tax-to-GDP ratios are below those of its structural and regional peers across most subcategories of tax (notably personal income tax revenue). Annex II provides definition of peer countries. 11. Guinea-Bissau’s tax policy framework has been largely unchanged for decades . Its personal income tax rate, at 12 percent, is the lowest among all its peers. Similarly, although the statutory corporate income tax (CIT) rate is 25 percent, similar to its peers, the effective CIT rate is only around 15 percent, partly due to exemptions and the application of lower rates in some sectors. Recent increases in domestic revenue are due to improved tax and customs administration. Nonetheless, generous tax exemptions and tax liabilities account for a large portion of foregone revenues, estimated at 2.7 percent of GDP. Successful experiences from other fragile states indicate that revenue administration measures are more effective when they are accompanied by tax policy reforms, such as raising value-added tax (VAT) (or sales tax) rates, increasing excise taxes, broadening the coverage of some taxes, and reducing tax exemptions. 12. Public spending in Guinea-Bissau has fluctuated markedly, largely reflecting political volatility and aid dependence. Overall spending averaged 19 percent of GDP in 2010–17, the bulk of which represents recurrent spending. Spending financed from domestic resources has increased in recent years but only a very small share of this has been channeled to capital spending. At below 4 percent of GDP, 3 Guinea-Bissau’s total public investment is among the lowest in Sub-Saharan Africa (SSA). The share of capital spending allocated to the social sectors has increased recently, perhaps reflecting changes in donor priorities as more than 95 percent of capital spending is foreign funded. The high degree of dependence on external assistance, combined with substantial fluctuation of donor support, remains a cause for concern given that a sustainable development strategy requires domestic ownership, among other factors. 13. Compared to its structural peers, Guinea-Bissau’s government expenditure on the social sectors is among the lowest, while security spending is among the highest. Wages and salaries averaged 5.1 percent of GDP over 2015–17, 36 percent of which is absorbed by the security sector. Spending on goods and services is low compared to peers, with the share of social sectors declining. Guinea-Bissau’s outsized security sector has absorbed 35 percent of total spending on goods and services over 2010 –17. Despite a decline since 2015, security spending accounts for about 16 percent of total government expenditure (or 3 percent of GDP), which is high for a country that faces no significant external threats or active internal conflict. By contrast, education and health together account for 22.3 percent (4.2 percent of GDP) of government expenditure. Transfers and subsidies have increased in recent years, partly due to the materialization of government debt guarantees. Contingent liabilities have proved to be an important source of fiscal risk, with calls on debt guarantees amounting to 2.2 percent of GDP in 2016–17. 14. The following policy recommendations emerge from the analysis. The Government may consider developing a reform strategy combining well-sequenced tax policy and administration measures. In the short term, the Government may want to minimize and narrowly target tax exemptions, recover tax arrears, and introduce excise taxes on alcoholic products. It is also important to continue broadening the tax base by curbing tax evasion and expanding tax payer registration, and strengthening key operations in the tax and customs departments. Over the medium- to long-term, Guinea-Bissau will have to establish a simple and transparent tax regime. In the short term, any reform strategy should take into account the country’s limited institutional capacity. To improve the efficiency of public spending, Guinea-Bissau should rethink its current allocation of public resources. Government may want to identify its spending priorities in the short term to allocate limited public resources more strategically. Public Sector Wage Bill 15. Guinea-Bissau has 4.7 security personnel per 1,000 inhabitants, compared to less than 1 frontline health worker. The total number of civil servants in Guinea-Bissau stood at 20,550 in 2005 and increased by 17 percent to 23,550 in 2017, 37 percent of whom are in the security sector. The actual size of the public sector workforce could be substantially larger given the large number of unofficial employees, many of whom are paid with budgetary resources intended for non-wage spending or ministries’ own revenues. In terms of sectoral distribution, the number of staff in the education sector has increased by 43 percent since 2010, driven by a rise in the number of teachers. While the staff headcount in the defense sector has remained broadly stable (at 2.3 personnel per 1,000 inhabitants or double the regional average of 1.2), the size of staff in the internal security sector has increased by a third between 2010 and 2017. 16. The structure of the civil service has changed significantly since 2005 partly due to retrenchment programs. Guinea-Bissau’s civil service is dominated by technicians, with very few employees at the top and bottom of the scale. Nearly 80 percent of civil servants are mid-to-low ranking technicians and officials, reflecting the impact of a downsizing program that reduced the number of unskilled workers. However, the salaries of many of the dismissed employees (roughly 1,000 public sector workers) continue to be paid by the Government but they are classified under transfers and do not appear on the official wage bill. After a large-scale salary harmonization in 2004 aimed at reducing wage 4 differentials between the top grades and the rest, salary levels were adjusted only once, in 2012, for those at the bottom of the salary scale (up by 58 percent in nominal terms). 17. The official wage bill in Guinea-Bissau is somewhat small relative to GDP and in line with those of peer countries, but it remains far above the WAEMU convergence criteria of 35 percent of tax revenues. The wage bill accounted for 57.5 percent of tax revenues over 2010 –17. According to official figures, wages and salaries have decreased steadily, from 5.9 percent of GDP in 2004 to 4.5 percent in 2017. The share of wages and salaries allocated to the social sectors has grown substantially on the back of higher spending on education. Wages in the education sector have increased from 1.1 percent of GDP in 2010 –13 to 1.7 percent in 2014-17. Although domestic revenues increased from 8 percent of GDP in 2013 to 13 percent in 2017, little of this fiscal space was used for government priorities, including public investment and essential social services. 18. However, official figures do not provide a full picture of total government employment and compensation. Several spending items, including pensions, bonuses, salary top-ups, health benefits, and salaries of staff in autonomous state institutions are classified under different budget lines and therefore not included in official figures. When these items are considered, the total government employee compensation amounts to nearly double the official wage bill, averaging 8.9 percent of GDP over 2010 –17. This puts Guinea-Bissau’s actual wage bill among the highest compared to its structural and regional peers. Bonuses paid to MEF staff to incentivize tax collection (incentivos) have increased rapidly, reaching 1.3 percent of GDP (or 10 percent of total tax revenue) in 2017. However, the contributions of incentivos to raising tax revenue is questionable as they are not linked to individual performance. 19. Scenario analysis indicates that shocks could make the actual wage bill rise above 10 percent of GDP within 3 years, making it unaffordable and raising the need to curb employee compensation. In the absence of measures that materially impact wage levels or employment, Guinea-Bissau’s total actual wage bill is projected to remain at around 8 percent of GDP over the next six years. However, given that tax revenue averaged 9–10 percent of GDP in recent years, shocks (for instance, to fiscal revenue) could put a major strain on public finances. Given Guinea-Bissau’s poorly diversified economy and persistent political instability, the risk of such a shock occurring is not negligible. Alternative wage bill policies (including attrition-based restructuring and resizing of the civil service, and reforming tax collection bonuses) are associated with significant expenditure savings. However, such measures are not proper substitutes for medium-term structural reforms to improve existing wage and employment policies. 20. The wage bill analysis gives rise to the following policy recommendations . A large cohort of public sector employees are expected to retire between 2018 and 2025 which could be a good opportunity for the Government to consider an attrition-based restructuring and resizing of the civil service and address the misallocation of human resources to non-priority sectors. The recent civil service census could be used to carry out a comprehensive functional review to identify overlapping functions and redundancies, and apply hiring freezes until such review is completed. In the medium- to long-term, awarding notch wage increases5 and promotions only to high-performing individuals and limiting non-wage benefits could increase productivity while at the same time having a positive net fiscal effect. Further, the Government may consider completing the public sector organigram, carrying out a census of irregular employees, and preparing multi-annual staffing plans to better control the number of public sector employees. 5 Awarding notch wage rise represents an increase in pay due to workers moving up a level within their grade in the wage table – and promotions. 5 Public Financial Management, the Treasury Committee, and SIGFIP 21. The Government has set out PFM reform plans in its program, but implementation has been limited. Sound PFM system should ensure control, prioritization, accountability and efficiency in the management of public resources. Despite modest progress over 2009-13, the country scores below other fragile states on nearly all PFM measures. Political instability meant that no budget has been approved since 2015 but even before then budget credibility was very limited. The budget prepared by the MEF is used in a very limited way to guide expenditure priorities. There is generally limited information on the financial performance of parastatals, including contingent liabilities and transfers from government. Although efforts have been made recently to improve budget planning and build medium-term forecasting capacity, the gaps between budget appropriations and execution remain large. Investment is almost entirely funded by donors and there are no domestic systems to appraise and prioritize projects. 22. Spending control has recently improved with the operation of the Treasury Committee (TC) but expenditures have been made based on immediate needs rather than prioritization of social or economic returns. After several years of being sidelined or non-existent, the TC has been operating more effectively since the end of 2016, reducing arrears and all but eliminating non-budgeted expenditures (DNTs). The TC makes real-time decisions on expenditure based on cash availability and some ad hoc prioritization criteria. Weak budget planning and forecasting, and lack of capacity mean that line ministries have little influence over spending priorities once immediate expenditure needs such as salaries are met. In addition, there is lack of comprehensiveness and transparency as the budget classification is not fully in line with WAEMU guidelines; it has an additional category, “other current expenditures”, which reached 14 percent of total government expenditure in 2016. External control of government accounts is essentially non-existent. The Court of Auditors last published an opinion on the government’s accounts in 2010, and the National Assembly does not have the capacity to provide budget oversight. 23. The TC’s role goes beyond traditional treasury management and the concentration of decision-making powers in the MEF mean that sectoral priorities like health and education can be overruled by fiscal considerations. Line ministries are no longer authorized to make spending commitments unless previously approved by the committee. The TC monitors cash inflows and estimates cash outflows on a weekly basis to estimate the funds available and authorize payments. It controls cash outflows by defining a ceiling for payments and commitments, and monitors the accumulation of payment arrears. It prioritizes expenditure to ensure the most essential categories of spending such as salaries are authorized first, although in practice some non-essential expenditure (such as foreign travel costs for government officials and revenue collection bonuses) tends to be prioritized and investment is given a low priority. The TC is also hampered by limited information and a lack of forward-looking analysis. 24. The Government’s integrated financial management information system, the Sistema Integrado de Gestao Financeira Publica (SIGFIP), is operational but weaknesses prevent its full use . SIGFIP does not currently provide all the functionality of a government financial management information system. Most expenditures are made outside the system, through a paper-based approval mechanism, and then recorded in SIGFIP after some significant delay, and there is no automatic integration of information from other systems. Financial statements and other reports cannot be directly extracted by the system without further frequent and substantial adjustments, putting the reliability of the budget expenditure into question. SIGFIP is available only through a local area network (LAN), which prevents access to users outside the MEF premises. Users from other ministries must physically go to the MEF either to register their expenditure documents or to simply consult a process. 6 25. In terms of policy recommendations, experience from successful cases suggests that PFM reforms should focus on a realistic number of technically achievable reforms . Guinea-Bissau will need stronger political leadership on its PFM reforms and may want to develop an operational plan for reforms, with clear and measurable indicators to be monitored. It is crucial to build on the progress made by the TC and improve its analytical capabilities, including cash forecasting over longer timeframes and liquidity monitoring, and develop a cash buffer policy to prevent debt defaults. Line ministries should be encouraged to control and prioritize expenditures, and develop a priority matrix to guide expenditure decisions taken in the TC. To improve the quality and reliability of public financial data, the Government may consider upgrading SIGFIP so it can be used for the entire expenditure chain, discontinuing parallel paper processing where possible. It is important to make the system web-accessible so that line ministries can use it, integrating it with other systems, and training national technicians to eliminate dependence on third parties. Education 26. Public education spending has increased over 2010 –17, but Guinea-Bissau spends less than its peers. Education spending stands at 2.2 percent of GDP or 12 percent of total government expenditure over 2010-2017. Public education spending has been volatile, reflecting fluctuations in fiscal revenue and donor support. Over 90 percent of expenditure goes on salaries, leaving investment in classrooms or educational inputs to donors and non-governmental organizations. The education budget remains largely centralized, with resources at the subnational level limited to the collection of school fees from secondary education. The sector is heavily dependent on external sources, but aid management has been undermined by fragmentation, limited project management capacity, and unreliable disbursement. 27. Despite rising spending and improved access to basic education, the education system is characterized by low completion rates and alarmingly low levels of learning . Late entry, and high repetition and dropout rates mean that Guinea-Bissau has a primary education completion rate of 29 percent. A third of children between the ages of 6 and 11 have never attended school. Poor teacher performance and shortage of trained teachers remain critical problems. Spending is tilted towards post- primary education, with primary education receiving less than half of the resources but accounting for about 70 percent of total enrollment in public schools. Although access to primary education is relatively equitable, only 13 percent of students in secondary schools, and just 4 percent of students in higher education, are from the poorest quintile. The regional distribution is also regressive, with the poorest regions receiving lower funding per student and having higher student-teacher ratios than the wealthiest ones. 28. Out-of-pocket payments are believed to be an important source of funding, but little is known about the amount and usage of school fees . The education system has three main sources of funding: public, private, and external funding. Private funding comes largely from out-of-pocket payments for school fees from Grade 7 up, but also some informal payments from students. Despite some functions being devolved to regional directorates, fiscal decentralization lags behind, with the education budget remaining highly centralized. There is no financial reporting on the amount of school fees collected or how they were spent, and little record of what donations have been received by schools. Neither regional directorates nor schools receive any transfers from the central administration. Provisions are made on an ad hoc basis depending on resource availability and lack a strategic planning approach as well as clear distribution criteria. 29. PFM systems in the education sector reflect the wider problems in Guinea-Bissau. Although the Government has recently introduced a medium-term budget planning perspective, the sectoral plan is more of a “fundraising” tool than a planning and budgeting document, and the Ministry of Education (MOE) 7 lacks capacity to participate in budget preparation. The absence of a functional financial management system at the MOE hampers the ministry’s access to real-time financial information that the Administrative and Financial Directorate (DAF) could use to administer programs effectively, formulate budgets, and expedite payment transactions. Budget classification is not aligned with the educational structure, which hampers the tracking of spending by education level. Poor planning, lack of capacity and delayed payments meant the education budget was, on average, underspent by 8 percent over 2010–17. Very few schools have been externally audited. The MOE’s internal control systems are weak, largely due to outdated laws, policies, and missing procedures. Public procurement procedures have generally not been complied with, which may have led to higher costs and poor-quality purchases. 30. Policy recommendations include improving governance and institutional arrangements, building planning and budgeting capacity, and implementing financial management systems in the MOE. The Government may want to update the organic laws of the MOE and develop procedure manuals that document the mandates and responsibilities of central and decentralized entities in the sector. It may also consider documenting the amount and distribution of school fees, and eliminating the practice of making non-budgeted expenditures using non-registered internal revenues. Budget transparency could be improved by making financial information and service delivery performance available to the public. The MOE may want to ensure value for money in public procurement by preparing a procurement procedures manual, with emphasis on open and competitive procedures, monitoring of results, and access to appeal and redress arrangements. In the medium term, it is critical to enhance capacity in the areas of budgeting, human resources, and payroll, and to implement a standard financial management system across central and decentralized entities, connecting them to the centralized SIGFIP system. 31. The Government could improve the efficiency of resource use by better prioritizing expenditures and improving the tracking and monitoring of resources and performance. Increasing the budget share of non-salary school-level spending will be important to improve learning outcomes. Repetition, dropout rates, and delayed entry may be partially addressed through a mandatory enrollment age and automatic promotion at least within the primary education level. Resources need to be better targeted to improve equity in access to education with short-term measures to reduce the number of out- of-school children, for example, through vouchers, cash transfers, and awareness programs. Human resource management could also be strengthened with the implementation of a common personnel database and competency-based training plans. Providing teachers with incentives to teach in remote areas could also help reduce regional inequalities in the medium term. Health 32. Government spending on health makes up only 1.9 percent of GDP, which is significantly lower than some of its structural peers. At 6.9 percent of GDP, total health spending (including donor resources) in Guinea-Bissau is comparable with its peers, but the composition of its spending is problematic. Government health expenditure is not only low but also hugely skewed towards staff costs. Roughly 90 percent of the health budget is spent on personnel costs. Guinea-Bissau relies on out-of-pocket (OOP) payments and external resources to fund health services. Formal and informal OOP payments have become institutionalized and are seen by health workers as a means of supplementing small and irregularly paid salaries. On average, households spend 15 percent of their non-food expenditure on health. High OOP payments expose households to financial shocks linked with higher mortality rates, while dependence on volatile external resources has implications for sustainable health financing. 8 33. Guinea-Bissau’s health outcomes are at or below those of benchmark countries. At 57.4 years, life expectancy is lower than the average for low income countries (LICs) and SSA. Despite some progress, Guinea-Bissau’s infant, under-five, and maternal mortality rates are among the highest in the world. Only 45 percent of births take place in health facilities. Poor health outcomes partly reflect the low level of health spending and high OOP payments. OOP payments have not resulted in improved access to health services. The number of births assisted by a qualified health professional is low in regions with high OOP payments (Gabú and Bafatá). On the supply side, challenges include an acute shortage of qualified health workforce, weak infrastructure, and shortage of medical equipments. On the demand side, there are several barriers, including under-the-table payments, poor perceptions of service quality, and cultural factors. 34. Despite modest improvements in recent years, PFM systems in the health sector remain weak. Guinea-Bissau has recently elaborated the National Plan for Health Improvement (PNDS) and for the first time, in 2017/18, prepared a three-year budget based on this plan. However, budget preparation process has not yet been institutionalized in the Ministry of Public Health (MOH), which lacks the staff and skills to develop a budget. The budgeting process in the MOH is more a “fundraising” tool than a planning and budgeting document. Budget execution (including donor-funded investments) fell as low as 50 percent in 2014 and 42 percent in 2015, although this improved to 70 percent in 2017. The flow of funds to and from decentralized health entities are not clear and not reported in the public accounts, and internal and external audit is practically non-existent. The weak PFM framework has led to limited funds flowing to the health sector, stimulating OOP and informal payments, and undermining service provision. The financial management information systems in the health sector also remain unreliable and represent a constraint to developing an effective PFM system. 35. The Government may consider the following policy recommendations. It is important to strengthen the budget process through building capacity in the MOH, among others. Budget transparency could be improved by making financial information (including the transfer of funds from the MEF to the MOH and decentralized units, as well as the distribution and use of fees collected form health centers) and service delivery performance available to the public. The Government may also want to ensure value for money in public procurement through preparing a procedures manual, with emphasis on open and competitive procedures, monitoring of results, and access to appeal and redress arrangements. Payroll controls could be improved by, among others, procuring SIGRHAP for the MOH and updating the profiles of staff in the Ministry of Public Administration database. The Government could reinforce the procedures applied in internal audit by providing IGAS6 with the necessary budget and operational independence. 36. Improving health service delivery will also require public resource tracking systems and strengthening regulations to eliminate informal payments . The MOH will need to enhance the coordination and provision of services through integrated frontline teams and work with donors to improve coordination of external funding. Strengthening human resource management will require targeted sectoral workforce policies, defining career pathways and revising remuneration policies, and improved procedures to recruit health workers. Further, the MOH may want to play a central role in coordinating donor efforts by identifying national priorities and requesting donors to respond accordingly. Security 37. Guinea-Bissau is one of the most coup-prone and politically unstable countries in the world, and faces several security challenges. Although the armed forces have recently refrained from interfering 6 IGAS is an acronym for Inspeção-Geral para Assuntos da Saúde (General Inspection for Health Affairs). 9 in politics and the 2016 Conakry Agreement has begun to bring some return to stability, relations between the political elite and the military remains fragile. Guinea-Bissau performs badly in international measures of personal security, the rule of law, and combating corruption, and gender based violence is endemic. The long-running separatist movement in Casamance, Senegal, (on Guinea-Bissau’s northern border) is currently in ceasefire, but drug trafficking in the wider region is a cause for concern as it has become a significant transit route to Europe and North America. With high poverty, a large portion of its territory offshore, and the involvement of well-connected elites in trafficking, Guinea-Bissau is particularly vulnerable to this trade. The country also lacks the capacity to tackle international money laundering, and is both a source and a destination for trafficked children. 38. Despite a series of attempts at security sector reform (SSR), there has not been a successful process that addresses the efficiency, effectiveness, governance, and accountability of the security forces. Over the last decade, a number of international partners including ECOWAS, the UN, and the EU have supported national efforts at SSR. In 2006, a national strategy was devised to resize and modernize the sector to reflect the country’s needs, and to provide for veterans from the Liberation War. Implementation of this strategy led to some improvements in salary payment processes and the rehabilitation of some barracks. In 2015, a pensions and gratuity law was approved, to provide momentum for the restructuring and potential downsizing of the army, but the pension fund proposals have been seen as at odds with the extreme poverty of the wider population. 39. Security spending averaged 2.5 percent of GDP over 2015–17, which is high for a country that faces no significant external threats or active internal conflict, as well as compared to its peers. The security sector (military and internal) accounts for 37 percent of total public servants, 35 percent of the wage bill, and 15 percent of the national budget; however, actual policing and justice continue to be left to non-state and customary authorities. Since 2015, personnel expenses have, on average, made up 68.5 percent of total government security spending, while operational expenditure (notably the purchase of goods and services) accounted for 30.5 percent of total spending. Capital expenditure has amounted to nearly nil in recent years. Compared to its structural and regional peers, Guinea-Bissau’s military is among the lowest in absolute numbers of armed force personnel. However, as a share of the labor force, the size of its army is second only to Burundi. 40. Only 14 percent of the army are private soldiers, compared with almost a third of high- ranking or field officers who account for 60 percent of total army wage bill . The security sector is rapidly ageing, with little new recruitment and an inverted pyramid hierarchical structure. Around 25 percent of serving personnel are over the official retirement age of 60, with another 15 percent due to reach retirement age within the next 5 years. Such top-heaviness is a sign of a sector that is not greatly functional or representative of a country’s needs. The police do not have as disproportionate a concentration of senior officers as the army, but non-commissioned officers make up almost half of the police ranks, with constables accounting for just over 15 percent. Salaries in the police force vary considerably between the lowest and highest ranks, and are also supplemented by 20 percent allowances on top of base salaries. 41. Measures that result in the retirement of security personnel in accordance with existing labor law would be associated with substantial savings—a projected reduction of US$46.3 million (roughly 3.5 percent of GDP) in the wage bill over five years. Guinea-Bissau needs to balance its development priorities against its security spending. Assuming all other things remain unchanged and considering fiscal revenue projections, Guinea-Bissau’s security spending will remain only marginally affordable. Breaking the security sector reform deadlock will be essential if Guinea-Bissau is to escape its current “stop-go equilibrium” brought about by periodic political crises. The Government may want to take advantage of the fiscal space from rising revenues to address the outstanding issue of pensions and Liberation War veterans, exploring 10 options for downsizing the security forces and fresh recruitment to rebalance their hierarchical structure. This may be combined with a clear strategy to rebalance its budget away from salaries and aim to modernize and professionalize its security sector institutions. 42. The Government may want to take advantage of the fiscal space from rising revenues to address the outstanding issue of pensions and Liberation War veterans, exploring options for downsizing the security forces and fresh recruitment to rebalance their hierarchical structure. This may be combined with a clear strategy to rebalance its budget away from salaries and aim to modernize and professionalize its security sector institutions. 11 Table of Contents Chapter 1: Overview of Guinea-Bissau’s Fiscal Position ............................................................................16 Recent Macro-Fiscal Developments .............................................................................................................................. 16 Domestic Revenue: Trends and Key Drivers ............................................................................................................... 19 Challenges in revenue mobilization .............................................................................................................................. 24 Public Spending in Guinea-Bissau .................................................................................................................................. 26 Policy Recommendations................................................................................................................................................... 31 Chapter 2: The Public Sector Wage Bill ........................................................................................................34 2.1 Overall Trends in the Wage Bill ....................................................................................................................................... 34 2.2 Civil Service Headcount and Sectoral Composition ................................................................................................ 35 2.3 Civil Service Grade and Pay Structure ........................................................................................................................... 37 2.4 Computing Total Public Sector Compensation ......................................................................................................... 39 2.5 Wage Spending: Challenges and Opportunities ...................................................................................................... 43 2.6 Scenario Analysis: Medium-Term Wage Bill Projections ....................................................................................... 45 2.7 The Fiscal Impact of Alternative Wage Bill Policies ................................................................................................. 45 2.8 Options for Structural Reforms ....................................................................................................................................... 47 Chapter 3: Public Financial Management, the Treasury Committee, and SIGFIP ..................................51 3.1 The Quality of Public Financial Management in Guinea-Bissau ......................................................................... 51 3.2 The Treasury Management Committee ....................................................................................................................... 57 3.3 The Integrated Financial Management Information System: SIGFIP ................................................................ 63 3.4 Policy Recommendations................................................................................................................................................... 67 Chapter 4: Education .......................................................................................................................................71 4.1 Overview of Guinea-Bissau’s Education System ....................................................................................................... 71 4.2 Performance of the Education System ......................................................................................................................... 73 4.3 Education Spending: Trends and Links with Sector Priorities ............................................................................. 78 4.4 Public Financial Management (PFM) in the Education Sector ............................................................................ 83 4.5 Management of Human Resources ............................................................................................................................... 88 Chapter 5. Health .............................................................................................................................................91 5.1 The Health System in Guinea-Bissau ............................................................................................................................. 91 5.2 Health Outcomes: Trends in Guinea-Bissau and Benchmarking ........................................................................ 93 5.3 Health Care Financing ......................................................................................................................................................... 97 5.4 Public Financial Management (PFM) in the Health Sector ................................................................................. 100 Chapter 6: Security Sector ........................................................................................................................... 107 6.1 Historical and Sectoral Context ..................................................................................................................................... 107 6.2 Evolution of Guinea-Bissau’s Security Sector ........................................................................................................ 1223 6.3 Affordability and Sustainability ..................................................................................................................................... 122 6.4 Recommendations .............................................................................................................................................................. 128 Annex I: Policy recommendations of World Bank (2017) Public Expenditure Analysis .................................................. 133 Annex II: Definition of Peer Countries ................................................................................................................................................ 134 Annex III: Lessons from Successful Cases .......................................................................................................................................... 135 Annex IV: Data Tables ................................................................................................................................................................................ 136 Annex V: Staff Numbers per Ministry.................................................................................................................................................. 140 Annex VI: Wage Bill Projection Model ................................................................................................................................................ 141 Annex VII: Public Sector Wage Tables ................................................................................................................................................ 143 Annex VIII. Organigrams of the Security Sector…………………………………………………………………………………………………131 Figure 1.1: Guinea-Bissau: Growth Trends and Drivers ................................................................................................................... 17 12 Figure 1.2: Guinea-Bissau: Selected Economic and Fiscal Trends............................................................................................... 19 Figure 1.3: Guinea-Bissau and Peers: Trends in Doemstic Revenue ............................................................................................ 7 Figure 1.4: Guinea-Bissau and Peers: Tax Rates ................................................................................................................................... 8 Figure 1.5: Guinea-Bissau and Peers: Overall Trends in Public Spending ............................................................................... 11 Figure 1.6: Guinea-Bissau and Peers: Composition of Public Spending .................................................................................. 13 Figure 1.7: Guinea-Bissau: Trends in Public Spending .................................................................................................................... 15 Figure 1.8: Guinea-Bissau: Efficiency Frontier ..................................................................................................................................... 16 Figure 2.1: The Public Wage Bill in Guinea-Bissau ............................................................................................................................ 35 Figure 2.2: Guinea-Bissau: Trends in Public Employment .............................................................................................................. 38 Figure 2.3: Structure and Trends in Public Employment in Guinea-Bissau ............................................................................. 39 Figure 2.4: Trends in the Wage Bill, and Domestic Revenues (% of GDP) .............................................................................. 41 Figure 2.5: Tax Collection, Public Employment, and the Wage Bill ............................................................................................ 42 Figure 2.6: Trends in Compensation and Public Employment ..................................................................................................... 44 Figure 3.1: CPIA Scores on PFM Cluster Indicators, SSA and FCS (2016) ................................................................................ 53 Figure 3.2: Treasury Core Functions: Managing Flows and Stocks ............................................................................................ 61 Figure 3.3: Budget-Treasury Relationship ............................................................................................................................................ 61 Figure 4.1: Primary Completion Rates in Africa and Peer Countries ......................................................................................... 74 Figure 4.2: Retention Rates Among 16-22 Year-olds, by Gender, Area, Wealth Quintile and Region ........................ 74 Figure 4.3: Out-of-School Rate by Age (2014) ................................................................................................................................... 74 Figure 4.4: Out-of-School Rate by Wealth Quintile, Area and Age Group (2013 –14) ....................................................... 74 Figure 4.5: Enrollment Concentraiton Curves by Education Level.............................................................................................. 74 Figure 4.6: Benefit Incidence of Spending, All Levels of Education ........................................................................................... 74 Figure 4.7: Government Subsidy to Education from Current Expenditure by Wealth Quintile ...................................... 74 Figure 4.8: Gender Equality in Education by Wealth Quintile and Education Level ............................................................ 74 Figure 4.9: Government Subsidy to Education by Gender and Wealth Quintile .................................................................. 74 Figure 4.10: Source of Public Spending for Education .................................................................................................................... 74 Figure 4.11: Distribution of total spending by sector ...................................................................................................................... 74 Figure 4.12: Guinea-Bissau and Peers: Education Spending ........................................................................................................ 74 Figure 4.13: Benchmarking Primary Level Outcomes ...................................................................................................................... 74 Figure 4.14: Trends and Required Education Expenditure ............................................................................................................ 74 Figure 4.15: Breakdown of Education Expenditure ........................................................................................................................... 74 Figure 4.16: Guinea-Bissau and Peers: Education Expenditure by Level .................................................................................. 74 Figure 4.17: Current Expenditure by Category (2010-15) .............................................................................................................. 74 Figure 4.18: Total Expenditure by Category (2010-17) ................................................................................................................... 74 Figure 4.19: External Expenditure by Sector ........................................................................................................................................ 74 Figure 4.20: Education Investment by Source .................................................................................................................................... 74 Figure 4.21: Education Expenditure and Student Population by Region ................................................................................ 74 Figure 4.22: Student-Teacher Ratio and Poverty Level by Region ............................................................................................. 74 Figure 4.23: Financial Flows in the Education Sector ....................................................................................................................... 69 Figure 5.1.a: Distribution fo Health Facilities per 1,000 Population ........................................................................................... 92 Figure 5.1.b: Under-Five Mortality Rate and Geographical Density of Facilities .................................................................. 92 Figure 5.2: Evolution of the Health Workforce by Category (1996 –2016) .............................................................................. 92 Figure 5.3: Guinea-Bissau and Peers: Life Expectancy at Birth (1990–2015) .......................................................................... 94 Figure 5.4: Guinea-Bissau and Peers: Maternal Mortality Ratios (1990–2015)...................................................................... 94 Figure 5.5: Guinea-Bissau and Peers: Under-Five and Neonatal Mortality Rates (1990-2015) ...................................... 94 Figure 5.6: Guinea-Bissau and Peers: Immunization Coverage Rates (1990-2015) ............................................................. 94 Figure 5.7: Guinea-Bissau and Peers: Life Expectancy at Birth (1990–2015) .......................................................................... 95 13 Figure 5.8: Guinea-Bissau and Peers: Maternal Mortality Ratios (1990–2015)...................................................................... 95 Figure 5.9: Guinea-Bissau and Peers: Under-Five and Neonatal Mortality Rates (1990-2015) ...................................... 95 Figure 5.10: Guinea-Bissau and Peers: Immunization Coverage Rates (1990-2015) .......................................................... 95 Figure 5.11: Guinea-Bissau and Peers: Health Spending per Capita ......................................................................................... 98 Figure 5.12: Guinea-Bissau and Peers: Government Health Spending (2015), % total health spending ................... 98 Figure 5.13: Guinea-Bissau and Peers: Out-of-Pocket Payments (2015), % total health spending .............................. 98 Figure 5.14: Guinea-Bissau and Peers: Trends in Out-of-Pocket Payments (2000–15) ..................................................... 98 Figure 5.15: Guinea-Bissau and Peers: External Health Spending, % total health spending ........................................... 98 Figure 5.16: Financial Flows in the Health Sector ........................................................................................................................... 103 Figure 6.1: Political Violence and Protest in Guinea-Bissau (1998–2018) ............................................................................ 109 Figure 6.2: Political Instability in Guinea-Bissau and Impact on GDP (1971–2014).......................................................... 109 Figure 6.3: Governance Indicators (2006, 2011, 2016) ................................................................................................................. 109 Figure 6.4: Ibrahim Index of Governance (2012–16) ..................................................................................................................... 109 Figure 6.5: Drug Trade Routes to Western Africa .......................................................................................................................... 113 Figure 6.6: Distribution of Personnel by Sector .............................................................................................................................. 117 Figure 6.7: Distribution of Army Ranks ............................................................................................................................................... 120 Figure 6.8: Army Payroll, by Rank ......................................................................................................................................................... 120 Figure 6.9: Army Personnel, by Age ..................................................................................................................................................... 120 Figure 6.10: Reverse Cumulative Distribution of Army Personnel, by Age (2018) ............................................................ 120 Figure 6.11: Distribution of Police Officers, by Rank ..................................................................................................................... 121 Figure 6.12: Police Payroll, by Rank ..................................................................................................................................................... 121 Figure 6.13: Police, by Age ...................................................................................................................................................................... 121 Figure 6.14: Reverse Cumulative Distribution of Police by Age (2018) ................................................................................. 121 Figure 6.15: Guinea-Bissau Expenditure by Sector (US$) ............................................................................................................ 123 Figure 6.16: Armed Forces Personnel: Regional and Peer Countries .................................................................................... 125 Figure 6.17: Relative Size of Armed Forces: Regional and Peer Countries .......................................................................... 125 Figure 6.18: Total Military Expenditure: Regional and Peer Countries, constant 2015 US$ ......................................... 125 Figure 6.19: Relative Military Expenditure: Regional and Peer Countries, % of GDP ...................................................... 125 Figure 6.20: Relative Military Expenditure, % of Government Spending .............................................................................. 125 Figure 6.21: Guinea-Bissau Security Expenditure: Baseline ........................................................................................................ 127 Figure 6.22: Projected Domestic Revenue versus Security Sector Expenditure ................................................................. 127 Figure 6.23: Cumulative Difference in Personnel and Operations Expenses: Scenarios (2018 –23) ........................... 128 Box 1.1: Tax Potential versus Actual Tax Collection ......................................................................................................................... 23 Box 2.1: The Structure of the Guinea-Bissau Civil Service ............................................................................................................. 36 Box 2.2: Civil Service Structures in Zambia and Cameroon ........................................................................................................... 37 Box 2.3: Update on Biometric Civil Service Census and HR Database Reconciliation ........................................................ 43 Box 3.1: A Short Overview of Recent PFM Reforms ......................................................................................................................... 52 Box 3.2: The WAEMU Budget Classification and Guinea-Bissau ................................................................................................. 56 Box 3.3: Priority Expenditure Categories ............................................................................................................................................... 59 Box 3.4: Factors Affecting the Budget Cycle ....................................................................................................................................... 60 Box 3.5: Treasury Management in the Face of Unrealistic Budgeting ...................................................................................... 62 Box 6.1: International Sanctions ............................................................................................................................................................ 108 Box 6.2: International Comparisons of the Ratio of Supervisors to Subordinates ........................................................... 119 14 Domestic revenue mobilization Continue improving tax and customs administration Minimize and better target tax exemptions. Recover outstanding tax arrears through stricter enforcement mechanisms. Broaden the tax base through increasing the number of registered tax payers. Build administrative and institutional capacity at the tax and customs departments Introduce tax policy reforms Consider revising tax rates, including income and excise tax rates. Government expenditure Rethink the allocation of government spending Allocate greater resources to social spending and infrastructure. Align government spending with national policies or priorities Identify spending priorities to facilitate the strategic allocation of public resources. Strengthen links between the MEF and line ministries to allow greater alignment of government spending with national policies or priorities. 15 Chapter 1: Overview of Guinea-Bissau’s Fiscal Position1 Economic growth remained high in recent years, averaging 6 percent over 2015-2017, largely reflecting strong global cashew prices. Guinea-Bissau’s tax-to-GDP ratio increased from 8.0 percent in 2010 to 10.3 percent in 2017. However, tax collection level remains below potential and those of its structural peers, and its tax policy framework remains largely unchanged from colonial times. Going forward, Guinea-Bissau may want to combine improvements in tax and customs administration with well-sequenced tax policy reforms. Over the medium term, the government should consider raising income and excise tax rates, and establishing a simple and transparent tax regime. On the expenditure side, volatile spending reflects the country’s political instability and aid dependency. Wages (official and unofficial) make up the bulk of expenditure —a significant portion of which is absorbed by the security sector. It will thus be essential to rethink the existing allocation of resources to better reflect Guinea-Bissau’s national priorities and improve the efficiency of spending. 1. The chapter is organized as follows. It first discusses recent macro-fiscal developments in Guinea- Bissau. Section 1.2 analyzes key trends in domestic revenue mobilization and benchmarks Guinea-Bissau against relevant peer countries. Section 1.3 discusses challenges in revenue mobilization. Section 1.4 discusses public spending and its major drivers according to economic and functional classification, including through international benchmarking. Finally, Section 1.5 offers policy recommendations to increase tax revenue and improve the efficiency of public spending, drawing on lessons from successful experiences elsewhere. Recent Macro-Fiscal Developments 2. Guinea-Bissau’s economy is mainly agricultural and almost entirely dependent on a single cash crop—cashew. Economic activity is dominated by the production and sale of unprocessed cashew. Cashew nuts account for more than 95 percent of the country’s total export revenue. Unlike most economies in Sub-Saharan Africa (SSA), the structure of its economy has barely changed over the past two decades. Agriculture represents more than 50 percent of gross domestic product (GDP) while industry accounts for 15 percent, with services making up the remainder. This starkly contrasts with the wider African experience where the services sector has recently come to dominate economic activity, while agriculture, though still important, commands a relatively smaller share. 3. Growth has remained low and volatile for decades, keeping the country stuck at low-income levels (Figure 1.1.b). Guinea-Bissau’s income per capita grew, on average, at less than 1 percent per annum over 2000–17. On the demand side, private and government consumption were the major drivers, with a small contribution from investment. Net exports made a negative contribution, reflecting less diversified export portfolios and vulnerabilities to adverse terms of trade developments. On the supply side, agriculture remains the main engine of growth. While labor accumulation and total factor productivity (TFP) accounted for the bulk of growth, lower capital accumulation reduced it (Figure 1.1.e). Stabilization policies and external factors drove growth, while structural factors (notably financial disintermediation) reduced it (Figure 1.1.e).2 1 Prepared by Fiseha H. Gebregziabher with contirbutions from Jorge Retana de la Peza (Consultant, GMTA2) and Melanie Marie Laloum (Consultant, GMTA2). This Chapter partly draws upon the World Bank (2017) Public Expenditure Analysis. 2 The model explains growth using three sets of variables: structural factors (infrastructure, financial intermediation, trade, education, government size, institutions), stabilization policies (inflation, exchange rate), and external conditions (terms of trade, commodity prices). 16 Figure 1.1: Guinea-Bissau: Growth Trends and Drivers a. Real GDP Growth in Guinea-Bissau, WAEMU, and Sub- b. Real GDP Growth and Growth Volatility (2000–17) Saharan Africa (2005–17) 10 14% CAR Low, volatile High, less 8 12% Standard deviation of growth YEM growth volatile growth 10% SLE 6 PSE ZWE 8% ARM 4 UKR Percent AFG 6% MNG 2 4% GNB 0 MMR 2% 2008 2013 2005 2006 2007 2009 2010 2011 2012 2014 2015 2016 2017(e) ETH -2 0% Guinea-Bissau WAEMU SSA -2% 0% 2% 4% 6% 8% 10% 12% -4 Average real GDP growth Source: MEF and World Bank. Source: MEF and World Bank. c. Supply-Side Growth Decomposition (2000–17) d. Demand-Side Growth Decomposition (2000–17) 10 6 8 4 6 Percentage points 4 2 Percent 2 0 0 2007 2005 2006 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017(e) -2 2000-2016 2000-2005 2006-2010 2011-2017 -4 -2 Private consumption Government consumption -6 Agriculture Industry Investment Net Exports Services Real GDP growth Real GDP growth Source: MEF and World Bank. Source: MEF. e. Growth Determinants (2000–16) f. Growth Accounting (2000–17) SSA 5 Togo Senegal 4 Niger Mali 3 Guinea-Bissau Guinea 2 Percent Gambia Cote d'ivoire 1 Cabo Verde Burkina Faso 0 Benin 2000-2017 2000-2005 2006-2010 2011-2017 -1.0 0.0 1.0 2.0 3.0 -1 Percentage points -2 Capital stock Labor Structural factors Stabilization policies TFP Real GDP growth External conditions Residual (Unexplained growth) Source: World Bank staff estimates. Source: World Bank staff estimates. 4. However, growth rebounded strongly during the last few years, driven by favorable international prices and good cashew production. Real GDP growth averaged about 6 percent in 2015– 17 17, compared to 2.8 percent in 2000–14 (Figure 1.1.a). Cashew exports surged from 175,000 tons in 2015 to about 195,000 tons in 2016, and are estimated at 175,000 tons in 2017. The decline in production in 2017 was more-than-offset by a sharp increase in international prices. Higher cashew prices have led to a rapid growth in construction activity and consumption. Inflation held steady at below 2 percent, largely reflecting low global fuel and food prices, and a relatively prudent macroeconomic management. 5. The external current account surplus has narrowed, from about 2 percent of GDP in 2015–16 to 0.5 percent in 2017, as other exports remained subdued. The swing in the current account balance from a deficit of 5 percent of GDP in 2013 to a surplus in 2014 –17 was driven by high international cashew prices and lower prices of imported food and oil products (Figure 1.2.b). This was in spite of a sharp decline in donor grants, reflecting the withdrawal of support by some donors to force a resolution of the political instability and the reversal of the 2015 illegal bailout of two commercial banks by government. Total budget and project support plunged from 10 percent of GDP in 2014 to 4 percent in 2017. 6. Improved revenue mobilization and tighter control of spending helped reduce the fiscal deficit, from 4.3 percent of GDP in 2016 to 1.5 percent in 2017. Total expenditure declined from 22 percent of GDP in 2014–16 to about 20 percent in 2017, mainly reflecting a decline in transfers. The authorities have managed to align outlays with available resources, with no new non-concessional external borrowing and no new payment arrears (external and domestic) in 2017 . Measures to improve tax and customs administration, and broaden the domestic revenue base helped boost tax revenue.3 However, contingent liabilities have proved to be an important source of fiscal risk, with calls on government debt guarantees in 2016–17 amounting to 2.2 percent of GDP, leaving fewer resources for other spending. 7. However, fluctuations in international cashew prices continue to have significant growth and fiscal implications given the country’s high export concentration in cashew. Cashew receipts are a strong determinant of the current account balance and they are a main provider of foreign exchange (Figure 1.2.b). Cashew exports are also an important source of fiscal revenue, with cashew export taxes accounting for about 9 percent of tax revenue. (Figure 1.2.a). Overperformance in government revenue due to favorable cashew prices in the early 2010s and 2016–17 led to increases in government spending, whereas the opposite happened when cashew prices plunged in 2013-15. Global cashew exports have increased substantially over the past decade, reducing both prices and the terms of trade for Guinea-Bissau. Guinea- Bissau maintains the second least diversified export portfolio in SSA. Agricultural diversification, through moving up the cashew value chain and capitalizing on non-traditional crops, will be critical to reducing its vulnerability to external headwinds. 8. Fiscal policy has long been undermined by political instability through its impacts on trade, donor flows, and large swings in the proceeds from the sale of fishing licenses . Donor grants and EU fishing compensation account for about 25 and 10 percent of domestic revenue, respectively. Total budget and project support fell from 12 percent of GDP in 2014 to 6 percent in 2017 due to the suspension of aid by some donors in the context of the political turmoil. This contributed to the sharp decline in total government expenditure from 22.9 percent of GDP in 2015 to 16.8 percent in 2017. By contrast, an unprecedented donor support that followed the coming to power of a democratically elected government in 2014 led to an upsurge in public spending, although this was reversed in the wake of renewed political instability in 2015. The uncertainty that political turmoil creates also increases the risk aversion of local 3 In 2017, the authorities embarked on several policy measures to broaden the domestic revenue base, including the assignment of taxpayer identification numbers (TINs) and the rollout of a new sales tax invoice. 18 entrepreneurs and foreign investors alike. IMF (2015) estimates that Guinea-Bissau’s GDP per capita could have been 65-90 percent higher had it not been for political fragility. Figure 1.2: Guinea-Bissau: Selected Economic and Fiscal Trends a. Fiscal Balance and International Cashew Prices b. Cashew Prices and the External Current Account (2010–17) Balance (2011–2017) 30 2000 10% 2000 25 5% 20 1500 1500 Cashew prices(US$ per ton) 15 0% US$ per ton 1000 10 1000 Percent -5% 5 500 0 500 -10% 2010 2011 2012 2013 2014 2015 2016 2017(e) -5 -15% 0 Net Exports of goods and services -10 0 Current account Total expenditures Total revenues and grants Inflation Fiscal balance Cashew price (rhs) Cashew price (rhs) Source: MEF. Source: MEF. Domestic Revenue: Trends and Key Drivers 9. Given Guinea-Bissau’s massive development needs, it will have to enhance domestic revenue mobilization. It has one of the lowest levels of investment among fragile and WAEMU peers, while the quantity and quality of its key infrastructure and human capital remains low. It will thus be critical to create fiscal space through improving revenue mobilization. The country also needs to boost revenues because donor support has historically been volatile as evidenced by the marked reduction in grants in recent years. Weaknesses in tax collection led to reliance on taxing cashew exports. Cashew-specific taxes and fees directly generate about 10 percent of domestic revenue, exacerbating the vulnerability of its fiscal revenue to fluctuations in global prices. In addition to the direct fiscal impact, a sustained increase in revenue could help foster social stability in a conflict-affected country like Guinea-Bissau. 10. Despite improvements in recent years, tax revenue remains below potential and is dominated by sales tax revenue. Guinea-Bissau’s tax-to-GDP ratio increased from 8.0 percent in 2010 to 9.2 percent in 2015 and peaked at 10.3 percent in 2017 (Figure 1.3.a). However, the increases in 2015–16 largely represent one-off revenues. Tax revenues account for the bulk of total domestic revenue. Tax revenue comprises direct and indirect taxes with about 70 percent of tax collection stemming from indirect taxation (Figure 1.3.c). Sales tax represents the largest share of tax revenue and increased from 2.6 percent of GDP to 3.8 percent between 2010 and 2017. Trade tax and corporate income tax (CIT) also increased. In contrast, personal income tax and excise tax remained low and stagnant. Altogether, Guinea-Bissau performed below potential in tax revenue collection, indicating it has scope to enhance its revenue. 11. Guinea-Bissau lags behind its peers and regional comparators in tax revenue collection. Guinea-Bissau is benchmarked against some countries identified as the country’s structural peers using the 19 World Bank’s “Find Your Friends” tool.4 These are Burundi, the Central African Republic (CAR), the Comoros, The Gambia, Sierra Leone, and Tajikistan.5 Where appropriate and possible, the country is also compared with its WAEMU peers. See Annex 0B for the definition of peer countries. Guinea-Bissau’s tax-to-GDP ratio is below the level of its structural peers, except Sierra Leone and the CAR (Figure 1.3.b). Its tax revenue (in percent of GDP) also falls short of the averages for WAEMU countries and fragile states by about 6.2 percentage points (ppts) and 3.7 ppts, respectively. Guinea-Bissau’s tax-to-GDP ratio also trails far behind countries with similar per capita GDP (Figure 1.4.c). Similarly, Guinea-Bissau’s tax-to-GDP ratio is below the minimum ratio of 13 percent (a “tipping point”) that is associated with growth acceleration.6 12. Similar to its structural peers and regional comparators, Guinea-Bissau’s tax revenue is dominated by indirect taxes, symptomatic of a less equitable tax system . Guinea-Bissau has a lower proportion of indirect taxes compared to its peers, and even WAEMU countries, which have substantially higher tax-to-GDP ratios (Figure 1.3.d). However, a relatively high share of indirect taxes has equity implications given that they are essentially levied equally on tax payers, without due consideration of their income. While it is difficult to make conclusive suggestion from international comparisons regarding the appropriate mix of direct versus indirect taxes, there is a widespread consensus that the composition of revenues should shift from consumption to personal income taxes with economic development. 13. Guinea-Bissau’s tax ratios are below its peers across most subcategories of tax revenue. Sales and trade taxes account for the bulk of tax revenue in Guinea-Bissau. Over 2010–17, sales tax revenue averaged 2.7 percent of GDP, higher only than that of the Central African Republic. Taxes on international trade amounted to 2.3 percent of GDP over the same period, which is higher than those for some of its peers.7 However, its personal income tax (PIT) revenue, at 0.6 percent of GDP, is the lowest amongst its structural peers (Figure 1.3.f). For the same period, CIT revenue averaged 1.6 percent of GDP, lower than those for most of its fragile and WAEMU peers. Guinea-Bissau also had the lowest excise tax collection compared to peers, at 0.7 percent of GDP in 2017 (Figure 1.5.e). 14. CIT revenue has slightly increased in recent years, despite a very small tax base, while revenue from PIT remained low and stagnant (Figure 1.3.e). CIT revenue steadily increased from about 1.5 percent of GDP in 2015 to nearly 2 percent in 2016-2017, primarily due to an increase in the corporate tax rate from 15 percent to 25 percent and the retention of taxes at source in certain areas. Corporate taxes are collected on presumed profits due to lack of financial statements. Less than 5 percent of the businesses in the taxpayer’s registry maintain financial statements and about 80–90 percent of CIT receipts come from large taxpayers.8 PIT is collected by employers without the assignment of a tax registration number which undermines control of tax obligations and leaves room for non-compliance and rent seeking.9 4 Guinea-Bissau’s structural peers are countries that share similar characteristics: a population of less than 10 million, real GDP per capita below US$600, high dependence on agriculture, below-average life expectancy, below-average trade with the world, and below-average fiscal revenue. 5 As data may be missing for some of these countries, not all of them will be included throughout the benchmarking exercise. 6 See Gasper et al. (2016). The tipping point estimated in IMF Working Paper – Gasper, Jaramillo, and Wingender (2016) – is a minimum tax-to- GDP ratio of about 13 percent that would enable a country to adequately spend on some of its important development programs. Although this result has been shown to be statistically significant across a panel of countries, the precise number is not a hard-and-fast rule and should be interpreted with caution as it may vary country by country. 7 However, Guinea-Bissau’s trade tax revenue is significantly lower than the average of 3.8 percent of GDP in other WAEMU countries and 3.5 percent in other fragile states. 8 World Bank (2017) Public Expenditure Analysis. 9 Property taxes are almost non-existent and are regulated by legislation dating back to 1920 and 1966. Currently, owners who rent out houses or apartments are taxed 15–18 percent on the rental income, but coverage is very low. 20 Figure 1.3: Guinea-Bissau and Peers: Trends in Domestic Revenue a. Guinea-Bissau: Domestic Revenue (2010–17) b. Benchmarking: Tax Versus Non-tax Revenue 25 35 Tax revenue Non tax revenue 30 Tax revenue 20 25 Non-tax revenue 20 Percent of GDP Percent of GDP 15 15 10 10 5 5 0 0 2010 2011 2012 2013 2014 2015 2016 2017 Source: MEF. Source: MEF, World Bank, and IMF. c. Guinea-Bissau: Direct Versus Indirect Tax d. Benchmarking: Direct Versus Indirect Taxes (2010–17) 12 20 Indirect tax Direct tax 10 15 Direct tax Indirect tax Percent of GDP 8 10 Percent of GDP 6 5 4 0 2 0 2010 2011 2012 2013 2014 2015 2016 2017 Source: MEF, World Bank, and IMF. Source: MEF. e. Guinea-Bissau: Composition of Tax Revenue f. Benchmarking: Composition of tax revenue (2010- (2010–17) 17) 15 20 PIT CIT Corporate Income Tax Personal Income Tax Trade tax Excise tax Sales tax Trade tax 15 Sales tax Other tax 10 Percent of GDP 10 Percent of GDP 5 5 0 0 2010 2011 2012 2013 2014 2015 2016 2017 Source: MEF. Source: MEF, World Bank, and IMF. 15. Trade tax is an important source of revenue with import taxes taking the lion’s share . Taxes levied on imports of raw material and finished goods, and cashew exports represent a quarter of total tax revenue. The effective import tax rate10 rose from 8.0 percent in 2010 to 9.2 percent in 2017, largely due to 10 The effective import tax rate is a proxy for the efficiency of tax collection and is calculated by dividing import tax revenue by import value (free-on-board; fob). 21 stricter controls over fuel imports and rationalization of fuel tax exemptions. Cashew export taxes increased from 0.5 percent of GDP in 2010 to 0.9 percent (7.5 percent of total domestic revenue) in 2017, reflecting favorable cashew prices and increased export volume.11 A cumulative levy of 11.5 percent. is applied on cashew exports at the prevailing domestic reference price. Although international experience with cashew export taxes is mixed, many cashew-exporting SSA countries do not have a special tax on cashews given it may have distortionary effect on trade and income of farmers. Figure 1.4: Guinea-Bissau and Peers: Tax Rates a. Benchmarking: PIT and CIT Tax Rates b. Excise Rates in WAEMU Countries 70 60% Excise-beer 60 PIT CIT 50% Excise-cigarettes 50 40% 40 Percent 30% 30 20% 20 10% 10 0% 0 Source: MEF, World Bank, and IMF. Source: MEF, World Bank, and IMF. c. Benchmarking: Tax Revenue and Income per Capita (Average 2014–16) 40 35 30 Tax revenues (% GDP) 25 20 WAEMU 15 Fragile states 10 Guinea-Bissau 5 0 0 20 40 60 80 100 GDP per capita (PPP, US$ '000) Source: World Bank staff calculations. 16. Taxation on goods and services accounts for about 30 percent of tax revenues, although there is room for improvement given weak administrative capacity. The sales tax base is concentrated on a few formal activities in urban areas. Most small and medium-sized companies have poor book-keeping capacity which hampers enforcement of tax compliance and encourages the underpayment of taxes. The Revenue Authority’s limited audit capacity has further exacerbated the problem. Some progress was made in 2017 to broaden the domestic revenue base, including through the assignment of taxpayer identification numbers (TINs) and the rollout of a new sales tax invoice. However, more remains to be done to improve tax and customs administration and enforcement mechanisms. 11 This was also supported by strengthened border control measures which somewhat reduced smuggling of cashew nuts to Senegal. 22 17. Non-tax revenues are low and volatile, driven by revenues generated from fishing licenses. Fishing licenses, on average, account for 65 percent of non-tax revenues, while EU fishing compensation accounts for 28 percent.12 Non-tax revenues declined sharply, from about 26 percent of total domestic revenue in 2015 to 19 percent in 2017, mainly due to the non-disbursement of EU fishing compensation (about CFAF 4 billion or 0.6 percent of GDP) reflecting protracted disagreements over license fees. Revenue from fishing licenses is low compared to other neighboring coastal countries such as Senegal and Mauritania.13 Fishing remains an area of untapped potential, with license fees either not enforced or not reflecting the real catch capacity of foreign vessels (small or large) entering Guinea-Bissau’s waters. Other sources of non-tax revenues are the sale of timber and tourism revenues. Timber sales did not materialize in 2017 due to a CITES14 ban related to illegal logging, which was however lifted in January 2018. 15 Box 1.1: Tax Potential versus Actual Tax Collection Panel econometric analysis was carried out to estimate the tax potential in Guinea-Bissau and benchmark countries. The analysis uses a stochastic time-invariant tax frontier approach. The model regresses the tax-to- GDP ratio of a sample low-income countries on variables that proxy for possible tax bases and other factors that could affect a country’s ability to raise tax revenues. The analysis included several variables: (i) GDP per capita in PPP; (ii) agriculture-to-GDP ratio; (iii) trade openness measured by import to GDP; (iv) control of corruption; (v) political stability; (vi) age dependency ratio; and (vii) share of natural resources rents in GDP. Figure A: Tax Potential and Tax Effort (2006–15) Figure B: Tax Potential and Actual Performance 20 60 25 Actual Potential GAP 50 20 Percent of tax potential 15 40 15 Percent of GDP Percent of GDP 10 30 10 20 5 5 10 0 0 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Actual Tax gap Tax effort (lfs) Source: MEF and World Bank staff calculation. Source: MEF and World Bank staff calculation 18. Tax rates in Guinea-Bissau are generally at the lower end compared to its regional peers (Figures 1.4.a and 1.4.b). The PIT rate in Guinea-Bissau, at 12 percent, is the lowest in the WAEMU region. The country’s CIT rate is more comparable with those of WAEMU peers. The WAEMU recently introduced a common definition of the corporate tax base within the range of 25 to 30 percent. However, despite Guinea- Bissau’s 25 percent statutory CIT rate, the effective CIT rate is roughly only 15 percent16, partly due to 12 The latter represents payments by the EU to the fishery sector as part of its Fisheries Partnership Agreement concluded between the EU and Guinea-Bissau; other funds under the Agreement are used to purchase fishing licenses. 13 World Bank (2015) Country Economic Memorandum. 14 CITES stands for the Convention on International Trade in Endangered Species of Wild Fauna and Flora. 15 The CITES (the Convention on International Trade in Endangered Species of Wild Fauna and Flora) ban was imposed because government failed to put in place national measures necessary for implementing the obligations required under the Convention. 16 World Bank (2017) Public Expenditure Analysis (PEA). 23 exemptions and the application of lower rates in some sectors.17 However, firms are much more likely to identify tax rates and tax administration as a major obstacle to doing business in Guinea-Bissau than the WAEMU average.18 Excise tax rates on non-alcoholic beverages and beer are largely consistent with other WAEMU countries but excise duty on cigarettes is the lowest in the region. 19. Despite recent improvements, tax collection in Guinea-Bissau remains far below potential, which is roughly estimated at about 18 percent of GDP. A panel econometric analysis for low-income countries (Box 1.1) indicates that the tax gap (difference between tax potential and actual tax collection) for Guinea-Bissau stands at over 7 percent of GDP. This is one of the highest tax gaps among its structural peers, only slightly lower than Sierra Leone and the Central African Republic. Its tax effort – the actual tax rate as a percent of estimated tax collection – has increased from 35 to 57 percent between 2006 and 2017 thanks to higher tax collection in recent years.19 However, there is a large scope for further increase tax revenue. Tax buoyancy20 is also low for all tax categories compared to structural and regional peers. 21 20. The experiences of other fragile states indicate the importance of tax policy reforms, combined with measures to improve revenue administration . Guinea-Bissau’s recent increase in domestic revenue has been largely due to improved tax administration and tighter customs controls. However, successful examples from fragile states elsewhere suggest that revenue administration measures are more effective when accompanied by tax policy reforms. The fragile states that managed to significantly increase tax revenue did so through a combination of revenue administration and policy reform measures. These include raising VAT (or sales tax) rates, increasing excise taxes, reducing exemptions, improving tax compliance, recovering tax arrears, and computerizing tax and customs administration. Annex III details the measures behind successful examples in three SSA countries, Burkina Faso, Liberia, and Rwanda.22 Challenges in revenue mobilization 21. Guinea-Bissau’s tax policy framework has been largely unchanged for decades and there is room to boost revenue mobilization through policy reforms . The country has made few changes to the tax system adopted during the colonial era, in the early 1960s (IMF, 2018).23 There were some attempts to introduce tax policy reforms in the 1980s and 1990s. The income tax rate was revised in the 1980s, although it remains very low. In the 1990s, Guinea-Bissau introduced a general goods and services tax, and excise duty. Despite some changes to various taxes over the past two decades; however, the core structure of the tax system has remained for the most part unchanged. 22. Generous tax exemptions and tax liabilities account for a large portion of foregone revenues, estimated at around 2.7 percent of GDP.24 The Government has granted tax exemptions to the diplomatic corps and other international organizations and the investment law exempts new foreign direct investment 17 For example, Guinea Bissau charges a lower CIT rate for companies who produce and sells mineral on the local market relative to companies which export. 18 World Bank (2018) Doing Business. 19 This represented an improvement from a low level but remains significantly lower than its fragile peers such as The Gambia (87 percent) and Burundi (83 percent). 20 Tax buoyancy measures the efficiency and responsiveness of revenue mobilization to real GDP growth or national income. A tax would be considered buoyant if the increase in revenues is proportionally more than a rise in national income or output. 21 IMF (2015) estimates Guinea-Bissau’s tax buoyancy at 0.73, below WAEMU (0.76) and Sub-Saharan African fragile states (0.80).21 22 Note that, although some other SSA countries have achieved substantial increases in tax revenue, they are more advanced than Guinea-Bissau or they have very different economic structures and political contexts. 23 Ministry of Economy and Finance (2015). “Brief overview of the main taxes collected by the Directorate-General of Taxes and Duties (DGCI).” 24 IMF (2017). Information collected during IMF mission on Guinea-Bissau’s fiscal system estimated foregone annual revenues from tax and customs exemptions at 2.7 percent of GDP (1.7 percent in the form of customs revenue and 1 percent in tax revenue). 24 (FDI) from import taxes. In 2012, the authorities established a tax exemptions committee at the MEF but its capacity to verify compliance with tax obligations and enforce penalties in case of non-compliance remains weak. Similarly, tax arrears remain high, at 1.4 percent of GDP in 2016, the bulk of which was accounted for by state enterprises. Efforts to redeem tax arrears and streamline tax exemptions have been hampered by vested interests in the business and political establishment. 23. Despite some improvements in 2017, there remain substantial capacity constraints in the customs and tax departments. The customs department is seriously affected by governance weaknesses, with the basic functions of customs clearance not well managed or executed. Inefficiencies in the tax and customs administration (poor filing, registration and staffing) are a challenge. Unstable and uncertain customs procedures, often executed by customs agents who received none of the formal training they needed, and weak information systems limit the effectiveness of revenue collection. In addition to these capacity constraints, the tax department has been exposed to political fragility and patronage. Neither the tax or customs administration is connected to the Integrated Financial Management System (SIGFIP), which impedes revenue forecasts and proper functioning of the Treasury Single Account (TSA). Weaknesses in both departments, combined with the perception of rampant corruption, have discouraged tax compliance. Figure 1.5: Guinea-Bissau and Peers: Overall Trends in Public Spending a. Guinea-Bissau: Domestic Versus Externally Financed b. Benchmarking: Overall spending vs. GDP per capita Public Expenditure (2010–17) 30 35 1000 30 800 25 GDP per captia 20 20 600 Percent of GDP Percent of GDP 15 400 10 10 200 5 0 0 0 2010 2011 2012 2013 2014 2015 2016 2017 Domestic expenditures External expenditures Total public spending GDP per capita Total public expenditures Source: MEF. Source: MEF and World Bank. c. Guinea-Bissau: Capital Versus Current Expenditure d. Benchmarking: Capital Versus Current Expenditure (2010–17) 20 40 1.2 Capital Expenditures 1.0 Capital to recurrent ratio Current expenditures 30 15 0.8 20 0.6 Percent Percent of GDP 0.4 10 10 0.2 0 0.0 5 0 2010 2011 2012 2013 2014 2015 2016 2017 Current Capital Capital to recurrent ratio (rhs) Source: MEF. Source: MEF and World Bank. 25 Public Spending in Guinea-Bissau 1.4.1 Overall Trends in Public Spending 26. Total public spending has fluctuated markedly, largely reflecting political volatility and aid dependence. Overall spending averaged 19.4 percent of GDP between 2010 and 2017 (Figure 1.5.a). Public spending declined between 2010 and 2013, mainly attributed to the 2012 military coup that led to the immediate suspension of donor aid. Following the return of political stability in 2014, albeit short-lived, and the ensuing increase in donor-funded investment, public spending increased substantially. However, the sharp recovery was reversed in 2015, with capital spending bearing the brunt of the decline in government expenditure (Figure 1.5.c). Guinea-Bissau’s level of public spending is lower than peers with similar GDP per capita (Figure 1.5.d). Only half of public spending is covered by the government’s own resources. 27. The gap between recurrent and development spending continues to widen . The structure of expenditure has been dominated by recurrent spending, which accounts for the bulk of total public expenditure, averaging 13.7 percent of GDP over 2015–17 (Figure 1.5.c). Recurrent spending accounted for about 64 percent of total expenditure, leaving 35 percent for capital spending (domestic and external). Domestically financed capital spending represents less than 1 percent of GDP. The ratio of capital to recurrent spending declined, from 0.7 in 2010 to 0.4 in 2017, reflecting slower growth in capital spending. Compared to its peers, Guinea-Bissau’s recurrent-to-capital spending ratio is at the lower end (Figure 1.5d). 28. The share of capital spending allocated to the social sectors has increased as resources have shifted away from the economic sectors, signaling changing donor priorities . Investment in the social sectors grew from 35.8 percent of total investment in 2010–13 to 44.6 percent in 2014–17. Health was the main beneficiary of this rise, with capital spending increasing by 0.9 percent of GDP. By contrast, capital spending in the economic sectors (notably agriculture and rural development) has fallen. Given that 95 percent of capital spending is foreign funded, this rebalancing is likely to reflect changes in donor priorities. 29. Higher domestic spending in 2014–17 was primarily due to an increase in recurrent spending with only a small share allocated to investment (Figure 1.5.c). Spending financed by domestic resources increased from 11 percent of GDP in 2010–13 to 15 percent in 2014–17, while externally funded spending grew only slightly from 6 to 7 percent of GDP during the same period. However, the rise in domestic expenditure was almost entirely used to finance an increase in recurrent spending. This trend, combined with the substantial fluctuation in donor grants, raises concerns as a sustainable development strategy will require significant domestically financed investment and domestic ownership. 1.4.2 Composition of Government Expenditure Economic Classification 30. Guinea-Bissau’s wage bill is the largest spending item and has slightly increased in recent years (Figure 1.6.a). The wage bill rose from an average of 4.7 percent of GDP in 2010 –14 to 5.1 percent in 2015–17. However, the wage bill fell from 43 to 37 percent as a share of total recurrent expenditure over the same period, reflecting faster growth in overall recurrent spending, with a larger share allocated to transfers and ‘other current expenditures’. The increase in the wage bill was mostly due to the education sector; the education wage bill increased from 1.1 to 1.8 percent of GDP between 2010 and 2017. Salary spending in the defense and internal security sectors remain high at 2.9 percent of GDP, consuming 36.7 percent of the total wage bill over 2010–17, on average. 26 Figure 1.6: Guinea-Bissau and Peers: Composition of Public Spending a. Guinea-Bissau: Public Spending by Economic b. Benchmarking: Public Spending by Economic Classification (2010–17) Classification 16 25 Other current Interest 14 Transfers Goods & services 12 20 Wages and salaries 10 Percent of GDP Percent of GDP 8 15 6 4 10 2 5 0 2010 2011 2012 2013 2014 2015 2016 2017 0 Comoros Togo Republic Leone Gambia Mali Niger Guinea- Sierra Bissau Central African Interest payments Other current expenditure Transfers Goods & services Wages and salaries Source: MEF. Source: MEF, World Bank, and IMF. c. Guinea-Bissau: Staff Costs (Percent of Total d. Guinea-Bissau: Spending on Goods and Current Expenditure) Services Versus Total Current Expenditure 16 14 12 Percent of GDP Non-staff 10 current Wages and expenditures salaries 8 32.8% 41% 6 4 Rep. dos serviços, 2.5% 2 Embassy Pensions 0 9.6% 2011 2010 2012 2013 2014 2015 2016 2017 Salaries, 2.2% NA Salaries Incentivos, 3.6% 8.8% Current expenditures Goods & Services Source: MEF. Source: MEF. e. Guinea-Bissau: Public Spending by Functional f. Benchmarking: Public Spending by Functional Classification (2010–16) Classification 100% 35 30 80% 25 Percent of GDP 60% 20 15 40% 10 20% 5 0% 0 2015 2010 2011 2012 2013 2014 2016 General Adm. and Finance Defense, Order and Security Admin. and Dev. of Infrast. Health and Social Action Teaching, Train. and Research Sovereignty Services Public order & safety Health Defense Production and Commerce Culture, Sports and Leisure Education Others General admin Source: MEF. Source: MEF, World Bank, and IMF. 27 31. Spending on goods and services nearly doubled during 2010–17, but the share allocated to key social sectors declined, raising concerns about the sustainability of service delivery (Figure 1.6.d). Purchases of goods and services grew rapidly, from 1.9 percent of GDP in 2010 to 2.8 percent in 2017. However, while defense and internal security, on average, accounted for 35 percent of the total spending on goods and services, the ministries of health and education represented only 4.2 and 3.0 percent, respectively. In fact, the share of spending on goods and services in education fell sharply, from an average of 4.3 percent in 2010–13 to 1.7 percent in 2014–17. In addition, nearly a quarter of the spending on goods and services was dedicated to civil servants’ remuneration and benefits, with 15.2 percent spent on salary top-ups for top officials (representaçao dos serviços) and 8.6 percent on foreign travel. 32. Transfers and subsidies increased from 2.7 percent of GDP in 2010–13 to 3.2 percent in 2014– 17. The increase was driven by spending on autonomous services, the bulk of which represents salaries and benefits for staff at embassies, and other autonomous institutions. Moreover, spending on pensions for civil servants, which are fully funded through the budget, accounted for 38 percent of total transfers in 2010-17. Although the coverage of the pension scheme is low, it represents a fiscal strain. The increase in transfers was also due to calls on government debt guarantees in 2016 (to Guiné Telecom—1.4 percent of GDP) and 2017 (to the utility company EAGB (Electricidade e Aguas da Guine-Bissau) – 0.8 percent of GDP). 33. Other current spending increased rapidly, driven by tax collection bonuses and high levels of non-titled expenditure (DNTs)25 in 2014–16, notwithstanding improvements in 2017. Civil servants at the MEF receive salary top-ups calculated as a percentage of the total tax revenue. Tax collection incentives and refunds26 have grown in recent years, reaching 1.3 percent of GDP in 2017 (Figure 1.6.f). In addition, expenditures on non-regularized expenditures (DNTs) and payments for medical treatment abroad for officials and civil servants (encargos com saude), surged from 0.6 percent of GDP in 2010–13 to 1.1 percent in 2014–17 due to high levels of DNTs in 2014–16. Improvements in public financial management in 2017 saw spending on this subcategory fall sharply to 0.5 percent of GDP. 37. The size of the official payroll and spending on “goods and services” is low compared to Guinea-Bissau’s fragile peers, while transfers were higher than all peer countries. The public sector wage bill is below the level of most of its structural peers and the WAEMU average. Guinea-Bissau’s wage- to-current spending ratio is also amongst the lowest within benchmark countries. However, the actual wage bill (including nonwage compensation) surpasses those of peers. Guinea-Bissau’s spending on goods and services was also far below its structural peers and WAEMU counterparts. In contrast, transfers averaged 3.7 percent of GDP over 2015–17, higher than all its structural peers, except The Gambia. Functional Classification 38. Guinea-Bissau’s high level of security spending poses a barrier to allocating more public resources towards priority economic and social sectors . Security spending accounted for an average of 16 percent of total expenditure (or 3 percent of GDP) in 2010 –17—falling from 18.7 percent of total spending in 2010–13 (a period of heightened political instability) to 13.9 percent in 2014–17. When external loans and grants are excluded, the share is significantly higher, with security spending averaging 23 percent of domestically funded expenditures over 2010–17. While defense spending fell in real terms by 15.2 percent, spending on internal security increased by 41.7 percent during the above-mentioned period. 25 DNTs are defined as any treasury outlay not properly accounted for by the National Budget Directorate and/or not included in the budget. 26 Referred to as ‘incentivos para a cobrança de receitas and Restituições de receitas e incentivos’ in Portuguese. 28 Guinea-Bissau’s outsized security sector absorbed 36.1 percent of the public-sector wage bill and 35 percent of total spending on goods and services between 2010 and 2017. Figure 1.7: Guinea-Bissau: Trends in Public Spending a. Guinea-Bissau: Spending by Administrative b. Guinea-Bissau: Breakdown of Spending by Classification (2010–17) Administrative Classification (2010–17) 20 100% Percent of Total Domestic Spending 80% 15 Percent of GDP 60% 10 40% 5 20% 0 0% 2013 2010 2011 2012 2014 2015 2016 2017 2015 2010 2011 2012 2013 2014 2016 2017 Administration Defense and Security Administration Defense and Security Economic Education Economic Education Health Health Source: MEF. Source: MEF. c. Benchmarking: Public Versus Private Investment d. Guinea-Bissau: Spending on Goods and Services by Major Sectors (2010–17) 16 3.0 Private investment 14 Public investment 2.5 12 10 2.0 Percent of GDP Percent of GDP 8 1.5 6 4 1.0 2 0.5 0 0.0 2010 2011 2012 2013 2014 2015 2016 2017 Administration Security Social Economic Source: World Bank. Source: MEF. 39. ‘General administration and finance’ accounts for 27 percent of total public expenditure, while health and education represent less than 25 percent (Figure 1.7.a and 1.7.b). The rapid growth in spending on ‘general administration and finance’ indicates the growing importance of centralized expenditure over spending at ministry levels. Spending on the social sectors (education and health) has grown but investment and service provision remain dependent on external funding. Social spending increased from 3.8 percent of GDP in 2010–13 to 5.7 percent in 2014–17. Education has particularly experienced a rapid growth, driven by an increase in the number of teachers and higher remuneration. Over half of the social spending was financed with external grants and loans and nearly all the capital was donor funded. This dependence on external funding has made the provision of basic social services vulnerable to aid volatility. 40. Public investment is low and somewhat inefficient 27, even compared to fragile peers. Guinea- Bissau’s gross investment averaged less than 6 percent over 2014–17, compared to averages among its 27 World Bank (2017) Public Expenditure Analysis. 29 fragile peers which ranged from 12.6 percent in the Central African Republic to 20.5 percent in Burundi. The country has also one of the lowest public investment levels, at 3.7 percent of GDP in 2014–17, compared to an average of 7.3 percent for structural peers (Figure 1.7.c). Total health and education expenditures remain low, together accounting for a little over 5.5 percent of GDP. Education spending is particularly low, representing 2.8 percent of GDP, compared to 4–6 percent for the country’s structural peers. 41. Given limited fiscal space, Guinea-Bissau needs to improve the efficiency of public spending to achieve its development goals. Low levels of spending, coupled with low efficiency, translate into weak social outcomes (Figure 1.8). In health, not only does Guinea-Bissau spend less than almost any other SSA country but it also achieves poor outcomes for every dollar spent. Life expectancy at birth is among the lowest while maternal and infant mortality rates are among the highest in the world. Guinea-Bissau has lower life expectancy than many countries that spend the same amount on healthcare. Similarly, in education, 44 percent of children of primary school age are out of school, almost double the WAEMU average.28 The education system is also characterized by low completion rates and alarmingly low levels of learning.29 Figure 1.8: Guinea-Bissau: Efficiency Frontier a. Health Spending, Efficiency Frontier b. Education Spending, Efficiency Frontier 90 100 85 Upper secondary enrollment in 2015 80 80 Life expectancy at birth in 2015 75 60 70 65 40 60 stochastic frontier 20 stochastic frontier 55 DEA frontier DEA frontier benchmarks benchmarks 50 0 0 5 10 15 20 0 2 4 6 8 10 Health spending (% of GDP) in 2015 Education spending (% of GDP) in 2015 Source: World Bank staff estimates. Source: World Bank staff estimates. 1.4.3 Government expenditure: Budgeted Versus Executed Spending 42. Budget execution performance was affected by overruns in some categories and underspending in others. Over 2010–17, there was a slight underspend in overall budget execution. However, this masks a shift from underspending by 19 percent in 2010–13 to overspending of 21 percent in 2014–16, albeit this declined to 9 percent in 2017. The highest overruns were in “other current spending”, averaging 60 percent in 2010–17, which indicates the excessive use of this category—intended for miscellaneous expenditure—to execute non-budgeted expenditure. Budgeted investment saw underspending of 14.2 percent in 2010–17 due to the dependence of capital spending on volatile aid flows. 43. Strengthening the improvements in public financial management (PFM) seen in 2017 will be critical to better aligning the execution of expenditure with budget appropriations. Substantial 28 Guinea-Bissau Ministry of Education (2015). 29 It is important to note, however, that efficiency gains in these sectors are more likely to come from improved outcomes in the short term or targeted rationalization of expenditure in the medium to long term. 30 deviations from budget appropriations across all spending items over the past years point to weaknesses in preparing realistic budgets and containing execution within the budget envelope. In 2017, a cautious fiscal stance and improved expenditure controls led to a more disciplined budget execution. Efforts to improve PFM practices, including the re-establishment of a Treasury Committee, contributed to the disappearance of DNTs in 2017, which had been a persistent problem for several years. Continued improvements in budget preparation should result in more realistic budgets and spending execution. 1.4.4 Public Investment Program 44. The reliance on aid has resulted in large swings in investments with adverse implications for sustained growth and development. Guinea-Bissau’s public investment program (PIP) is almost entirely funded by external aid, of which grants account for nearly 75 percent, with the rest financed by concessional loans. The major donors include the UN System, the West African Development Bank, the World Bank, and the African Development Bank. Its experience with aid volatility reflects the wider pattern of international assistance to fragile countries. Countries that have experienced periods of fragility, violence, or conflict are, in general, more likely to experience higher volatility in aid flows. Donor support fell by almost half (47 percent) between 2010 and 2012, and by another 15 percent in 2016 in response to military coup and political crisis, but rose during the brief post-conflict transition (2014/15). These interruptions have severely affected the country’s public investment agenda and derailed reform efforts that are often aid funded. 45. Efforts to devote more domestic resources to mitigate the impact of “stop-go aid” have met with little success, with the contribution of domestic spending remaining very modest . The government’s contribution to the PIP increased slightly in 2012 and 2013 and again in 2015 to make up for the temporary suspension of aid. Nevertheless, the country is far behind the WAEMU convergence criteria that set the share of domestically funded investment to tax revenue at 20 percent. Guinea-Bissau’s ratio remained below an average of less than 2 percent, as recurrent expenditures grew apace with revenue collection, leaving little space for public investment. The lack of government-financed public investment also reflects serious capacity constraints in identifying, appraising and implementing investment projects which resulted in a dearth of bankable investment projects. Policy Recommendations Domestic revenue mobilization 30. Drawing on lessons from successful peers elsewhere, Guinea-Bissau needs to develop a reform strategy featuring a well-sequenced set of tax policy and revenue administration measures. In the short term, given limited technical capacities and political challenges, any reform strategy need to take into account administrative capacity. Short-Term Measures - Minimize and better target tax exemptions, particularly those linked with substantial foregone revenue. - Strengthen monitoring to enhance compliance through tax audits and penalties, among other measures. For example, a firm could be excluded from bidding processes or exemptions if it did not comply with its tax obligations. - Better monitor large taxpayers. Government may want to ensure that companies do proper accounting and financial reporting. The large taxpayer office needs to have a clear roadmap to inspect all firms through such measures as tying up their financial statements with the volume of their businesses. 31 - Broaden the tax base. Implement measures to curb tax evasion and increase the number of registered tax payers. - Recover outstanding tax arrears through stricter enforcement mechanisms. - Consider introducing a special excise tax on selected alcoholic products. Excise taxes do not generally require fundamental changes to the tax system and thus are relatively simple to implement. Proposed Areas for Medium-Term Reforms - Consider introducing tax policy reforms. A comprehensive reassessment of the tax code is needed to establish a simple and transparent regime, reduce compliance costs and ensure effective implementation. - Rationalize tax rates in line with the WAEMU fiscal directives. As trade and investments in the region have picked up, it will be important to harmonize tax and customs regulations with WAEMU standards. - Build administrative and institutional capacity. The tax and customs departments should strengthen their technical skills through on-the-job training and supported by the digitization of transactions to facilitate tax payments and monitor non-compliance. Government Expenditure - Rethink the current allocation of resources. The need to allocate greater resources to infrastructure and social spending brings to the fore the fiscal implications of continuing to spend 15 percent of the budget and 37 percent of the wage bill on security.30 - Obtain a better understanding on the size and composition of public sector staff. As a first step towards a civil service reform, the government may wish to deepen and expand the coverage of the 2017 audit of public administration. 30 Large security spending may reflect a less-productive use of scarce resources as the country does not face significant external threats. 32 Policy Recommendations – Public Sector Wage Bill Technical Timeframe feasibility policy impact Public employment policy - Use the recent civil service census to carry out a comprehensive functional review to identify Medium Medium term overlapping functions and redundancies. - Carry out a census of irregular public sector employees to inform future policy options aimed Low Medium term at regularizing their situation and/or reducing their numbers. Wage setting and pay policy - Use hiring freezes selectively to reduce overstaffing in certain sectors. High Short term - Freeze or limit the number of promotions to slow wage increases Medium Short term -Carry out a comprehensive review of non-wage compensation to ensure that pay is Medium Medium term competitive, equitable and transparent. Tax collection bonuses need to be a central part of this. HR payroll control - Complete organigram, including terms of reference for each position. Medium Short term - Include the remaining civil servants not yet included the census, a total of 1,023 military staff. High Short term - Initiate the process of removing civil servants from payroll when they reach retirement age. High Short term - Implement a fully functional Human Resources Management System and link it to SIGFIP. Low Medium term -Strengthen transaction controls, payroll reconciliations, and auditing. Medium Medium term Accounting and reporting - Update the existing budget classification system to ensure that all employee compensation High Short term expenditures are appropriately accounted for. 33 Chapter 2: The Public Sector Wage Bill1 While the official public sector wage bill in Guinea-Bissau is small relative to the size of its economy, these figures do not provide the whole picture—total compensation of government employees could be more than double the official figures. The share of the wage bill allocated to the social sectors has risen, particularly in education, while the share of the security sector, although still high, has fallen. While Guinea-Bissau’s wage bill is projected to remain reasonably affordable, an economic shock could cause it to rise over 10 percent of GDP, putting a significant strain on the country’s public finances. This risk could be mitigated by: (i) reviewing public employment policies to slow hiring and promotions, and reviewing non-wage compensation, with tax collection bonuses forming a central part of this.; (ii) strengthening human resource management and accounting procedures; and (iii) building on and extending the recent civil service census to review public sector employment and prepare staffing plans. 1. The chapter is organized as follows. It first provides an overview of Guinea-Bissau’s public sector wage bill. Sections 2.2–2.3 analyze the civil service headcount, pay structure, and sectoral composition of public sector employees. Section 2.4 estimates total public sector compensation by incorporating items that are unaccounted-for in the official wage bill, including pensions, bonuses, salary top-ups, health benefits, and salaries of staff in autonomous state institutions. Section 2.5 makes projections of Guinea-Bissau’s wage bill under different scenarios. Finally, Section 2.7 concludes with a discussion on options for reforms. 2.1 Overall Trends in the Wage Bill 2. The official wage bill in Guinea-Bissau is small relative to GDP and in line with peer countries. Official government figures show that wages and salaries represented, on average, 4.9 percent of GDP over 2010–17. The wage bill accounted for 57.5 percent of tax revenues over the same period, which exceeds those of peer countries and puts Guinea-Bissau far away from meeting the WAEMU convergence criteria of 35 percent. In addition, the official wage data exclude several budget categories related to staff compensation which would otherwise increase the government’s wage bill significantly. In fact, estimates indicate that total government employee compensation could amount to nearly double the official wage bill, reaching 9.0 percent of GDP and 93.5 percent of tax revenues in 2015, putting Guinea-Bissau among the countries with the highest wage bills (Figures 2.1a and b). 3. Wages and salaries in the public sector have declined as a share of revenues, but it remains far above the WAEMU convergence criteria. The official public sector wage bill has decreased as a share of GDP, from 5.9 percent in 2004 to 4.5 percent in 2017 (Figure 2.1.c). Moreover, as domestic revenues have risen during the same period, wages and salaries also fell as a share of domestic revenues, from 82 percent in 2007 to 34 percent in 2017. The decline accelerated after 2014 due to higher real GDP growth and a rise in domestic revenues, whereas wage levels and employment in the public sector increased at a slower pace. However, Guinea-Bissau has an undiversified export base and a volatile political system, and remains vulnerable to external and internal shocks. As has happened in the past, revenues could fall abruptly, making the wage bill hard to sustain. Therefore, future wage growth should be prudent and carefully planned, to ensure its sustainability and leave more fiscal space to finance essential public services and investment. 1 Prepared by Kjetil Hansen (Senior Public Sector Specialist, GGO) with contirbutions from Jorge Retana de la Peza (Consultant, GMTA2). 34 4. The share of wages and salaries allocated to the social sectors has grown rapidly, largely due to increased spending on the education sector (Table 2.1). The hiring of new teachers has seen the wage bill of the education sector jump from 1.1 to 1.7 percent of GDP between 2010 –13 and 2014–17 and nearly double as a share of wages and salaries. Meanwhile, the security sector ’s share of the wage bill has fallen, albeit slowly, from 41.9 percent in 2010 to 36 percent in 2017. Figure 2.1: The Public Wage Bill in Guinea-Bissau a. Government Employee Compensation as a Share of b. Government Employee Compensation as a Share GDP (2015) of Tax Revenues (2015) 16 100 90 14 Percentage of tax revenues 80 12 70 Percentage of GDP 10 60 50 8 40 6 30 4 20 10 2 0 0 Kenya Senegal Rwanda Timor-Leste Uganda South Africa Mauritius Guinea-Bissau Guinea-Bissau (Tot. compensation) Kenya Rwanda Senegal Uganda Timor-Leste South Africa Mauritius Guinea-Bissau Guinea-Bissau (Tot. (Official) compensation) (Official) Source: MEF, IMF and World Bank staff estimates. Source: MEF, IMF and World Bank staff estimates. c. Wage Spending in Guinea-Bissau (2004–17) 100 20 80 15 60 Percent Percent 10 40 5 20 0 0 2007 2004 2005 2006 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Official Wage Bill (% of domestic expenditures) (Left axis) Official Wage Bill (% of domestic revenue) (Left axis) Official Wage Bill (% of GDP) (Right axis) Source: MEF and World Bank staff estimates. 2.2 Civil Service Headcount and Sectoral Composition 5. The number of official government staff has increased only slowly over 2005–18, but the number of unofficially employed staff could be large . As noted in Chapter 1, the total number of civil servants, including military and security staff, stood at 20,550 in 2005 and increased by 17 percent to 23,550, according to the latest civil service census carried out in 2016/17. The number of civil servants per 100 population remained steady at 1.3, putting Guinea-Bissau slightly above some of its WAEMU peers, including Senegal and Guinea. However, there is evidence that the actual size of the public sector workforce 35 could be substantially larger given the large number of unofficial employees, many of whom are paid with budgetary resources intended for non-wage spending or ministries’ own revenues.2 6. In terms of sectoral distribution, the number of staff in the education sector has increased by 43 percent since 2010, on the back of a rise in the number of teachers . The Ministry of Education also relies on additional teachers hired on a temporary basis (contratados) for the school year (around 6 months every year). Temporary staff are not included in civil servant numbers. In 2016/17, the number of temporary teachers stood at 2,919 or 12 percent of the government’s employee headcount. Temporary teachers are paid lower salaries than their permanent counterparts and are only hired for the duration of the school year. However, repeated teacher strikes over lower pay and poor working conditions, and income unpredictability have often reduced the length of the school year, and shortened the duration of temporary teacher contracts. Improving the quality of the public education system will probably require a shift to a more sustainable model which incentivizes performance, favors capacity building, and retains talent. Table 2.1: Share of Wages and Salaries by Sector (2010–17) Avg. Avg. 2010 2011 2012 2013 2014 2015 2016 2017 2010–13 2014–17 As a percent of wages and salaries As a percent of GDP Administration 17.4 15.0 17.0 16.2 15.6 17.2 16.8 15.8 0.9 0.8 Finance & Economy 5.1 4.3 4.1 2.9 5.2 6.4 3.9 2.4 0.6 0.2 Justice 4.6 4.7 8.1 8.2 6.6 6.9 6.7 6.3 0.3 0.3 Security 45.0 44.5 42.1 46.2 39.3 37.9 35.0 36.0 2.3 1.9 Defense 28.5 25.7 23.3 24.8 20.6 20.2 20.6 21.3 1.4 1.0 Interior Security 16.5 18.9 18.7 21.4 18.7 17.8 14.4 14.8 0.9 0.8 Social 33.5 36.0 36.3 32.8 41.2 40.9 44.3 44.4 1.6 2.2 Education 23.6 26.2 25.6 21.3 31.3 31.0 33.6 35.1 1.1 1.7 Health 9.6 9.6 10.5 11.4 9.6 9.5 10.4 9.0 0.5 0.5 Economic 4.1 4.4 4.7 4.8 3.9 4.0 4.0 3.8 0.1 0.2 Agriculture and Rural dev. 1.9 1.5 1.3 1.2 1.0 1.0 0.9 0.9 0.1 0.0 Commerce, tourism, and art 0.8 1.1 1.1 1.2 1.0 1.0 1.1 1.1 0.0 0.1 TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 4.8 5.1 Source: MEF. Box 2.1: The Structure of the Guinea-Bissau Civil Service The government currently uses five separate career structures: general civil service, judicial, health, defense, and internal security. Each has a different salary table defining grades and pay scales. Most government employees are in the civil service, which is structured into four groups (0 to IV), which are further split into different categories, as follows: - Group 0: Categories P1 to P6: Includes the president, vice-president, and ministers - Group I: Categories A to E: Includes secretary generals, director generals, and advisers - Group II:– Categories A to C: Includes service directors and regional delegates - Group III:– Categories A to J: Includes technicians and officials - Group IV:– Categories A to B: Non-administrative staff (unskilled staff) Source: MEF. 2 At the Ministry of Economy and Finance, for instance, the total number of staff is more than double (139 percent higher) than the official employee headcount. Moreover, the number of employees at autonomous government institutions and state-owned enterprises, paid either through the government’s budget, sectoral taxes/fees (i.e. telecoms licenses) or own revenues (i.e. port fees, utilities), is un clear, but could be large and growing, given that their hiring processes are less onerous and more vulnerable to nepotism. 36 7. While the staff headcount in the military sector has remained broadly stable, the size of staff in the internal security sector has increased by a third since 2010 (Figure 2.2.a and b). Guinea-Bissau has 4.7 security personnel per 1,000 inhabitants, including about 2.3 military personnel per 1,000 inhabitants or nearly double the regional average of 1.2. This compares to less than one frontline health worker per 1,000 inhabitants. The security sector includes the military and paramilitary (police and gendarmes, the national guards, and the security intelligence service). Given the security sector accounts for 37 percent of public sector personnel and absorbs about 35 percent of wages, reforming it will be crucial to any strategy aimed at reducing the wage bill, although past reform efforts have faced strong resistance. 2.3 Civil Service Grade and Pay Structure 8. Guinea-Bissau’s civil service is dominated by technicians, with very few employees at the top and bottom of the scale. Nearly 80 percent of civil servants belong to Group III, consisting of mid-to-low ranking technicians and officials (Box 2.1). Only 7.8 percent of civil servants belong to the lower non- administrative categories. This reflects the fact that civil service retrenchment initiatives over the last 15 years sought to reduce the wage bill by decreasing the number of unskilled workers and outsourcing the functions they performed (i.e. cleaning and security). However, the salaries of many of the workers dismissed as part of these downsizing programs continue to be paid by the Government but do not appear on the official wage bill, as their salaries are classed under transfers. At the other end of the scale, high-ranking officials in Groups I and II make up less than 13 percent of the civil service. The number of positions in these grades is relatively fixed, while a lack of defined career paths with pre-set and functioning promotion mechanisms limit career progression. Box 2.2 presents the civil service structures of Cameroon and Zambia. Box 2.2: Civil Service Structures in Zambia and Cameroon Although no single best model for staff distribution exists, mid-career and technical staff in charge of operational and technical work often constitute the core part of public service. In Zambia, most staff are distributed across the grades with a concentration within the range of F to L grades while in Cameroon, the civil service has a top-heavy structure (Grade A1/2) with a large number of contractual workers (Category 8). Civil Service Structure in Zambia (left) and Cameroon (right) In Cameroon, in the wake of structural adjustments that limited or froze the hiring of new staff, many technical staff, mainly teachers, were hired as contractual workers in the late 1980s and 1990s. These now constitute a third of the workforce. This has led to a situation in which the number of staff in managerial positions (Grade A) is almost as high as those in technical positions (Grades B and C). Cameroon’s low basic wages, inverted civil service structure, and reliance on poorly paid contract workers without clear career prospects, explain low civil service morale and ineffectiveness. This example underlines the importance of well-planned recruitment, career management of staff, and adequate remuneration policies. Source: MEF. 37 9. Salary tables have remained broadly unchanged since 2004, with upward adjustments only for the lower grades. There is no automatic or regular mechanism to adjust public sector wages in Guinea- Bissau, for instance, by tracking price levels or cost of living indicators. Wage level adjustments are decided on an ad hoc basis and, in fact, have been limited over the last decade. After a large-scale salary harmonization in 2004 aimed at reducing vast wage differentials between the top grades and the rest, which saw a large increase in average wages, salary levels were adjusted only once, in 2012. This most recent adjustment—designed to increase the wages of those at the bottom of the salary scale—led to an increase in the average wage of staff in Group IV by 58 percent in nominal terms (Figure 2.3.b). Figure 2.2: Guinea-Bissau: Trends in Public Employment and Wage a. Employee Headcount by Sector (2010–17) b. Percentage of Population Employed by Government 7% 25000 6% 20000 5% 4% 15000 3% 2% 10000 1% 5000 0% 0 2010 2014 2015 2016 2017 Defense Interior security Education Health Adminstrative and economic Source: MEF, MPA, and World Bank staff estimates. Source: ILOSTAT, World Bank, MEF, and MPA. c. Structure of the Civil Service by Grades d. Average Wage per Grade % of total headcount Multiples of GDP per capita Group 0 (Cat. P1-P6) 13.0 Group 0 (Cat. P1-… 0.46% Group 1 (Cat. A-B) 9.9 Group 1 (Cat. A-B) 0.46% Group 1 (Cat. C-E) 6.8 Group 1 (Cat. C-E) 2.7% Group 2 (Cat. A-C) 3.7 Group 2 (Cat. A-C) 9.6% Group 3 (Cat. A-C) 3.2 Group 3 (Cat. A-C) 46% Group 3 (Cat. D-F) 2.0 Group 3 (Cat. D-F) 25.6% Group 3 (Cat. G-J) 1.3 Group 3 (Cat. G-J) 7.4% Group 4 (Cat. A-B) 1.0 Group 4 (Cat. A-B) 7.8% Average 2.8 0% 10% 20% 30% 40% 50% 0 2 4 6 8 10 12 14 Source: MPA and MEF. Source: MPA and MEF. 10. The average wage in the civil service has increased by 80 percent in nominal terms since 2005 despite salary levels remaining constant for most grades (Figure 2.2.d). The structure of the civil service has changed significantly since 2005 due to retrenchment programs (Figure 2.3.a). The share of staff in Group IV fell from 25 to 8 percent, while Group III increased from 63 to 79 percent. Thus, most low-skilled workers were substituted by workers in professional categories. In contrast, the proportion of workers in the higher-grade groups (Group 0 to Group II) remained mostly unchanged. This perhaps reflects a fixed number of positions at those levels and a lack of promotion opportunities for workers in Group III. The average wage in Groups I and II also remained constant, while the average wage for civil servants in Group III nearly doubled, reflecting the fact that more workers were in the higher paying categories within the 38 group (Figure 2.3.b). Although these changes resulted in a large average wage increase in nominal terms, they have barely allowed the average wage to keep up with GDP per capita. 11. The average wage in the defense sector has grown by 90 percent in nominal terms as the structure has become increasingly top-heavy. Two-thirds of the Bissau-Guinean military is composed of officers, while only 33 percent are soldiers, down from 46 percent in 2005 (Figure 2.3.c). In addition, while the average wages of top officials have remained unchanged, those of mid-ranking officers and soldiers have grown significantly over the last 13 years, further driving the increase in the army’s average wage (Figure 2.3.d). A reform to rebalance the structure of the army, reducing the proportion of officers to 30 percent, and cutting down its size, has been on the agenda for decades, but has so far proved unsuccessful. This reform will be essential if Guinea-Bissau is to reduce the army’s wage bill while making the military more modern, efficient, and able to meet its objectives. Figure 2.3: Structure and Trends in Public Employment in Guinea-Bissau a. Structure of the Civil Service by Group b. Average Gross Monthly Wage by Grade % of total civil servants CFAF ‘000s 700 Grupo 0 600 2005 Grupo I 2005 500 2017 2017 400 Grupo II 300 Grupo III 200 Grupo IV 100 - 0% 20% 40% 60% 80% 100% Grupo 0 Grupo I Grupo II Grupo III Grupo IV Average Source: MPA, MEF, and IMF. Source: MPA, MEF, and IMF. c. Structure of the Defense Sector by Rank d. Average Gross Monthly Wage in the Defense % of total military personnel Sector by Rank CFAF ‘000s General General Colonel Colonel 2005 Majors Majors 2017 Captains 2005 Captains 2017 Tenants Tenants Subtenants Subtenants 2nd Sergeants 2nd Sergeants Soldiers Soldiers Average 0 10 20 30 40 50 - 200 400 600 800 Source: MPA, MEF, and IMF. Source: MPA, MEF, and IMF. 2.4 Computing Total Public Sector Compensation 12. Official statistics do not provide a full picture of total government employment and compensation, and their fiscal implications. According to official data, the government’s wage bill has been decreasing as a share of domestic revenues and expenditures. The average public sector wage has 39 also remained low and stable relative to GDP per capita. Public employment as a share of the population is broadly in line with other countries in the region. However, these figures do not give the complete picture and fail to explain why, despite rising revenues, little fiscal space has been created for other expenditure items, such as investment and purchases of goods and services. In fact, the budget execution data for wages and salaries omit large portions of public spending dedicated to compensations and benefits of civil servants. There are indications that unofficial government employment could be large and employment in the public sector is likely to be substantial higher than official data suggest. 13. When nonwage benefits are accounted for, the wage bill nearly doubles, averaging 8.9 percent of GDP over 2010–17, making Guinea-Bissau’s wage bill one of the highest among its peers (Table 2.2 and Figure 2.4). Several spending items are classified under different budget lines. Civil service pensions, wages and benefits for employees at autonomous state institutions, wages and benefits for members of parliament, and wages of redundant staff (pessoal excedentario) are classified under transfers and averaged 2.4 percent of GDP over 2010–17. Bonuses paid to MEF staff to incentivize tax collection as well as medical benefits for civil servants are recorded under “other current expenditures”. These items have increased rapidly, reaching 1.3 percent of GDP in 2017. Finally, travel expense allowances (ajuda de custo) and salary top-ups for top officials (representaçao de serviço), as well as miscellaneous wages without budget space in other lines are classified under “purchases of goods and services” and averaged 0.5 percent of GDP. Table 2.2: Composition of Total Public Sector Compensation (2010–17) Avg. Avg. 2010 2011 2012 2013 2014 2015 2016 2017 2010–13 2014–17 As a % of domestic expenditure As a % of GDP Official wages and salaries 40.1 41.9 38.4 47.3 39.8 32.5 31.2 37.9 4.7 5.0 Other compensation items 36.1 36.0 32.3 32.4 29.0 28.3 26.5 32.4 3.9 4.1 Civil service pensions (transfers) 10.7 9.3 9.4 12.1 7.6 6.7 8.0 9.2 1.2 1.1 Autonomous services wages (transfers) 3.9 4.0 3.6 2.5 4.4 3.5 3.5 3.8 0.4 0.5 Private individuals (transfers) 4.1 3.1 3.7 4.1 2.6 1.9 1.8 1.4 0.4 0.3 National assembly wages (transfers) 3.5 3.3 3.1 4.4 3.1 2.9 3.2 3.5 0.4 0.5 Tax collection incentives (other current) 8.7 11.5 7.7 4.3 7.2 9.6 6.8 10.9 0.9 1.2 Medical benefits (other current) 0.1 0.3 0.1 0.3 0.3 0.3 0.3 0.6 0.0 0.1 Travel expenses (goods & services) 0.0 0.0 0.6 0.7 0.5 0.6 0.6 0.8 0.0 0.1 Salary top-ups (goods & services) 3.3 3.6 3.2 2.8 2.6 2.0 1.8 2.0 0.4 0.3 Other wages (goods & services) 1.9 0.9 0.7 1.2 0.8 0.8 0.6 0.2 0.1 0.1 Total compensation 76.2 77.9 70.8 79.7 68.8 60.9 57.7 70.2 8.6 9.2 Source: MEF, World Bank staff estimates. 14. Civil service pensions are a significant fiscal burden, despite having a very low coverage rate . Public pensions constitute the largest compensation item not included in the official figures, averaging 1.1 percent of GDP and 9.1 percent of domestic expenditure between 2010 and 2017. Despite their large fiscal cost, their coverage is very low: only 2.8 percent of the active workforce3 contribute to the system and less than 4 percent of individuals of retirement age (60 years old and older) receive a pension. 15. Tax collection incentive payments absorbed nearly 10 percent of domestic revenues in 2017, while benefitting only 0.2 percent of the working-age population.4 These payments are calculated as a share of tax revenue, with employees at the Revenue Office (Directorate General of Taxes and Customs; DGCI and DGA) receiving 6 percent of tax revenues collected by the DGCI, those at Customs receiving 1 3 World Bank (2016): Sistema de Pensões na Guiné-Bissau: Principais Desafios e Perspetivas para o Futuro. 4 The population of Guinea-Bissau between the ages of 15 and 60 is 997,000, according to the UN’s World Population Prospects 2017. 40 percent of export taxes and 1.5 percent of import taxes collected, and staff from other units of the MEF receiving 3 percent of tax revenue from both the DGCI and the DGA. Incentivos are high, at CFAF 5.3 million (nearly US$10,000) per employee in 2017, which is equal to 13.4 times GDP per capita and 4 times the average public sector wage (Figure 2.5.a and b). More senior staff receive a larger share, and incentivos are not linked to individual performance. These bonuses like provide only a small incentive for increasing tax collection, yet they come at a very high cost.5 Despite the ministry officially employing 773 staff at the end of 2017, bonuses were being paid to 1,852 individuals—an indicator of the size of unofficial employment. Figure 2.4: Trends in the Wage Bill, and Domestic Revenues (% of GDP) 16 14 12 10 8 6 4 2 0 2010 2011 2012 2013 2014 2015 2016 2017 Official Wage Bill Actual Wage Bill Revenues (Excl. Grants) Source: MEF and World Bank staff calculations. 16. Nearly 1,000 public sector workers made redundant as part of a retrenchment program have remained on the payroll indefinitely. The failure of a donor-sponsored civil service downsizing program, aimed at cutting the wage bill by reducing the number of unskilled workers and creating alternative private sector employment has led to surplus workers ( excedentarios). These “workers” remain on the payroll even though they are not required to work. As a result, the government is paying for a service provider to carry out these functions such as cleaning and security, while still paying the salaries of the civil servants formerly employed to do so. Any future retrenchment program should be better designed and implemented to increase its effectiveness at shifting workers to the private sector and achieving fiscal savings. 17. The wages of staff at diplomatic missions absorbed 2.5 percent of domestic revenues in 2016, which may be a high cost for maintaining 20 embassies and consulates. Staff in diplomatic missions abroad are paid under the autonomous services (serviços autonomos) budget line, which is classified under transfers rather than wages and salaries. In 2016, Guinea-Bissau had 20 diplomatic missions (embassies and/or consulates) abroad, employing a total of 202 staff, at a cost of 2.2 billion CFAF (US$4.1 million). The average salary for employees of diplomatic missions was 29 times GDP per capita and 10 times that of other public sector workers. In contrast, the wages and salaries of the 2,208 employees in the health sector amounted to 3.2 billion CFAF (US$6 million). Health professionals received, on average, one seventh of the average wage of embassy and consulate staff. 18. Medical benefit payments to civil servants are large, benefit only a small number of individuals, and lack the appropriate safeguards to avoid mismanagement . Civil servants are entitled to receive financial assistance from the government in case of a medical emergency that requires evacuation to seek adequate treatment overseas. This medical benefit ( encargo com saude) is non-contributory and 5 Since unofficial staff members, who in many cases have not had to go through the legally mandated selection and hiring processes, also receive these payments, the risk of corruption is also high. 41 financed through the budget. It is normally paid out as a lump sum of 1.5–2.5 million CFAF (US$2,800– 4,700) on the condition that the civil servant presents official documents to justify the need to seek treatment abroad. Not all civil servants enjoy equal access to this benefit, as payment normally requires approval from a high-ranking officials at the MEF. As a result, the way medical benefits are awarded is not transparent and is vulnerable to mismanagement. While this benefit is budgeted for under wages and salaries, actual expenditure has vastly exceeded the allocated budget, which has led to executions under different budget categories, including under ‘other current expenditures’. Figure 2.5: Tax Collection, Public Employment, and the Wage Bill a. Tax Collection Incentives and Income per b. Compensation of MEF Staff Relative to Capita (2010–17) Others (average over 2010–17) Multiple of GDP per capita 6 20 Defense 15 in millions of CFA 4 Health 10 2 Education 5 0 0 MEF (unofficial) 2010 2011 2012 2013 2014 2015 2016 2017 Incentives per head (left axis) MEF (official) GDP per Capita (left axis) Incentives as a multiple of GDP per capita (right axis) 0 5 10 15 20 Source: MEF and World Bank Staff estimates Source: MEF and Staff Estimates c. Public Employment and the Wage Bill Source: MEF, MPA, World Bank Staff estimates. 19. International comparisons suggest that while the wage bill absorbs a large share of government spending, it benefits a disproportionately small share of the population. Guinea-Bissau’s government formally employs only 1.3 percent of the population, but wages amounted to 36 percent of current expenditures in 2015 according to official figures, or nearly double that, 66 percent, when taking into account total compensation and benefits for civil servants (Figure 2.5.c). While the average wage is broadly in line with international standards, Guinea-Bissau’s low revenue mobilization capacity means that 42 a large share of scarce public resources is being spent on staff remuneration. This signals that Guinea- Bissau’s public sector is too large for its fiscal capacity. Increasing revenue collection, while reforming the public sector to control wage growth and to increase its efficiency, could create the fiscal space necessary for investment while improving service delivery. Box 2.3: Update on Biometric Civil Service Census and HR Database Reconciliation Between 2015 and 2016, the Ministry of Public Administration (MPA) carried out a biometric civil service census, which sought not only to identify individual employees but also to assign them to positions in the administration’s organigram. The completion of this census led to an initiative to reconcile it with the payroll database in use at the MEF, with the aim of creating a unified civil service database, as well as to identify anomalies, such as duplicate entries and ghost workers on the payroll. From August 2017, the government decided to block salary payments to individuals on the payroll whose details were inconsistent, incomplete or could not be appropriately verified, and urged those who did not receive their wages to verify their details with the MPA. While most salary payments that had been initially blocked were subsequently verified and unblocked, this initiative led not only to a fully reconciled and complete database, but also to the deletion of 1,065 anomalous entries (ghost workers, duplicates, etc.) from the payroll. As a result, the savings to the wage bill savings are projected to amount to 1.9 billion CFAF (US$3.5 million) in 2018, or over 5 percent of the official 2017 wage bill. In addition, the existence of a single complete civil service database could enable the MPA to better carry out its functions. In 2018, it is expected to have defined all positions with concrete terms of reference, identified vacancies and redundant positions, and designed a pilot performance assessment project. These initiatives could form the basis for a more efficient civil service in the future, in which recruitment is done on the basis of strategic and operational needs, individuals have the appropriate qualifications, and civil servants are awarded promotions and notch salary increases based on performance. However, the implementation of a fully functional human resources information management system will be necessary to guarantee the feasibility and sustainability of these advances. Source: MEF and MPA. 2.5 Wage Spending: Challenges and Opportunities 20. Increasing the number of civil servants may contribute to better service delivery; however, higher headcounts do not necessarily lead to better public service outcomes . Outcomes in the education sector in Guinea-Bissau have seen little improvement even though the number of teachers has increased rapidly over the last 8 years. The challenge of managing the wage bill in Guinea-Bissau, therefore, is not just an issue of reallocating and increasing spending in the social sectors but also about improving the efficiency and effectiveness of public wage spending. 21. An overall lack of human resource capacity is a major drag on public service outcomes, despite some recent initiatives. The Ministry of Public Administration is responsible for human resources management (HRM). Its mandate includes approving organigrams (staffing structures), defining career paths, recruitment, performance assessment, and promotions. However, a combination of mismanagement, political instability, and lack of capacity and resources has left the ministry unable to fulfil its mandate. As a result, HRM functions have been performed by the MEF on an ad hoc basis, with limitations including: lack of adequate controls on salary payments, politicized recruitment processes, no performance assessments, arbitrary pay rises and promotions, and a lack of up-to-date organigrams. A recent project to update the civil service census and create a unique HR database to be audited regularly could be the basis for major improvements in the future (see Box 2.3). Moreover, a new salary structure for health sector workers, which defines clear career paths and sets out progressive salary notch increases based on years of service, could serve as a blueprint for similar initiatives across the whole public administration. 43 22. The challenge of improving the efficiency and effectiveness of wage spending—and managing its overall fiscal burden—is not unique to Guinea-Bissau and is faced by all countries in the world. In the end, the path chosen to navigate this difficult terrain must necessarily reflect national priorities as well as local conditions and constraints, including budgetary resources, institutional capacity, labor market structure, demography, and geography. Many of the choices facing policy makers embody trade-offs that can be difficult to evaluate. Wage and employment policies are often shaped by forces unrelated to the core objective of improving public service outcomes. Such forces can include (i) political pressure to award ad hoc increases in wages or to expand the size of the public workforce in the run-up to elections; and (ii) fiscal pressure that forces governments to enact crude and untargeted cost-cutting measures in response to shocks. In these respects, Guinea-Bissau is no different to many other countries. Figure 2.6: Trends in Compensation and Public Employment a. Total Compensation Spending b. Age Distribution of Public Servants Projections (2017–23), % of GDP Overall (2017) 10.5% 4.5% 4.0% 10.0% 3.5% 9.5% 3.0% 9.0% 2.5% 8.5% 2.0% 1.5% 8.0% 1.0% 7.5% 0.5% 2017 2018 2019 2020 2021 2022 2023 BASELINE 0.0% SLOWER GDP GROWTH(2/3 OF BASELINE) (2) SLOWER GDP GROWTH(1/2 OF BASELINE) (3) 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65 68 EXTERNAL OR INTERNAL SHOCK (1) Age Source: World Bank estimates. Source: MFPRAT, MEF, and World Bank staff estimates. Note on Figure a: Based on data provided by the authorities and on the IMF’s most recent World Economic Outlook projections. (1) Assumes nominal GDP contracts by three percentage points in 2018 and follows a similar recovery trajectory after the most recent recession in 2012. (2) Assumes nominal GDP grows at two thirds of the rate projected by the IMF. (3) Assumes nominal GDP grows at half of the rate projected by the IMF. c. Age Distribution of Public Servants by Sector (2017) 6% 5% 4% 3% 2% 1% 0% 26 20 22 24 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 -1% Civil Servants Defense Education Health Source: MFPRAT, MEF, and World Bank staff estimates. 44 2.6 Scenario Analysis: Medium-Term Wage Bill Projections 23. In the absence of ad hoc measures that materially impact wage levels or employment, the total compensation bill is projected to stay constant at around 8 percent of GDP over the next six years (see Annex VI). This expected stabilization of the wage bill is not due to easing fiscal pressure from wage spending but to projections of robust economic growth. Even minor fiscal shocks such as slightly slower-than-anticipated rate of growth, could easily push the wage bill beyond 10 percent of GDP by 2023 (Figure 2.6.a). A more pronounced internal or external shock, like the one the country experienced after the coup in 2012, could see the compensation bill rise above 10 percent of GDP within three years. Given that domestic revenue stood at around 13 percent of GDP in 2017, such a shock could put public finances under significant stress. In addition, given the undiversified nature of its economy, and its reliance on imports of staple foods, as well as persistent political instability, the risk of such a shock occurring is not negligible. 24. Shocks would increase the need to limit the wage increase in real terms, curb non-wage benefits, and reconsider the size of employment, particularly in the security sector . In the short term, constraining wage growth can be accomplished by limiting the number of promotions awarded which, in the absence of outright wage increases, have become the only mechanism for wage progression. Awarding only high-performing individuals notch wage increases and promotions could increase productivity while having a positive net fiscal effect. Guinea-Bissau’s currently large non-wage benefits, such as tax collection bonuses and salary top-ups for top civil-servants, create significant inequalities within the system. Limiting them, or linking them more directly to performance, could increase equity and productivity, as well as creating fiscal space which could serve as a buffer in the event of an exogenous shock. 25. Age distribution analysis suggests some potential timings for the strategic restructuring and rejuvenation of public service (Figure 2.6.b and c). The retirement age in Guinea-Bissau is 60 years old for civil servants. The current age distribution of civil servants means there will be a large batch of retirees between 2018 and 2025 and another one after 2037 (Figure 2.6.b). In light of the current unbalanced structure, which does not reflect human resource needs in key service delivery sectors, these periods could be opportunities for the government to consider an attrition-based restructuring of the civil service and address the fundamental structural and organizational problems such as duplication of roles or sectoral imbalances. It could also be an opportunity to recruit competent and well-qualified young professionals. 2.7 The Fiscal Impact of Alternative Wage Bill Policies 26. Presented below are estimates of the fiscal impact of measures that could be deployed in the short term to create fiscal space for much-needed investment and to cope with exogenous shocks. Fiscal impact numbers are cumulative, in percent of GDP, relative to the baseline projection. It should be emphasized that these measures are not presented as recommendations—in general, across-the-board measures are more popular during times of economic crisis and are not a proper substitute for structural reforms aimed at wage levels or the “right sizing” of public employment. Instead, the measures are presented to illustrate the fiscal impact that could accrue from the portfolio of tools available to the government for adjusting wage spending and remuneration practices in the short term. - Attrition: in the absence of mechanisms to retrench redundant workers, attrition-based policies are the only available mechanism for reducing the size of the workforce in absolute numbers. A challenge, particularly in the defense sector, has been the unwillingness of high-ranking officers to retire, in the expectation of a higher pay-out as part of a reform, and to avoid losing the privileges and powers 45 attached to the position. This limits the fiscal and structural impact of an attrition policy without a more comprehensive reform effort. Table 2.3: Estimated Fiscal Impact of Attrition Policies (2018–23) Attrition 2018 2019 2020 2021 2022 2023 Fix the size of the workforce at its 2017 value 0.17 0.50 0.97 1.57 2.30 3.15 Replace two of three departing workers 0.16 0.47 0.94 1.57 2.35 3.27 Replace one of three departing workers 0.18 0.55 1.10 1.83 2.74 3.83 - Tax collection bonuses (incentivos): while cutting these payments in half would provide immediate fiscal relief, it would also be very contentious and would not contribute to better performance. Alternatively, bonuses could be calculated as a share of any increase in revenues relative to the previous year. While this would produce lower savings initially, reducing resistance to the reform, the savings would be incremental. This approach would also link pay to performance more directly and would insure the government against a fall in revenues. Table 2.4: Estimated Fiscal Impact of Reforming Tax Collection Bonuses (2018 –23) Tax collection bonuses (Incentivos) 2018 2019 2020 2021 2022 2023 Cut by half as a share of revenue 0.58 1.18 1.81 2.45 3.12 3.82 One quarter of revenue increase 0.41 1.04 1.72 2.53 3.28 4.05 One third of revenue increase 0.17 0.62 1.12 1.77 2.33 2.91 Half of revenue increase -0.33 -0.28 -0.17 0.16 0.31 0.47 - Defense and security sector reform: an effective implementation of the defense sector reform according to a plan by the Economic Community of West African States (ECOWAS) from 2014/15 (see Box 2.3) would have an immediate fiscal impact, creating fiscal space for spending on priority sectors. Extending this reform to the internal security sector would add a further 70 percent of fiscal savings. Table 2.5: Estimated Fiscal Impact of Defense and Security Sector Reform (2018 –23) Defense and Security Sector Reform 2018 2019 2020 2021 2022 2023 Defense reform by 2020 as per 2015 ECOWAS plan 0.00 0.00 0.00 0.50 0.99 1.46 Extending ECOWAS reform to both Defense and Security 0.00 0.00 0.00 0.85 1.67 2.47 27. Across-the-board measures, such as those discussed above, may generate meaningful fiscal relief in times of economic crisis but could be problematic from a labor policy perspective. Across- the-board measures aimed at wage levels, for example, can introduce or exacerbate disparities in the relative compensation paid to workers of varying skills, experience, and performance, potentially encouraging workers in critical positions and higher-performing workers to leave public employment. It is important to note that the existence of a wage premium for the average public sector worker does not necessarily imply that all public sector workers receive higher compensation than their private sector counterparts with comparable skills and experience. 28. Similarly, across-the-board measures aimed at employment (such as attrition-based policies) can result in the loss of critical expertise or skills, ultimately leading to their reversal or to a deterioration in the quality of essential public services. To some degree, these problems can be addressed by applying differential policies to different categories of workers or through the use of targeting (focusing on nonessential positions or exempting critical sectors). However, neither of these are proper substitutes for medium-term structural reforms to improve existing wage and employment policies. 46 2.8 Options for Structural Reforms 29. Over the medium to long term, structural reforms are the way to improve the efficiency and effectiveness of wage spending. Although total public employment is broadly comparable to peer countries, there is considerable room for improvement in how this workforce is distributed. As discussed above, reforms that make it easier for the government to retrench redundant workers would help reduce overstaffing and the duplication of functions across the public sector. Easing restrictions on redeployment would enable the government to better match personnel resources to staffing needs. Reforms that address wage-setting policies offer the potential to provide better incentives for performance by rewarding high- performing individuals and reduce existing disparities in the relative compensation of workers with varying skills, experience, and performance while also reducing average wage costs. 30. It is important to carry out a comprehensive functional review of key sectors, using the recent civil service census, to identify overlapping functions and reduce redundancies . The Government should consider tightening hiring processes and set a clear procedure for the approval of employees given the existing control practices are ineffective. In addition, a census of irregular employees should be conducted to inform future policy options aimed at regularizing their situation and/or reducing their numbers. An in-depth organizational and HR capacity assessment also needs to be conducted to determine capacity gaps across geographical areas. In the medium-term, Government should consider preparing multi-annual staffing plans as part of the process of approval and control of number of public sector employees, which could be another effective deterrent to wage bill increases. In the long-term, shifting from a lifetime employment model to a more flexible position-based model would be necessary for employment to be adjusted in response to changing needs, or demographic and technological changes. 31. The following measures could help strengthen payroll control. The administration’s organigram will have to be completed, including terms of reference of reference for each position, including details of required qualifications, skills and experience for each role. Civil servants not yet included in the census, a total of 1,023 military staff, need to be included in the payroll. It is also critical to strengthen transaction controls, payroll reconciliations, and auditing, particularly with respect to wage-related payments made in cash. Government may want to simplify and streamline the process of retirement and entry into the public pension scheme. The process of removing civil servants from payroll should be initiated when they reach retirement age. Government may also consider implementing a fully functional Human Resources Management System linked to the Integrated Financial Management System (SIGFIP) in the MEF. Underlying all these efforts is the urgent need to strengthen the quantity and quality of data available for fiscal analysis, budget formulation, and the monitoring of public sector employment and compensation. 32. Structural pay reforms have the potential to both offer fiscal relief and improve workforce incentives and fairness, thereby improving morale. The goal behind all wage-setting policies, of course, is to provide sufficient remuneration to attract and retain qualified workers and to create incentives for them to perform to the best of their ability over the course of their careers. In the public sector, however, this goal is often easier stated than met. There are many reasons for this, but one of the most relevant is the strong job protections enjoyed by public sector workers. In Guinea-Bissau, as in many developing countries, a job in the public sector is effectively a guarantee of lifetime employment. Other challenges include evaluating performance in the public sector when outcomes are harder to measure, and designing and implementing wage-setting policies that motivate civil servants to work harder and better. 33. In the medium- to long-term, it is important to link remuneration to performance than having incremental salary scales as in Guinea-Bissau. However, reforms should be designed and implemented taking into account the country’s limited administrative and institutional capacity. Structural 47 pay reforms should improve hiring to match actual needs and stop ad hoc promotions without adequate controls and no performance assessment. Instead, notch wage increases should be considered, so that increases are awarded only to the highest-performing workers. Implementing such notch increases will require: (i) developing clearly-defined performance criteria for all categories of workers across the entirety of the public sector; and (ii) designing and successfully implementing of a formalized and—crucially— transparent process for periodically evaluating the performance of workers against these criteria. 34. The enabling legislation required to introduce performance-based pay is already in place, but is yet to be implemented. Decree-law Nº7/2012 from October 19, 2012 establishes the principles and mechanisms to assess the performance of public sector employees as well as the role of performance assessment in career and salary progression. In addition, Decree-law Nº9/2012 of the same date defines the structure of public sector careers and the rules governing hiring and movement within and across careers. Introducing performance-based pay may not require further legislative or regulatory authority, thus government could focus solely on issues of design and implementation. 35. Over a longer horizon, making performance-based pay succeed will also require revising all existing wage tables as part of the process of conducting periodic pay reviews . The objective of a pay review is to establish pay parity: the principle that workers with comparable qualifications and experience ought to be paid comparably, regardless of where they work. As part of this process, Guinea-Bissau’s public sector wage tables should be revised so that each notch increase is small enough to be consistent with the government’s fiscal capacity but still large enough to create meaningful incentives for workers to improve their performance. So far, only wage tables for frontline health workers have been adapted to the new legislation, including the possibility of notch increases after every three years of service on the condition of good performance, although performance assessment systems are not yet in place. Table 2.6: Estimated Fiscal Impact of Structural Pay Reforms (2018–23) Introducing notch salary increases and abolishing ad hoc promotions 2018 2019 2020 2021 2022 2023 Increase wages to match inflation -0.10 -0.32 -0.65 -1.12 -1.74 -2.52 Award notch increases to one third of workers each year -0.08 -0.24 -0.48 -0.80 -1.20 -1.68 End ad hoc promotions and stop wage drift 0.21 0.63 1.24 2.04 3.02 4.17 36. The civil service census and organigram completed in 2017 could be a first step towards the implementation of performance-based management. Although performance-based management is currently stipulated in law, ad hoc promotions are the only avenue for career and salary progression. Promotions tend to be awarded without appropriate controls or due process (as the capacity and processes are not in place) and thus are prone to nepotism and corruption, and are not conducive to appropriate performance incentives. However, the recently completed census and organigram could be a powerful tool to start implementing regular performance assessments, as the MPA can now identify the responsibilities of each individual, and who they report to. Using this information, a first pilot performance assessment survey was conducted in 2017, with positive results. 37. Performance-based pay has the potential to generate significant fiscal space while reducing the risk of mismanagement. For illustrative purposes – assuming that the process of structural pay reform had been completed and could be implemented in 2018 – introducing performance-based pay, resulting in awarding a notch salary increase a third of workers every year, while ending ad hoc promotions, would yield net fiscal savings of 0.13 percent of GDP in 2018 and cumulative fiscal savings of 2.49 percent of GDP by 2023 relative to the baseline projection (Table 5.6). This would also create enough fiscal space for annual inflation-indexed wage increases at no net fiscal cost, relative to the baseline. 48 38. To this end, Guinea-Bissau needs to strengthen its human resources management and information system (HRMIS) and its public financial management processes. Weak information systems and public financial management processes impede efforts to monitor and control wage spending, forecast labor needs and manage recruitment, and identify employment and payroll irregularities. 49 Policy Recommendations – PFM, the Treasury Committee and SIGFIP Technical Timeframe feasibility policy impact Strengthening the Treasury Committee and building PFM institutions - Strengthen political leadership on PFM reforms. Develop an operational plan for PFM High Short term reforms, with clear and measurable indicators to be monitored. - Develop a priority matrix to help guide, inform, and rationalize public expenditure decisions High Short term taken in the TC. - Support capacity building for cash forecasting and liquidity risk monitoring, and develop a Medium Short term basic cash forecasting tool based on historical treasury information. - Expand the time horizon for cash planning, for instance rolling cash forecasts for at least Low Short term the next two or three months and monitoring liquidity risk. - Improve budget preparation using solid data from the sectoral ministries. High Medium - Streamline budget execution processes, with timely release of funds for priority sectors and High Medium tightened controls on the quality of services rendered. Financial management information system, SIGFIP - Improve the quality and reliability of public financial data. High Medium term - Use SIGFIP for the entire expenditure chain, including commitments and payments and Medium Medium term discontinue parallel paper based payments where possible. - Introduce a financial programming model to consolidate TSA implementation and Medium Medium term streamline treasury management. - Make SIGFIP accessible to all line ministries by moving from a LAN-based to a web-based Medium Long term system. 50 Chapter 3: Public Financial Management, the Treasury Committee, and SIGFIP6 Since late 2016 the Treasury Management Committee—commonly known as the Treasury Committee (TC)— has been providing a measure of control over cash management and public expenditure in Guinea-Bissau. However, the quality of expenditure remains low, with substantial funds benefitting central public sector personnel in various forms, instead of being used for frontline public services. Guinea-Bissau thus needs to improve expenditure prioritization, while also strengthening formal public financial management (PFM) institutions over the medium term and gradually allowing them to play their role in budget preparation, execution, and control. Limited use of the integrated financial management system and a lack of external audits also leads to unreliable expenditure data. Short-term priorities should focus on: (i) preparing a credible budget to guide public expenditures; (ii) consolidating recent gains in expenditure management and controls through the TC while also strengthening formal expenditure controls; (iii) gradually strengthening external audit and controls; and (iv) increasing transparency and data availability. Underlying these recommendations is the urgent need to improve the reliability of public finance data. 1. The chapter is organized as follows. It first provides an overview of recent PFM reforms in Guinea- Bissau. It then goes into some depth on the Treasury Management Committee (henceforth, the Treasury Committee; TC), and the government’s integrated financial management system, Sistema Integrado de Gestao Financeira Publica (SIGFIP). The reason for focusing on these two elements is as follows: (i) the Treasury Committee is the central control point for all strategic public expenditure allocation and execution decisions—effectively taking over the role of the national budget as an instrument for strategic budget programming, and replacing other control institutions; and (ii) the current weaknesses in the use of SIGFIP are critical to understanding the dysfunctional nature of expenditure management, including the lack of reliable expenditure data and the absence of expenditure controls outside of the TC. 3.1 The Quality of Public Financial Management in Guinea-Bissau 2. The Government’s PFM reform priorities are outlined in the Government’s program (2016– 18), but implementation has been limited. This program builds on recent reform strategies, including the DENARP II 2011–15 (the strategy for poverty reduction) and the long-term development strategy (Terra Ranka). The reforms outlined in the government program7 are very elaborate, and are generally aspirational high-level objectives without a clear operational program. Moreover, there are no monitorable indicators and no regular steering committee meetings to measure progress. A PFM coordination unit has been established to support the reforms, but the unit has largely focused on technical aspects and donor coordination matters. Box 3.1 provides a short overview of recent PFM reforms. 3. Guinea-Bissau scores below other fragile and conflict affected states on nearly all PFM related indicators. Its scores in the 2009 Public Expenditure and Financial Accountability (PEFA) assessment were dismal, with one B (comprehensiveness of budget documentation) and the rest D or D+ (Table 3.1). Some important progress was made in the years that followed on budget credibility, budget 6 Prepared by Kjetil Hansen (Senior Public Sector Specialist, GGO) with contirbutions from Gabriel Yorio Gonzalez (Public Spector Specialist, GGO). The section draws on the 2009 Public Expenditure Management and Financial Accountability Review (PEMFAR) report and the 2013 Public Expenditure Financial Accountability (PEFA) document as well as the government’s medium-term priorities as laid out in the government’s 2014 Programa de Estabilização e Desenvolvimento. 7 The current government, nominated in early 2017, has not been able to pass its program since the National Assembly considers the current government illegitimate as it was not appointed in accordance with the Conakry Agreement signed between the president and the party in control of parliament (PAIGC) in October 2016. 51 comprehensiveness, and transparency, as well as controls in execution (Box 3.1), reflected in the improved scores on the 2013 PEFA assessment (one A, six Cs). Guinea-Bissau’s scores on the PFM related indicators of the Country Policy and Institutional Assessment (CPIA) are also substantially below the Sub-Saharan African average as well as the average for all fragile and conflict affected states (FCS; Figure 3.1). Box 3.1: A Short Overview of Recent PFM Reforms Guinea Bissau’s history of PFM reforms has been beset by chronic political instability and modest progress has seen quick reversals. Several PFM reforms have been initiated in recent years with some results achieved but most have been suspended or simply not implemented. Progress and reversals on PFM in Guinea-Bissau continues to be fluid, much like the continuously changing political and institutional landscape. At times progress has been made, only to be slowed or reversed by political instability and changes in administration. - Budgetary preparation and execution has been supported since late 2008 by the implementation of the budgetary software system SIGFIP, and adoption of a budgetary nomenclature in line with the West African Economic and Monetary Union (WAEMU) standards. While the system is in operation, a significant share of government expenditure is still being processed outside the system. - The monitoring of budgetary execution had improved through the preparation of quarterly budgetary execution reports, which were regularly compiled from mid-2009 and communicated to the National Assembly and the public. This practice was abandoned in June 2012 but has been reinstated in early 2016. - Budgetary control procedures were revamped. In early 2010, the financial controller office was separated from the directorate of the budget and placed under the direct supervision of the minister. This reform was not pursued after the coup. - Efforts to rein in exceptional budgetary procedures in 2010 were abandoned after 2012, preventing any reliable budget execution reporting. However, extra-budgetary expenditures are eliminated since early 2017. - Progress has been made in addressing shortcomings in accounting, with the introduction of a new accounting system at the Treasury in September 2010. The chart of accounts was subsequently strengthened in 2017 with the support of the IMF. - Weaknesses in civil service payroll procedures had been addressed with the publication in July 2010 of the results of the civil service census. All personnel data were to be entered in the new system by the end of 2011, setting the basis for a more rational management of the civil service. As part of World Bank support to cover salaries of teachers and health workers in early 2014, the so-called Folha Unica (the financial model of the payroll) has been updated in these sectors. - In June 2010, the government adopted a new public procurement law, consistent with WAEMU rules and procedures. - The customs administration system ASYCUDA++ has been installed in the three main customs offices with external technical assistance, although currently only 10 percent of the system is fully operational. 3.1.1 Credibility of the Budget 4. Budget credibility remains weak. The 2013 PEFA noted a very slight improvement over the 2009 assessment while pointing out significant weaknesses in revenue projections. The budget execution rates over 2015–17 show significant variation between planned and actual revenues and expenditures, even on budget items that should be relatively easy to forecast, such as interest payments. Notably, there have been large swings in execution rates ranging from 43 percent in 2015 to 140 percent in 2016. While wages and salary payments were relatively accurate, transfers were at 118 percent in 2015 and 197 percent in 2016, and ‘other current expenditures’ reached 251 percent of the budgeted amount in 2016. Although a positive development, total revenues exceeded the forecast by 131 percent in 2016, while total expenditures were at 136.4 percent. The primary balance worsened between 2015 and 2016, reaching a deficit of roughly 2 percent of GDP from a projected 1.16 percent – a forecast error of 125 percent. 52 Figure 3.1: CPIA Scores on PFM Cluster Indicators, SSA and FCS (2016) Source: World Bank (2017). Table 3.1: PEFA Assessment of Guinea-Bissau 2009 versus 2013 P1 A. PFM outturns: Credibility of the budget Score 2009 2013 P1-1 Aggregate expenditure outturn compared to original approved budget D C P1-2 Composition of expenditure outturn compared to original approved budget D D+ P1-3 Aggregate revenue outturn compared to original approved budget D A P1-4 Stock and monitoring of expenditure payment arrears D+ NC B. Key cross-cutting issues: Comprehensiveness and transparency P1-5 Classification of the budget D C P1-6 Comprehensiveness of information included in budget documentation B C P1-7 Extent of unreported government operations D NC P1-8 Transparency of inter-governmental fiscal relations NA NA P1-9 Oversight of aggregate fiscal risk from other public-sector entities D D P1-10 Public access to key fiscal information D D C. Budget cycle C(i) Policy-based budgeting P1-11 Orderliness and participation in the annual budget process D+ NC P1-12 Multi-year perspective in fiscal planning, expenditure policy and budgeting D+ D+ C (ii) Predictability and control in budget execution P1-13 Transparency and taxpayer obligations and liabilities D+ C+ P1-14 Effectiveness of measures for tax-payers registration and tax assessment D+ D+ P1-15 Effectiveness in collection of tax payments D+ D+ P1-16 Predictability in the availability of funds for commitment of expenditures D+ D+ P1-17 Recording and management of cash balances, debt and guarantees D+ C P1-18 Effectiveness of payroll controls D C+ P1-19 Competition, value for money and controls in procurement D D+ P1-20 Effectiveness of internal controls for non-salary expenditure D+ D+ P1-21 Effectiveness of internal audits D+ D+ C (iii) Accounting, recording and reporting P1-22 Timeliness and regularity of accounts reconciliation NC D P1-23 Availability of information on resources received by service delivery units D D P1-24 Quality and timeliness of in-year reports D D+ P1-25 Quality and timeliness of annual financial statements D+ NC C (iv) External scrutiny and Audit P1-26 Scope, nature and tracking of external audits D D P1-27 Legislative scrutiny of the annual budget law D NC P1-28 Legislative scrutiny of external audit reports D D D. Donor practices D-1 Predictability of direct budget support D D D-2 Financial information provided by donors for budgeting and respective D D reporting on program and project aid D-3 Proportion of aid that is managed by use of national procedures NA NA 53 5. Overall, the budget is used in a very limited way to guide expenditure priorities . Instead, the Treasury Committee makes real-time decisions on expenditure based on cash availability and some prioritization criteria. Due to the political impasse, the National Assembly has not functioned between 2015 and mid-April 2018 and no government program or budget has been approved during this time. Government is thus formally working on the basis of the 2015 budget allocations. However, the Ministry of Economy and Finance (MEF) prepared annual budgets for 2016-2017 and annual expenditure ceilings are set on the basis of the average between 2015 and the year in question. Budget execution improved in 2017 with budget execution reports now being produced on a quarterly basis. However, data from SIGFIP do not reflect actual expenditure by institutions or category since these are generally retrofitted having been first processed manually and then entered into SIGFIP after significant delays. Weaknesses remain in the way that expenses are recorded. There are large discrepancies between different reports (e.g. budget execution reports and the Table of State Financial Operations; TOFE), or between the same reports over time. 3.1.2 Comprehensiveness and Transparency 6. There are notable weaknesses in classification and coverage . No precise data exist on extra- budgetary operations, but it is assumed that these continue to be substantial, given that the revenues of many ministries account for only a part of their spending. There is very limited information on the financial performance of parastatals, including any contingent liabilities and transfers from the government, except for the energy and water company which has recently begun to submit monthly financial statements to the Treasury. The Government is reportedly carrying out a mapping of state-owned enterprise (SOE) contingent liabilities, and recently carried out audits of the largest SOEs in 2018. 7. Access to public finance information is limited. The Government’s website has not been updated since 2015. The MEF’s website does not work, the National Assembly’s website has been removed from the host server, and the Auditor General’s website contains very limited and outdated information. In this context, comprehensiveness and transparency remain limited at best. 3.1.3 The Budget Cycle Policy-Based Budgeting 8. The Government has made attempts to increase the quality of budget planning and forecasting. Staff from the economic studies unit have started producing quarterly forecast reports (nota conjuntura). Line ministries are also preparing sectoral medium-term expenditure frameworks for health, education and public works. In addition, a detailed budget calendar has been prepared for the 2018 budget exercise with support from the World Bank, and a new organigram has been prepared with detailed task descriptions for the various units within the ministry. Despite these recent efforts, the gap between budgeted amounts and actual spending continues to be very large. Budgeting for capital and operational spending are distinct and unlinked processes, and in practice do not allow for recurrent expenditure to be estimated. Predictability and Control in Budget Execution 9. Expenditure control significantly improved in 2017 with the work of the Treasury Committee, but the quality of expenditure remains very weak. The functioning of the TC since 2016 has greatly improved cash management. The government now prepares rolling monthly cash flow projections, and bank reconciliations take place on a weekly basis. However, the way in which expenditures are prioritized and approved is questionable. In the recent past, the extremely constrained fiscal situation meant that the 54 government had no resources for anything other than salaries, and some basic goods and services. Overperformance in revenues in 2017 have created some fiscal space for the government. However, without a system or criteria to determine spending priorities, it was left up to the Minister of Finance to make these decisions, with only limited consultation with other ministries. Most spending decisions were taken by the minister and then communicated in the TC meeting. Some issues are debated and negotiated among committee members in the meeting. The basis for decisions are often unclear, and mostly based on the minister’s or committee members’ opinion, rather than previously agreed strategic priorities. 10. This is sub-optimal from an investment perspective since basing expenditure on immediate needs may not always be the most efficient method in terms of social or economic returns. Investment spending is almost entirely financed by donors and uses donor systems to appraise and manage projects. The methods vary greatly amongst donors, but are general assumed to be acceptable, at least for the largest donors. The Government itself has no system in place to plan and prioritize public investments. It has no guidelines or manuals for preparing public investment projects, nor any established criteria screening and selecting projects. The only criteria for determining whether to include a project in the three-year rolling Public Investment Program (PIP) is whether the proposed project is broadly in line with government priorities as outlined in their strategic plans, and whether funding is available. Government’s own investment is largely confined to counterpart funding of donor-financed investments, office refurbishments, limited purchases of equipment/furniture, and some periodic road maintenance. Tax and Customs Administration 11. Tax and customs obligations are generally clear and comprehensive, although implementation deviates greatly from the law. Taxpayers have access to information on tax and customs rights, although these are often negotiated directly in the case of customs. There is also an appeal mechanism against the tax and customs authorities’ decisions. While sanctions exist for tax violations, taxpayer registration is not subject to effective control. The sanctions mechanism needs to be substantially upgraded to have a real impact on tax compliance. There is an ongoing program of tax audits, although they are not based on coherent risk evaluation criteria. The audits have not led to a significant improvement in either tax collection or the amount of penalties collected for infringements. 12. The customs administration remains weak and fragmented, and low customs revenues are likely a result of significant leakages. Despite having an electronic system to manage customs procedures (SYDONIA), import and export documentation continues to be managed manually. Customs valuations are not based on standardized prices despite a WAEMU reference price database being available. Instead values and duties are determined jointly between the exporter/importer and the customs authority, the Direcção- Geral da Alfândega (DGA), and the rates applied are lower than they should be for a number of high-volume products. The HR structure of the DGA further complicates management and control. The DGA is staffed by personnel reporting to both the MEF and the Ministry of Interior, who see their roles differently. Debt Management 13. The data on public debt are reasonably accurate but the debt division personnel are not yet able to use the debt management system well enough to create regular external debt reports. The result is frequent late debt payments and penalties. The World Bank is financing an upgrade to the Système de Gestion et d'Analyse de la Dette (SYGADE) (or Debt Management and Analysis System) to SYGADE 6, including a comprehensive technical assistance and training program provided by the United Nations Conference on Trade and Development (UNCTAD). In June 2017, the Council of Ministers approved the 55 decree clarifying the contracting authority and procedures in relation to government debt and guarantees, bringing Guinea-Bissau’s legislation in line with WAEMU debt management regulations. Public Procurement 14. The legal framework for public procurement was revised and expanded in 2010 to ensure that it conformed to the WAEMU 2005 guidelines. The present procurement system has four main elements: (i) a public tendering code introduced in 2002 and revised in 2012; (ii) the Central Public Procurement Unit (UCCP), covering all public bodies; (iii) the Director General for Public Procurement, charged with scrutinizing calls for bids before they are distributed; and (iv) the Public Procurement Regulatory Authority (ARCP), responsible for oversight of the system, resolving complaints by operators, imposing sanctions for violations, and proposing appropriate reforms. Box 3.2: The WAEMU Budget Classification and Guinea-Bissau Guinea-Bissau’s current budget classification does not follow that of the WAEMU Directive N° 08/2009/CM/UEMOA of June 26, 2009 on the state budgetary nomenclature (NBE) within the WAEMU. In addition, it is no longer an application of the 1998 WAEMU directive. This partly reflects Guinea-Bissau’s limited administrative and institutional capacity and stability. It is important to note that even WAEMU countries with stronger PFM systems and institutions have not yet implemented many of the new WAEMU PFM regulations. The 2009 directive introduced numerous innovations, including: (i) The deletion of titles. In the old NBE directive, all revenue was presented in a single title, namely title 0, and expenditure in 7 titles. The new directive deleted the titles because their titles sometimes coincided with the titles of the articles. (ii) The introduction of functional classification. The old directive provided for functional classification in its annex, but its application was not mandatory. With the new nomenclature, this classification is made mandatory. (iii) The introduction of classification by program. The new directive modifies how the budget is presented. It embodies the programmatic classification which is the essence of the reform. The entire budget should be classified by program, i.e. by grouping appropriations intended to implement an action or a coherent set of actions representing a clearly defined policy in a medium-term perspective. (iv) The withdrawal of budgetary revenues, medium and long-term borrowings, repayments of loans and advances and proceeds from the sale of assets . The new directive removed the proceeds of medium- and long-term borrowings from the table of budgetary revenue, and the repayment of loans and advances and proceeds from the sale of assets. These transactions are classified as cash transactions. (v) The withdrawal of budgetary expenditure, the amortization of public debt and loans and advances . The 2009 directive has removed depreciation of the public debt as well as loans and advances from the table of budgetary outlays and classified them as cash flow. Guinea-Bissau has drafted a state budgetary nomenclature which is consistent with the 2009 directive, but has not yet implemented it. Its application will require drawing up the correspondence between the new nomenclature and that currently in use. This work, which is in progress, aims to facilitate its implementation. In addition, a team is working on the consistency between the proposed NBE, the state Chart of Accounts, and the TOFE to aid understanding and harmonize the content of each title or wording. Note that: (i) receipts are classified according to their nature; the optional source classification is not yet applied; and (ii) expenditure is currently classified according to administrative and economic classifications, the functional classification has only nine functions instead of the 10, and classification by program is not yet operational. It is therefore urgent for Guinea-Bissau to move towards implementing the new nomenclature to bring its budgetary and financial information in line with the WAEMU guidelines. 15. Compliance with the public procurement regulations remains low . The UCCP and ARCP are covered by the public procurement framework law and, their structures and functions are described in two 56 decrees dating from August 2012. Most ministries still fail to prepare annual procurement plans and many units circumvent the UCCP. The UCCP itself does not publish any statistical information on its work. It handled 24 procurement processes in 2014, 94 in 2015 and 106 in 2016. In 2016, 75 percent of the procurement processes were externally financed: 69 of the processes were done through open competitive bidding, 35 through a restricted shortlist, and 2 through direct contracting. Twenty were for equipment, 26 works and 54 for services. The supreme audit institution continues to play a role in the pre-approval of procurement contracts as is the case in most Portuguese-speaking African countries (PALOPs). 3.1.4 Accounting, Recording, and Reporting 16. The central integrated financial management information system is operational, but significant weaknesses still prevent its full use. The SIGFIP system only works in the MEF due to connectivity and capacity issues in other ministries. Most expenditures are first made outside the system through a traditional paper-based approval system and then recorded in the system afterwards. Even when payments are processed through SIGFIP, the approval process is replicated both on paper and electronically. Revenues are captured through other systems (SYDONIA, SYGADE), and are then entered manually into SIGFIP after some delay. A technical solution to transmit information electronically between the different systems was developed in 2017 but is not yet being used. 17. The lack of clarity around some important budget items, notably “other current expenditures” and transfers, makes it difficult to determine how public funds are spent with any certainty. The budget classification system is similar, but not identical, to the 1998 WAEMU guidelines (Box 3.2). Unlike other WAEMU countries, Guinea-Bissau’s economic classification includes a section called “other current expenditure”. Expenses recorded under this category reached 14 percent of total expenditure in 2016, and averaged 11–12 percent of total expenditure in 2010–16. The data on these expenditures are incomplete. Around 45 percent of this category is classified vaguely as “other common expenses” while over 54 percent are linked to incentive payments for revenue collection. All payments under other current expenditures are paid on behalf of, or for the benefit of, the MEF. These gaps call into question the reliability of public financial data in Guinea-Bissau. 3.1.5 External Scrutiny and Audit 18. External control by the Court of Auditors (Tribunal de Contas; TdC) and the National Assembly is non-existent. The last opinion (parecer) the TdC published on the government’s accounts dates from 2010, and government expenditure is not subjected to any other external audit mechanism. The audits the TdC normally conducts on some central government bodies account for less than 50 percent of total government spending. The TdC has been receiving support from a joint initiative between the UNDP and the EU to strengthen audit and oversight institutions in PALOP countries since 2013. Meanwhile, the legislature’s role is theoretically to scrutinize fiscal and related policies, and detailed expenditure and revenue calculations. However, the budget prepared by government is very general and contains no clear policy explanations, and the assembly does not have the capacity to provide its own assessment of the material. This makes the legislature’s oversight extremely limited. 3.2 The Treasury Management Committee 18. Treasury Committees have been functioning on and off for the last 15 years in Guinea-Bissau. One of the first committees, in operation back in 2002, was not entirely successful due lacking the necessary authority to implement policies because of political interference. Similarly, in 2005 and 2007, Treasury 57 Committees (TCs) were instituted by regulation from the prime minister’s offices with clear funding guidelines for the committee,8 and a monthly treasury cash management. TCs were constrained by political instability and lack of fiscal discipline. Their control of spending was inefficient as commitments were made by the Minister of Finance and sent for verification to financial controllers, contrary to best practice. The devolution of payment responsibilities to the line ministries’ financial administrators led to substantial non - regularized payments at the MEF. The existence of multiple government bank accounts was also not conducive to efficient cash management. In addition, most line ministries collected earmarked revenue that were not included/reported in the budget documentation. By 2010–12 the Treasury Committee was basically confined to the office of the Minister of Finance, before being discontinued in April of 2012. 19. The Treasury Committee was re-activated at the end of 2014. This time it had a two-tier structure: an upper level “plenary” committee was supported by an operational technical committee in charge of screening expenditure proposals, verifying cash availability, etc.9 During this period, efforts were made to streamline the expenditure chain to help restore fiscal discipline and transparency. It took 10 steps to execute a payment for budgeted activities10 which needed a good reporting system to keep up with the demand for requests. However, SIGFIP was not set up to provide the data needed to undertake expenditure decisions in a timely manner. This—combined with difficulties in forecasting revenues, and political pressures overriding agreed expenditure priorities—resulted in less-than-optimal allocation of financial resources, while non-regularized expenditures (DNTs) and some expenditure arrears continued. 20. At the beginning of 2017, a new dynamic was introduced into the functioning of the TC. The incentives for having a normal budget execution procedure (avoiding DNTs and arrears) and the political willingness to support these arrangement, at its highest levels, were reconciled. On the one hand, the decision makers at the MEF (the Secretary of the Budget and the Treasury) responded to the same technical goals and new orientations. The political authorities supported the new procedures, diminishing political interference in the functioning of the committee. In turn, the MEF’s directorates and members of the TC improved the provision of data on such things as authorization requests, approved requests, liquidations, payments orders, partial disbursements, and unpaid authorized requests. All requests were being recorded in SIGFIP before proceeding. As a result of these new procedures, DNTs and arrears in the budget execution were phased out in 2017. 3.2.1 The Current Functioning of the Treasury Committee 21. The TC receives from the Budget Directorate a list of all expenditure requests that were introduced into SIGFIP by the financial administrators of different ministries . These requests are sent to Financial Control for compliance, regularity, and approval; after examination by the Financial Controller an opinion is issued for the attention of the Secretary of the Budget, who authorizes a draft. These draft commitments are sent to the Budget Directorate for attribution, via Financial Control. The draft commitments are discussed at the technical level of the TC. The guidelines for authorizing the liquidation and payment orders of a request start with a categorization of expenditure by priority that was introduced in January 2007. Using these categories, the technical level of the committee recommends that the minister approves payments, trying to ensure that the limited resources available are allocated to the most essential 8 Members of the TC included the office of the minister of finance, the directors of the tax, customs, budget, economic studies and Treasury, as well as representatives of the UNDP, EU and the IMF 9 The operational level was mainly comprised of the directors of the Budget, Tax Office, Customs, Financial Control, Economic Studies, the Treasury, the Central Bank and the IMF representative. The upper level, intended to set policies, was comprised of the Minister of Finance and representatives of international institutions. 10 The 10 steps needed in order to use the resources allocated in the budget—starting from the request of the Administrative and Financial Directorates (DAFs; the financial managers of line ministries) to the effective payment of the expenditure by cashing the payment order— involved steps by the Budget Directorate (twice), Financial Control (twice), and the offices of the Secretary of the Budget and the Treasury. 58 functions. Thus, the Minister of Finance and his team (Secretary of the Budget and Treasury) are the de facto decision makers. 22. The expenditure categories range from Group A to Group G, where A is the highest priority and G the lowest. The expenditure categories in Group A are considered essential: wages and pensions, food for defense and security personnel, fuel for defense and security institutions, transfers to the public utility company, current debt amortization and interest payments, and service providers. Thus, expenditure in Group A will normally be approved first, with resources set aside for their payment, and only when additional resources become available will expenditures in subsequent groups be approved and executed. 23. While this system has ensured Box 3.3: Priority Expenditure Categories that essential expenditures (such as Prioritization of Spending Categories salaries) are paid regularly and on time, GROUP A (Highest Priority) SALARY EXPENDITURES in practice many non-priority expenditures Wages and benefi ts Pens i ons (foreign travel costs for government Nati onal As s embl y s al ari es Embas s y s al ari es officials, cost of medical treatment abroad FOOD Mi ni s try of Defens e for officials, and revenue collection Mi ni s try of Interi or Heal th Mi ni s try (Food for hos pi tal pati ents ) bonuses) are still given top priority. There is Jus ti ce Mi ni s try (Food for j ai l i nmates ) Food for the Pres i dent’s and PM s ecuri ty detai l also the question of whether the definitions Food for the Cuban doctors conti ngent FUEL of spending priorities and categories are Mi ni s try of Defens e Mi ni s try of Interi or Offi ce of the Pres i dent and of the Pri me Mi ni s ter adequate. Investment expenditures, for Raul Fol l ereau Hos pi tal EAGB (El ectri cy and wáter company) example, have been given the lowest level Fuel for TV and radi o broadcas ti ng centers Others (Mi ni s try of Fi nance) of priority, while defense expenditures are ELECTRICITY CONSUMPTION EAGB always executed without delays. Box 3.3 AGGREKO (Provi der of el ectri ci ty generators for EAGB) MANAGEMENT FUNDS details all the expenditure groups and For government i ns ti tuti ons For hos pi tal s categories as they are used in the TC. SERVICE PROVIDERS CURRENT DEBT REPAYMENTS Domes ti c i nteres t payments External i nteres t payments 24. There are still areas for Domes ti c pri nci pal repayments External pri nci pal repayments improvements, such as the need to use GROUP B Payment of arrears from previ ous months the TC to align expenditures with the real Members hi p and other i nternati onal i ns ti tuti on fees GROUP C expenditure needs in the areas which were Forei gn travel cos ts for di gni tari es , offi ci al s and ci vi l s ervants Payment of medi cal treatment abroad for di gni tari es , offi ci al s and ci vi l s ervants programmed by the sectoral ministries. To Mi s cel l aneous expendi tures GROUP D improve the efficiency and impact of the REPAYMENT OF ARREARS PC CEDEAO TC, in particular its fiscal discipline, it will be PCS UEMOA BOAD important to let the sectoral ministries to AMORTIZAÇÃO DOS AVANÇOS BANCÁRIOS GROUP E identify their priorities rather than the MEF. Revenue col l ecti on bonus es for MoF s taff PCS UEMOA – Regi onal tax This would contribute more towards PC CEDEAO – Regi onal tax IT Tax identifying the sectors that are being GROUP F Repres entati on s ubs i di es (Sa l ary top-ups for top offi ci al s ) benefiting with the execution of the budget GROUP G – PUBLIC INVESTMENT WITH DOMESTIC RESOURCES (lowest priority) 3.2.2 The Conceptual Framework for Treasury Management Functions 25. As fiscal challenges have emerged in recent years, Guinea-Bissau has engaged in PFM reforms, including the reinforcement of treasury management committee. Guinea-Bissau’s budget preparation and execution faces control challenges, does not align expenditure and revenue, and lacks budget reliability. There are also limitations in terms of cash flow forecasts and predictability in the availability of funds. This section aims to provide options for strengthening the TC to improve not only cash management but the quality and prioritization of expenditure. 59 26. The TC’s role in Guinea-Bissau goes beyond traditional treasury management. The TC is an institutional mechanism used to gather information on the availability of cash for approved expenditure on a rolling basis. For example, line ministers are no longer authorized to make spending commitments unless previously approved by the TC. This is partly due to the fact that Guinea-Bissau uses cash-based accounting and SIGFIP’s lack of robustness. The TC has improved cash management and expenditure control. However, it still faces challenges over prioritizing and approving expenditure under clear and transparent rules, especially when revenues overperform. The MEF has a powerful role in the decision-making process of the TC, where fiscal balance considerations sometimes overrule spending and investment priorities. Box 3.4: Factors Affecting the Budget Cycle External factors: Economic shock or natural disasters • Forecasts & Estimations • Budget formulation Preparation sequencing • Separations of • Hard or soft budget functions constraints • Reporting and audits Evaluation Voting & & Control Approval • Log Rolling on voting • Public Finance Management Execution process • Commitment control • Sequential or Single-Subject • Accrual vs Cash accounting Rule on voting process • IFMIS The extent to which cash inflows are able to offset cash outflows will be influenced by several factors, usually related to unbalanced budgets or distortions generated along the budget cycle. Cash shortfalls can appear as a result of an external shock, such as a decrease in economic activity, an increase in international interest rates, or natural disasters. On the other hand, cash shortfalls can be created in the course of the budget cycle because of budget distortions related to forecasting errors, legislative processes, or a lack of robust PFM systems. During the budget preparation phase, the lack of forecasting and estimation capacity, the sequencing of the budget formulation, or budget constraints can lead to unrealistic budgets in which actual revenues are not enough to pay for the estimated expenditures. This situation will eventually generate cash flow imbalances and cash shortfalls. The voting process can create unrealistic budgets as well, especially when the voting sequencing (single-subject rule) does not limit the phenomena of logrolling,11 which could inflate revenue forecasting in order to allocate more expenditure.12 Depending of the magnitude of the overestimation, the approved expenditure will impose pressure on the cash flow since the expected revenue will not materialize. In the same way, cash shortfalls can emerge when the PFM systems are not strong and commitment controls are not in place.13 Having a robust integrated financial management information system is important for reducing over- commitment and its potential impact on the cash flow. 27. Treasury functions usually need to cope with distortions generated throughout the budget cycle that could have an impact on cash management (Box 3.4). Treasuries characteristically have four 11 Logrolling is likely to surge when the legislature conducts the discussion and voting process in a “package,” by addressing r evenues and expenditures in parallel. When this happens, the legislature is more likely to vote on appropriations first and define the total budget size later. Doing this increases the possibility of inflating the budget. 12 Ian Lienert (2010) identifies also the “common pool” problem observed at the budget preparation, which may be large at the parliamentary approval stage, in which the parliament is prone to introduce changes that increase spending or reduce taxes. For more reference: “Role of the Legislature in Budget Processes” by the IMF. 13 Flynn and Pessoa (2014) mentioned that expenditure arrears could be the result of failures at any or all stages of the PFM cycle, including unrealistic budgeting, among others. 60 core functions: (i) receiving the inflow of funds from the organization’s revenue stream (tax collection in the case of government, sales in the case of a private company); (ii) managing the outflow of funds (payments to suppliers, etc.); (iii) managing financial assets (cash); and (iv) managing liabilities (debt). The standard model for treasury management thus implies managing two flows and two stocks (Figure 3.2). 28. In principle, for each budgetary transaction there is a mirror transaction in the financial side, which implies a cash operation at the Treasury for each budget transaction (Figure 3.2). Revenue administration and collection, and debt issuances, will create cash inflows to offset the cash outflows (payments) created by budget expenditures. When budgets are unrealistic and unbalanced, cash inflows and outflows will not match and one possible outcome will be the shrinking of cash stocks and the increase of debt stocks. Cash may need to be rationed and payment arrears accumulate. Therefore, treasury functions aim to conduct cash management and liquidity risk management by monitoring the cash position and the demand for payments. Figure 3.2: Treasury Core Functions: Managing Flows and Stocks Source: World Bank. Figure 3.3: Budget-Treasury Relationship Source: World Bank. 29. Since the TC in Guinea-Bissau manages cash flow to comply with government obligations, it is important to identify unrealistic budgets or the lack of budget control in times of cash shortfalls. 61 To this end, the TC gathers information on revenue inflows to plan for expenditure outflows. The TC issues a weekly report including treasury information on cash stocks and cash liquidity, and cash inflows and outflows, but also budgetary information on the composition of revenues and expenditures, historical comparisons and trends, and short-term obligations and payment arrears (restos a pagar) following a treasury management logic (Figure 3.3). This way, the TC evaluates progress in tax collection, expenditures, and the liquidity position for borrowing decisions. Box 3.5: Treasury Management in the Face of Unrealistic Budgeting Some Challenges for Treasury Committees Inputs (FY) Outputs Realistic budget Unrealistic budget (End FY) • Budget control not conducted by the • Budget control conducted by the Treasury Treasury based on cash rationing • Treasury focus on cash/debt certain level of payment Balanced management and payments prioritization Budget • Characterized by inter-annual cash • Limited cash accumulation, accumulation => Money Market Desk moderate generation of payment for cash optimization with surplus arrears, unplanned debt, budget investments and debt issuance reliability affected • Budget reliability is not affected • Moderate use of debt • Inexistent budget control and • PFM weaknesses intensive cash rationing • Limited cash accumulation and Unbalanced • Payments arrears accumulation and limited commitment control Budget waiting list for payments • No budget control on the treasury • Public services highly affected • Payment arrears and waiting list for • No cash accumulation, rapid debt payments increases, budget reliability highly • Unplanned debt to offset unbalances affected Treasury functions are usually shaped by the fiscal challenges, budgetary practices, or PFM weaknesses they face. The oversimplified matrix above does not reflect causality but is a characterization of treasury functions under conditions of unrealistic versus realistic budgets, and the different treasury functions and responsibilities needed to achieve balanced versus unbalanced budgets. When facing unrealistic budgets, the functions demanded of the Treasury usually reflect its need to take an active role in controlling the budget through cash rationing to achieve a balanced budget at the end of the fiscal year. In this situation, governments are likely to experience limited cash accmumulation, moderate payment arrears, some unplanned debt to offset cash imbalances, and diminished budget reliability. In contrast, when a realistic budget is formulated and approved, the Treasury can take a more passive role since there is no need for budget control at the treasury level to achieve a balanced budget at the end of the fiscal year. In this situation, treasury functions are characterized by a greater focus on cash management and payments without cash rationing; cash is accumulated seasonally, facilitating cash optimization and programming; and budget reliability is not usually an issue. 30. The governance of the TC aims to create an institutional mechanism for centralizing fiscal and treasury information for decision making. The committee meets on a weekly basis and is chaired by Minister of Economy and Finance. It is comprised of several stakeholders such as the state secretaries, the 62 general directors, the chamber of commerce, and some representatives of the multilateral agencies such as the IMF, World Bank, the UNDP, and the EU. The TC is assisted by a secretary who oversees the gathering of technical information from several sources including SIGFIP and the revenue collection units. The TC issues a weekly report based on liquidity management for decision making ( Relatório Técnico Semanal) elaborated by the General Directorate of the Treasury (Direção Geral do Tesouro). 31. The major challenge the TC faces in Guinea-Bissau is dealing with a non-approved budget that does not reflect government priorities . Such a situation usually demands a more active role from the Treasury in managing and controlling budget execution (Box 3.5). When budgets are unrealistic, the Treasury usually faces a continuously low cash stock provoking cash rationing. Therefore, treasuries adopt a role beyond traditional treasury functions, such as participating more actively on the points of control for budget execution. For instance, they may issue cash ceilings to control commitments based on cash availability with the objective of limiting the generation of payment arrears and unexpected debt increases. The TC thus takes on a more active treasury role to control the budget. 32. Cash is monitored based on the revenue received and programmed expenditure. The TC monitors cash inflows using information about the revenue collected through commercial bank accounts, and estimates cash outflows using programmed expenditure. This information helps to estimate the overall cash position of the government. The TC can thus estimate the funds available and authorize credit releases to execute the budget and make payments. The TC also conducts trend analyses of government cash flows based on historic monthly data, and by comparing cash positions in previous years. This allows the TC to analyze the estimation errors between planned, collected, and effective revenues (collected revenues without reimbursements) in each of the revenue collection units (tax, customs, and fisheries). 33. Cash outflows are controlled by defining a ceiling for payments and commitments . The TC estimates the cash demand from the previous month’s authorized commitments and the current month’s payments, which allows it to estimate the demand for cash expenditure. The TC also has information about the commitments generated in each of the budgetary classifications: salaries, goods and services, transfers, current expenditure, debt service (capital redemptions and cost of interest), capital expenditure, and payments arrears. This information allows the TC to assess the evolution of the cash position, the liquidity at the Treasury, and financing needs. Finally, the TC monitors the accumulation of payment arrears of more than 90 days, benchmarking indicators for the management of short-term obligations. 34. Limited information and a lack of forward-looking analysis mean the TC faces major challenges to achieving its objectives. Although the TC analyzes the cash position weekly and programs spending by the line ministries accordingly, there are still non-programmed spending and delays in identifying the concepts generating cash outflows. The TC could consider producing more forward-looking analysis and strengthening its monitoring tools for controlling commitments. 3.3 The Integrated Financial Management Information System: SIGFIP 35. SIGFIP was developed and implemented by the Société Nationale de Développement Informatique (SNDI) company of Côte d'Ivoire in 2008–10. It is a proprietary system also in use in a number of WAEMU countries and is based on the principles of budgetary and financial management. As an off-the-shelf software package, it offers generic features and has not been customized to the specific needs of the country. SIGFIP was developed in French and has been gradually translated into Portuguese over the years, although most of the screens and reports still use many French terms. Although it is described as a PFM system, it only encompasses some of the components of a state financial 63 administration system; for example, it lacks any state property and internal control components, as well as integrated accountability and planning components with their associated instruments, tools, and policies. 3.3.1 System Modules 36. The SIGFIP consists of 3 modules: - The Budget Preparation module allows the initial charging of the state budget forecast for the next fiscal year by the General Budget Directorate and the definition of thresholds for each of the ministries, taking into account revenue collection forecasts and goals established by international regulatory bodies and other economic partners. Ministries can also participate in the budgeting process by identifying their expenditure needs. - The Budget Execution module allows the Administrative and Financial Directorates (DAFs) within line ministries to register and follow up the three phases of expenditure (commitment, settlement, and payment). They can then be validated and authorized by the Financial Control, General Directorate of Budget, and Treasury. Expenditures recorded under the “normal process” go through the three phases outlined above, while personnel or public debt-related expenses, where the settlement phase is recorded in an automatic way, go through the “simplified process”. In the past, DNTs could also be registered but currently all state expenditures are first registered in SIGFIP. - The Accounting module allows the recording and accounting control, based on the Plan of State Simplified Accounts (SBB), a process of expenditure execution only part on the payment stage. 37. SIGFIP’s Budget Preparation module has some limitations. Some are due to lack of functionality such as the ability to predefine budgetary limits that guide the introduction of values by economic nature/sector and guide the preparatory work. Others are because it is not intuitive and easy to handle some aspects, forcing the intervention of the computer area to load data from Excel files. 38. The business model implemented in the Budget Execution module is bureaucratic and complex due to the use of duplicated physical processes. An ongoing assessment of SIGFIP, supported by the World Bank Public Sector Restructuring Project, indicates that payments are processed in parallel through both the system and the old paper-based process. For example, the budgetary availability controls in the system are validated several times, which does not add any value to the business chain. All steps of the administrative process require information to be printed out to be signed and stamped (often by more than one person), generating costs from printer consumables and paper. In contrast, some actions and decisions are taken outside of SIGFIP, such as controlling financial availability or prioritizing expenses to be paid at any time, either because there is no integration of information from other systems or due to the need to maintain discretion in decision making. 39. The MEF has proposed changes to the responsibilities, organization, and functioning of the Directorate General of Budget with a view to strengthening it. However, this proposed reorganization still maintains the workflows that feed SIGFIP, which require multiple approvals of expenditure within the MEF, on paper and in the system itself. The MEF needs to propose a simplified budget execution processes, reducing the number of verifications and approvals needed, and cutting down paper handling. 40. The Accounting Module records do not guarantee full accountability as they do not include the entire expense cycle. One of the most worrying issues with the use of SIGFIP is related to decisions authorizing payment (payment warrants). A budget implementation monitoring system has to cover revenue collection and expenditure realization according to the principle that “the revenue is foreseen and 64 the expenditure is fixed”. It needs to ensure that the balance of the Treasury Single Account is equivalent to the bank balance, adjusted to account for any compensation and consolidation in transit. As revenue is not automatically entered into SIGFIP, the bank and bordereau excerpts of the Central Bank (BCEAO) and other banks holding public funds need to be consulted each day before issuing new payments, to ensure there are enough funds to cover them. The introduction of effective financial programming in SIGFIP, bank reconciliation, and the creation of the Single Collection Document (DUC) to cover all revenue collection would bring enormous benefits in the process of consolidating the ongoing public finance reforms. 41. Another fundamental aspect is accountability – SIGFIP should be able to produce the reports and mappings foreseen by law for the presentation of the General State Account. For the accounts presented by government to be considered credible, they must be extracted automatically from the system, making them less vulnerable to manipulation. The State Financial Operations Table (TOFE) and other financial statements should also be directly extractable from the system. Yet, as noted above, frequent and significant adjustments are made to budget expenditure reports. This, combined with the lack of external audit and control, puts the reliability of the budget expenditure data into question. In addition, efforts must be made to ensure that information-generating entities automatically and regularly feed into SIGFIP. This applies not only to the revenue collecting bodies and bank accounts, but also for debt management. 3.3.2 Technological Infrastructure 42. SIGFIP was developed using Oracle Forms and Reports 10g and has as data repository the Oracle 10g Database. It is developed using a 3-tier architecture, with well-differentiated data, application and front-end layers. Despite being accessible from a web browser, its tools are not entirely web-based as is the current trend in information systems. This encapsulation is managed through a Java applet installed on the computers on which it is accessed. The database and application servers are in a dedicated room, but this does not have some of the fundamental requirements of a data center, lacking temperature and humidity control, physical access control, and system fire detection, among other measures. 43. SIGFIP is available through a local area network (LAN), which prevents access to users outside the MEF premises. An experimental network connection was installed to the DAF of the MEF, which is physically close, but this connection is not always operational according to the users. It would require a different version of the application to make SIGFIP available through the Internet, and would also need upgraded infrastructure at the MEF, including additional security equipment. Users from other ministries must physically go to the MEF to use the system either to register their expenditure documents or to simply consult a particular process. Apart from not being comfortable, it is also logistically not ideal because it forces users to walk around the city with the documents in hand. The equipment in this room is often not operational or missing components, thus preventing its use. 44. The MEF has three informatics directorates: the Service Management of Integrated Public Finance Management System, which is responsible for SIGFIP, and one each for Customs and Taxes. However, staff technical skills and training are very limited, especially at the level of development tools that would allow local technicians to make changes to the system. This has made the country totally dependent on the company that developed it for maintenance of the system. Although MEF technicians have shown their motivation to learn, their knowledge of the rules of business is still quite limited, partly due to the absence of technical documentation left by the company. 65 3.3.3 The Way Forward 45. Developing a new state financial administration system will be critical. This is particularly due to: (i) the need to adopt new functional models such as those required by WAEMU directives (budget by program or equity accounting, implying the standardization of new budget classifiers and the adoption of a new chart of accounts); (ii) the introduction of a financial programming model that will consolidate the implementation of the CUT, streamline treasury management and define expenditure prioritization criteria; (iii) the need to drastically simplify business rules in order to improve budget execution processes; reorganize the functional competencies of the Budget, Treasury and Accounting areas; involve sectoral and internal control bodies; and reduce operating costs; and (iv) the need to ensure it hold all the necessary information for correct and timely accountability, based on pure accounting, and interconnected with or incorporating data from other information systems. Table 3.2: Applying the Lessons Learned from PFM Reforms in Fragile States to Guinea-Bissau Lessons Implications for Situation in Guinea-Bissau post-conflict countries Progress on PFM Determine the level of Guinea-Bissau’s commitment level is partial and fluctuating. The need for reforms is possible if government international recognition and financing are drivers for reform, but past experience incentives are in commitment to reform: has shown that politicians have chosen short-term benefits over longer-term place. (i) substantial (ii) strategic objectives. The current administration seems committed to regaining partial and fluctuating legitimacy with donors and is hoping for an eventual return to budget support. commitment (iii) weak Recent advances on expenditure controls are partly driven by this objective. As in or absent many other FCSs, there is a high turnover of finance ministers and donors must deal commitment. with a fluid reform environment with windows of opportunity constantly opening and closing. Budget execution Work across the This is also the case in Guinea-Bissau, where the Minister of Finance prioritized reforms progressed budget cycle with a improving expenditure controls. Budget execution is now tightly controlled through further than budget core focus on the recently established Treasury Management Committee. A Single Treasury planning or execution reforms. Account is also being put in place and bank reconciliations take place on a weekly accountability basis. However, tighter controls on spending does not necessarily indicate a reforms. commitment to improve the quality of spending and public services. It could merely reflect a desire to get a tighter grip on limited public resources in order to maintain previous patronage practices. Transparency and accountability of spending has received less attention and have not improved. PFM reforms tend to focus on finance ministries and other core institutions such as tax and customs, with less participation from other line ministries and accountability institutions such as parliament and control organs. Short-term capacity Sustained attention to Guinea-Bissau is relatively dependent on expatriate technical advisors in key substitution has developing capacity positions such as tax administration and customs. Each new minister brings in new been easier than and a stronger link to heads of department with relatively high levels of capacity, but they lack the technical building longer-term overall human staff to implement reforms. The extremely low capacity of existing staff is apparent. capacities. resource reforms. Low public sector salaries and irregular salary payments make it difficult to attract competent staff, and there is a near complete lack of a core cadre of technicians to maintain institutional memory and continuity. There is no Focus on identifying There is a tendency of donors to push for optimal or standard PFM reforms in standard sequence bottlenecks and Guinea-Bissau. Reform priorities have depended on the ministers in charge and of reforms; reform pursue realistic reform their power to impose their views on the institutions in their remit. The current focus design and steps rather than on expenditure controls makes sense in Guinea-Bissau given the previous practice progression depend pursuing optimal or of high and increasing extra budgetary expenses. It has been more difficult for the on existing systems standard sequences. minister to make progress on broadening the revenue base, since the tax and and commitment by customs departments are only partially under the MEF’s control. The focus should political and now be on consolidating recent progress on expenditure control and cash technical leaders. management, and building on this to improve the quality of spending with a view to improving public service delivery. Source: Fritz, V. (2012). 66 46. SIGFIP’s current technology base has several weaknesses that may jeopardize continuity of service. The new development should safeguard technological developments, in order to: (i) ensure a faster and more reliable system by migrating the database to Oracle 12c and redesigning the application to use a fully web-based interface, regardless of the server-side programming language; (ii) make it accessible to all users by moving to fully web-based access; (iii) ensure interconnection with the other systems and services with which it is necessary to establish exchanges of information, in a safe and timely manner; and (iv) train national technicians so as to eliminate technological dependence on third parties. The development of a new system should involve national technicians in the MEF in the definition and validation of business rules, and involves its computer technicians in the system development and exploitation team. 3.3.4 PFM Reforms in Fragile and Conflict Affected States 47. Taking stock of PFM reforms in other post-conflict countries provides some useful lessons. Progress can be achieved by focusing limited capacity on a realistic number of technically achievable reforms that enjoy clear political support. These lessons show that most progress was made on expenditure controls, since this tends to be the area of most immediate concern by reform-minded ministers of finance. This is also the case in Guinea-Bissau (Table 3.2). However, the tendency of donors in such unstable environments is to continue to drive broad and overly ambitious reform agendas without clear priorities that take into account the technical capacities or the political commitment for such reforms. The result is that recipients are often overwhelmend by a vast reform agenda, and in their desire to attract funding they often overcommit and underdeliver on an agenda that was not realistic in the first place and commands limited ownership. The stocktake also shows that there is no standard or prescriptive sequence of reforms, and that donors need to remain flexible and adjust reforms where there is real traction. Guinea-Bissau needs to build on its recent progress on expenditure controls and engage other line ministries to further extend its PFM reforms with a view to improving the quality of its spending and public service delivery. 3.4 Policy Recommendations 3.4.1 Strengthening the Treasury Committee 48. The Treasury Committee has proven to be a useful control on government spending. The TC’s analyses are based on actual cash receipts during the week, which allow it to effectively establish spending ceilings based on the commitments and payment obligations generated over a very narrow time horizon. In order to improve cash planning, international practices recommend running cash forecasts on at least a three-month rolling basis. This usually facilitates expenditure execution but with weak PFM systems, this increases the risk of overcommitting and generating payment arrears and debt. The TC should therefore aim to enhance its analytical capabilities paired with budget controls and clear rules to prioritize spending. Capacity Building for Cash Forecasting and Liquidity Risk Monitoring 49. Guinea-Bissau will need to improve its cash planning and liquidity monitoring to develop a cash management model. The TC could benefit from expanding its time horizon for cash planning, which will provide the Treasury with better analytical tools to cope with fiscal imbalances and improve budget execution, such as the rolling out of cash forecast for at least the next two or three months, and liquidity risk monitoring. A new cash management model should include establishing incentives for the budget executing units to control commitments, and defining a prioritization rule. For this, it will be important to build the analytical capacity of the Treasury and the TMC. 67 50. A basic cash forecasting tool could be developed based on historical treasury information instead of the non-approved budgets. The TMC has historic information of inflows and outflows based on the cash demands and the seasonal behavior of the revenue collection, captured in commercial bank accounts. This is rough information that could allow to create a simple model based on historical data and seasonality analysis to forecast cash flows in a two or three month rolling basis. This model could help the TC to hard-tune the cash position, while using the weekly analysis to fine-tune its cash management. However, given the limitations of the information, any cash forecasting will be likely to contain large errors. 51. The cash forecasting exercise should be paired with a cash buffer policy to prevent selective debt defaults or delays in debt servicing. Guinea-Bissau has incurred selective debt defaults or delays in paying debt service. A cash forecasting model with longer horizons could allow it to define a cash buffer policy linked to its debt service requirements. This would help the Treasury make forward-looking estimations for cash planning as part of a policy to reduce liquidity risk. Empowering Executing Units to Control Commitments and Prioritize Expenditure 52. The Treasury’s estimation of available funds should be complemented by incentives for the executing budget units to control spending commitments. Some countries have established an aggregate budget ceiling controlled by line ministries and pass the responsibility for commitment control to the budget executing units by making it mandatory to prioritize short-term commitments like payment arrears, current expenditure for service delivery, and capital expenditure in the calculation of the credit releases for monthly payments. Accordingly, the MEF would mitigate the risk of over commitment while at the same time reinforcing accountability of the budget executing units. In this regard, Guinea-Bissau could consider the example of Portugal, which uses a formula to authorize commitments based on cash availability and previous commitments. 3.4.2 Improving PFM Systems 53. Guinea-Bissau’s most urgent priority is to improve the reliability of its public finance data, and increase transparency. Without reliable public finance data, it is impossible to evaluate the impact of public policies, or whether spending reflects agreed priorities. The essential purpose of good PFM is to ensure that public finance data are reliable, and then to ensure transparency around public finance and hold public officials accountable for their management of public resources. Public finances in Guinea-Bissau are managed in a very opaque manner and there are no systems in place to hold officials to account. The result is a government that appears to manage public funds for its own benefit with impunity and little regard for the public they are supposed to serve. 54. The Government may want to ensure that SIGFIP is used for the entire expenditure chain, including commitments and payments, and discontinue parallel paper-based payments where possible. The use of the system should be supported by the correct recording of budgetary data. It also needs to ensure the automatic exchange of information between existing systems to minimize errors in manual data entry. The government should continue its good practice of weekly bank reconciliations and explore the possibility of setting up an inter-bank real-time gross settlement system to facilitate automated bank reconciliation. To bring greater coherence and transparency to the wage bill, the HR databases of the MEF and the Ministry of Public Administration (MPA) should be integrated and government should centralize responsibility for HR management in one body, with the powers to manage all HR-related benefits and expenditures. 68 55. It will also be important to strengthen political leadership on PFM reforms . The MEF should take the political lead on PFM reforms. Given the political importance of PFM reforms in Guinea-Bissau, political leadership is essential. The starting point should be the development of an operational plan for PFM reforms, with clear and measurable indicators to be monitored. Regular steering committees should then be held to measure progress, take strategic decisions, and resolve bottlenecks. 56. The expenditure criteria used by the Treasury Committee should be strengthened to enhance the overall quality of spending. Given the broader PFM role adopted by the TC, it is urgent that it review its current expenditure criteria in order to strengthen the quality and progressiveness of expenditures, with a view to increasing public spending on frontline public service delivery rather than on benefits for central government staff. 57. Finally, the Government should consider improving budget realism and make the budget a meaningful policy document based on real political negotiation and consensus, and use it for determining expenditure priorities. This will require political leadership and engagement in the budget process, to make sure it is filled with substance and that the final budget document enjoys the broadest possible political consensus and buy-in at the technical level. In the short term, the government should develop a priority matrix to help guide/inform and rationalize public expenditure decisions taken in the TC. In the medium to long term, the government should strengthen the public investment system by adopting guidelines for project preparation and clear criteria for project selection to be included in the budget. 69 Policy Recommendations – Education Measure Technical Timeframe of feasibility policy impact Budgeting preparation and execution - Build planning and budgeting capacity in the Ministry of Education through providing training and Medium Medium term developing procedures manuals that document the mandates and responsibilities of each entity. - Improve budget transparency by: (i) making financial information and service delivery performance High Short term available to the public; and (ii) documenting the distribution and use of school fees. - Improve payroll controls by: (i) procuring SIGRHAP for the MPA and MOE; (ii) updating the staff Medium Medium term profiles in the MPA database; (iii) addressing staff turnover and parallel work in the private sector. - Ensure value for money in public procurement by: (i) preparing a procedures manual, with emphasis High Short term on open and competitive procedures, and monitoring of results, and (ii) building staff capacity. - Reinforce the standards and procedures applied in internal audit by providing the General Inspection High Short term with the necessary budget and ensuring its operational independence. Financial information technology systems - Work in coordination with the MEF to connect the MOE to SIGFIP on a pilot basis. Medium Medium term Improve the efficiency and equity of resource use - Prioritize and increase expenditures. Increase the budget share of non-salary school-level spending High Medium term to improve learning outcomes, in particular basic education. -Implement programs and measures to reduce the incidence of out-of-school children. Institute pro- High Short term poor educational policy intervention program focusing on marginalized communities. Human resource management - Establish clear criteria for human resource allocation. Base teacher staffing on pre-determined set Medium Medium term of criteria including student-teacher ratio, classrooms, school size, and availability of facilities. - Provide incentives to teach in remote areas. Implement an incentive scheme that provides a package Medium Medium term of accommodation and facilitates career progression for teachers serving in remote areas. 70 Chapter 4: Education1 Guinea-Bissau has set out an ambitious education sector strategic plan, but this has been hampered by fiscal constraints and political fragility, among others. Despite increased spending and significant progress on improving access to basic education, outcomes remain poor relative to peers, and the distribution of resources regressive. The sector suffers from the same public financial management (PFM) challenges as the rest of the country with limited budgeting capacity, lack of transparency, and lack of compliance with financial procedures. To improve education performance and achieve better value for its limited resources, the Government may want to: (i) build budget preparation and execution capacity, and improve budget transparency; (ii) upgrade its payroll, procurement and audit procedures, as well as its accounting and management systems; (iii) improve the efficiency of public spending and devolve it away from the center; (iv) further improve access to education and reduce drop out and repetition rates; and (v) improve the distribution of teachers, particularly in remote and rural areas. 1. The chapter is organized as follows . It first presents an overview of Guinea-Bissau’s education system, focusing on sector strategy. Section 4.2 analyzes the performance of the education system across different dimensions, including completion and retention Rates, learning outcomes, returns to education, and distribution of resources. Section 4.3 discusses key trends in public education spending, its alignment with national priorities, and benchmarking against relevant peer countries. Section 4.4 discusses the strengths and weaknesses of public financial management (PFM) in the education sector and sheds light on links with service delivery. Finally, Section 4.5 concludes with policy recommendations. 4.1 Overview of Guinea-Bissau’s Education System2 2. The organizational structure Table 4.1: Guinea-Bissau’s Education System of the education sector is Grade Age Level characterized by a largely centralized (official) management structure. The education n/a 3–5 years Pre-school sector is managed by the Ministry of 1–4 6–9 years EB1 Basic education Education and Higher Education, and 5–6 10–11 years EB2 Youth, Culture and Sport (henceforth, 7–9 12–14 years EB3 MOE), responsible for: (i) the national 10–12 15–17 years Senior secondary education education policy and strategy n/a Technical and professional education formulation, planning, monitoring and n/a 18–21 years Higher and tertiary education evaluation; (ii) operations and 1 Prepared by a team consisting of Fiseha H. Gebregizabher (Economist, GMTA2), Sylke Von Thadden (Consultant, GGO), Nicolas Drossos (Consultant, GGO), and Nelly Rakoto-Tiana (Consultant, GED). The team has received useful inputs throughout the process from Kebede Feda (Senior Economist, GED07), Bernardo da Bernardo da Cruz Vasconcellos (Education Specialist, GED07), Patricia De Sousa Robalo P Rosa Merle (Consultant, GED07), and Giulio Schinaia (ODI Fellow, Guinea-Bissau Ministry of Education). 2 The PER is based on a mix of quantitative and qualitative data analysis, including survey and interviews conducted during field visits to five regions of Guinea-Bissau (Oio, Cacheu, Bafata, Quinara, and the Autonomous Sector of Bissau).2 The main data source includes: (i) the EMIS for 2014-2015; (ii) two household surveys, carried out in 2014 (MICS) and 2010 (ILAP), (ii) the BOOST data for 2010–17, and (iii) sector documents (Country Status Report (2015), PASEC (2015), and the draft Service Delivery Indicator report (2018)). However, there were several limitations: (i) the current budget nomenclature does not allow education spending to be disaggregated by level of education; (ii) the information system does not allow tracking of funding to regional offices, and the collection, distribution and use of school fees; (iii) recent data on technical education and professional training and higher education, teaching and non-teaching staff were not available. Thus, some of the analysis in this chapter uses data from the 2015 Education Status Report (RESEN). 71 administration; and (iii) supervision and control.3 Even though the ministry is partially decentralized with nine regional directorates of education (DREs), corresponding to the administrative division of the country, and two additional educational sub-regions (Ingore and the Bubaque Islands), the decentralization of fiscal management and resources to the regional directorates and schools is still lacking. 4 3. Guinea-Bissau’s formal basic education system is centered around three cycles (or nine years of schooling), based on the 2011 education reform (Table 4.1). Grades 1-4 (EB1) are the first cycle, Grades 5-6 (EB2) the second cycle, and Grades 7-9 (EB3) are the third cycle. The first two cycles, which are compulsory and free, constitute the full cycle of primary education. The third cycle is comparable to lower secondary school and requires students to pay a fee. Basic education is followed by three years of secondary education, both general and technical schools, before four years of higher and tertiary education. 4. The public sector is the main provider of basic and secondary education, and is characterized by low quality. Out of nearly 460,000 students in 2014/15, 68 percent were enrolled in public schools, 15 percent in private schools, 15 percent in community schools, and 2 percent in Islamic schools ( madrassa). Despite efforts to accommodate more students, private schools remain the largest provider of pre-primary education, representing 39 percent of children enrolled in 2014/15. At the basic and secondary education level, public schools accounted for 70 and 80 percent of enrolment, respectively. The private sector plays an important role in the provision of technical education and professional training and higher education. The performance of the public school system is low, reflecting lack of investment and reforms. Private schools, notably in urban areas, tend to perform better.5 The quality of teaching is, however, not necessarily better, as some of the public sector teachers also work in private schools. Guinea-Bissau’s Education Sector Strategy: Goals and Priorities 5. The government has set out an ambitious sector plan to achieve universal primary education . With the support of the UNESCO team in the Pole of Dakar, the Ministry of Education has developed an Education Sector Plan (Plano Sectorial da Educação; ESP) for the period 2017–25, identifying the challenges the sector faces and the necessary reforms. The plan focuses on: (i) increasing access to basic education; (ii) improving the quality of education at all levels; (iii) developing training (technical, vocational, and higher education) appropriate to the needs of the country; and (iv) strengthening the governance of the sector by improving monitoring, decentralization, and financial management. The ESP is implemented through three triannual plans with the first plan covering 2017–19. It is expected to be partially funded by available resources and by additional funding mobilized by the Global Partnership for Education (GPE). 6. Successful implementation of the ESP will depend not only on increased investment, but also on political and institutional stability . The MOE identified five priority actions for its first triannual plan (2017–19). These include improving financial and human resource management, strengthening the Education Management Information System (EMIS), and pedagogical supervision. Implementation of the ESP has been slow so far due to delays in preparation of the plan and mobilization of external aid to support some of the priority measures such as the in-service teacher training, school inspection and monitoring and 3 Until 2015, the ministry’s operation also included the Secretary of State budget for Culture, Youth and Sports (Ministry of N ational Education, Culture, Sports and Youth). Between 2015 and 2018, 2015, the Government created a separate Ministry of Sports, Youth, and Culture but which was merged with the Ministry of Education again in 2018. 4 There are also autonomous institutes attached to the central administration, including the National Institute for the Development of Education (INDE), the Institute for Technical and Vocational Training (INAFOR) (conducts politics) and the National Institute of Studies and Scientific Research (INEP), the public university Amilcar Cabral (UAC), and the National Public Administration School (ENA). 5 PASEC (2015), Draft SDI (2018). 72 evaluation system, and sector management capacities. 6 Government has made important progress in reforming the inspectorate system, laying the foundation for improved teaching, but more sustainable and increased funding is needed. Moreover, high staff turnover caused by political instability, coupled with frequent teacher strikes, has severely undermined efforts to improve the education system. 7 4.2 Performance of the Education System 7. Over the past years, Guinea-Bissau has made significant progress on access to primary education, but attendance rates are still lower than other SSA countries. The primary gross enrollment rate increased at an average of 4.7 percent a year between 2000 and 2010, from 71.8 percent to 113.7 percent. This was significantly higher than the SSA average of 1.7 percent. Guinea-Bissau also achieved gender parity in the first cycle of primary education in that period. However, the gender gap is still relevant for lower and upper secondary education where girls’ attendance is lower by 10 and 14 percentage points. Gross enrolment rates in EB1 are significantly higher than 100 percent, driven by late entry into Grade 1 and high repetition rates within the first cycle of primary education. Guinea-Bissau tends to have a lower net primary attendance rate than its peers and other SSA countries. 8. Access to secondary education differs significantly across regions as well as between urban and rural areas. Although regional primary gross enrolment rates are relatively high, only three regions (Bolomo, the Autonomous Sector of Bissau [SAB], and Biomba) registered gross attendance rates above the national level in 2014. Access to secondary education by region is somewhat mixed. Gabu, Bafatá and Tombali regions have gross attendance rates below 25 percent, whereas SAB recorded a ratio of 115 percent. Furthermore, gross primary attendance in rural areas is 37 percentage points lower than in urban areas. This gap further increases to 77 percentage points in lower secondary education. Late Entry and Repetition Table 4.2: Percentage of Repeaters (EB1) Disaggregated by Grade and Gender 9. Due to late entry and high repetition rates, only 13 percent of children are in Grade 1 at the age of 6 (Table 4.2). Grade level Male Female Total Grade 1 19.9 19.1 19.5 Cultural beliefs combined with early marriage and teen Grade 2 21.1 20.3 20.7 pregnancy strongly contribute to late entry and high repetitions Grade 3 19.3 18.2 18.8 rates.8 This generates costs to education system as well as direct Grade 4 18.3 18.0 18.1 costs to households and opportunity costs to students. EB1 19.8 19.0 19.4 Source: World Bank and MOE, 2014-15. Completion and Retention Rates 9. Primary completion rates in Guinea-Bissau are low. In 2014, with a primary completion rate of 29 percent, Guinea-Bissau ranked near the bottom for Sub-Saharan Africa (Figure 4.1). Among the factors preventing children from completing a full cycle of primary education are the frequent disruptions in service provision, incomplete availability of grades in schools, late entry, and high repetition and drop-out rates.9 6 The World Bank’s Quality Education for All project, which is jointly funded with the GPE and aims to support several of ESP’s key priority measures, will become operational mid-2018. 7 Prior to the nomination of a new government in May 2018, the former Minister of Education was nominated in August 2016. The minister has not had full legal power over the ministry for the last 1-2 years of these governments but just maintained the "day-to-day functioning" of the ministry. Changes at the top have been accompanied by changes of the directors of the Faculty of Law, the INEP, the UAC, the Chief Inspector, the permanent secretary, and the ministry's director of finance since 2016/17, not to mention the regional and school directors, who usually suffer turnover due to political factors. 8 World Bank, Quality for Education All Project, Draft Project Appraisal Document, 2018. 9 World Bank, Quality for Education All Project, Draft Project Appraisal Document, 2018 73 Figure 4.1: Primary Completion Rates in Africa Figure 4.2: Retention Rates in Grades 1–6 Among and Peer Countries 16-22 Year-olds, by Gender, Area, Wealth Quintile and Region 160% 100% 140% 80% 120% 100% 60% 80% 40% 60% 20% 40% 0% Girls Boys Rural Q2 Q3 Q4 Bolama Bissau Q1 (poorest) Gabú Tombali Bafatá Quinara Oio Urban Biombo Q5 (richest) 20% Cacheu 0% Mali Togo Sudan Guinea Cameroon Ghana Gabon Comoros Guinea-Bissau Gambia Liberia Benin Ethiopia Nigeria DRC Tanzania Kenya SLE Rwanda Zimbabwe Namibia Lesotho Chad Zambia Gender Area Wealth quintile Region Source: UNESCO Statistics based on households surveys. Source: World Bank calculation based on MICS 2014. Figure 4.3: Out-of-School Rate by Age (2014) Figure 4.4: Out-of-School Rate by Wealth Quintile, Area and Age Group (2013–14) 80% 50% Dropped-out Dropped-out 45% 70% 40% Have never attended school Have never attended school 60% 35% 30% 50% 25% 40% 20% 15% 30% 10% 20% 5% 10% 0% Urban Urban Q1 Q3 Q5 Q2 Q4 Q1 Q3 Q5 Rural 0% 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Age 6-11years 12-14years 15-17years Source: World Bank calculation based on MICS 2014. Source: World Bank calculation based on MICS 2014. 10. Retention rates are estimated to be low amongst girls, and rural and poor populations .10 Among 16–17 year-olds, about 89 percent have been to school and at least completed Grade 1. Among 21–22 year-olds, only 58 percent have attained Grade 6, resulting in an estimated retention rate of 64 percent. Disaggregated data show that girls, and children in rural areas and the bottom income quintiles do worst under the current system (Figure 4.2). Gabu, Tombali and Quinara regions have retention rates 40 percentage points lower than Bissau region, i.e. Autonomous Sector of Bissau. Out-of-School Children 11. Despite the Government’s efforts, about 50 percent of 6-year-olds have never attended school. Due to a range of sociocultural, economic, access and other barriers, almost one third of children between the ages of 6 and 11 have never attended school (Figure 4.3).11 This is particularly true for children living in rural areas and poor children. Among those who do start school, dropout rates gradually rose from 5 to 44 percent between the ages of 12 and 24. Nearly 40 percent of 6-11 year-olds from households in the poorest quintile have never attended school, compared to only 7 percent for children from the richest 10 The retention rate is the ratio between the access rates for the last grade and the first grade. 11 UNICEF (2018, forthcoming) “Study on Out-of-School Children in Guinea-Bissau”. 74 quintile (Figure 4.4). School dropout rates are also higher among children from the poorest quintile for 12- 14 year-olds and 15-17 year-olds. The gap is also large between urban and rural areas: 40 percent of rural 6-11 year-olds have never attended school, compared to 12 percent in urban areas. The dropout rate reaches 18 percent in rural areas for 15-17 year-olds while it is only 2 percent in urban areas. Figure 4.5: Enrollment Concentration Curves by Figure 4.6: Benefit Incidence of Spending, All Education Level (2010) Levels of Education (2013 Expenditure) 100 1 0.9 Line of equality Cumulative enrolment, % (total) 80 0.8 Cumulative benefits 2013 60 0.7 0.6 40 0.5 0.4 20 0.3 0 0.2 0 20 40 60 80 100 0.1 Cumulative wealth quintiles, % of population 0 Primary education Line of equality 0 Lowest Second Third Fouth Richest Secondary (total) Higher education quintile quintile quintile quintile Source: Authors’ estimate using ILAP 2010. Source: Based on ILAP 2010 and 2013 current expenditure Figure 4.7: Government Subsidy to Education from Figure 4.8: Gender Equality in Education by Current Expenditure by Wealth Quintile (2013) Wealth Quintile and Education Level (2010) 1600 10 Above zero : female advantage Percentage points (difference from 50%) 1400 0 1200 -10 Billion 2013 CFA 1000 800 -20 600 -30 400 -40 200 0 -50 Lowest Second Third Fourth Richest Primary Secondary Higher quintile quintile quintile quintile Lowest quintile Second quintile Third quintile Fouth quintile Richest Source: Authors’ estimate using ILAP 2010. Source: Based on ILAP 2010 and 2013 Current Expenditure Figure 4.9: Government Subsidy to Education by Gender and Wealth Quintile (2013 Expenditure) 1000 Total subsidy to girls Total subsidy to boys 800 600 400 200 0 Lowest Second Third Fourth Richest quintile quintile quintile quintile Source: Based on ILAP 2010 and 2013 Current Expenditure. 75 12. These internal inefficiencies result in substantial costs for the Government. The internal efficiency coefficient for Guinea-Bissau was estimated at 53.4 percent in 2010 which is significantly lower than the 80 percent benchmark target recommended by GPE. 12 In other words, nearly half of the scarce public resources allocated to primary education are lost. This can be attributed mainly to the high dropout rate and, to a lesser degree, the high repetition rate.13 Without students dropping out, the internal efficiency coefficient would have been 86 percent. Learning Outcomes 13. Learning outcomes are poor mostly due to the low quality of the teaching force . Based on the results of the regional Program for the Analysis of Education Systems (PASEC) assessment (2012), Grade 5 students in the regions of Tchico Té and Bolama scored 28 percent more in math and 26 percent more in Portuguese when their teachers had received proper training. Nearly 94 percent of teachers in Grade 5 were unable to answer all the questions in Portuguese and 98 percent in math. The low quality of the teaching workforce, combined with a weak mastery of Portuguese among most students and teachers, are contributing to the low level of learning between Grades 2 and 5 where critical content is taught.14 Inequality in Distribution of Public Resources 15. Access to public primary education is more equitable compared to secondary education and above. Only 13 percent of the students in secondary schools are from the poorest quintile while 23 percent are from the top quintile. In higher education, the distribution becomes highly unequal, with 39 percent of students coming from the richest quintile and only 4 percent from the poorest. The concentration curves further show the level of inequality in enrollment by education level (Figure 4.5). Increasing spending on higher education levels therefore benefits the rich. 16. Overall, public spending in education tends to favor wealthier households (Figures 4.6 and 15 4.7). In 2013, the government spent about CFAF 6,067 billion on all education levels, excluding investment, administrative units, and vocational training. An average benefit analysis across wealth quintiles suggests that 22 percent of the benefits went to the richest quintile (around CFAF 1,353 billion) whereas the lowest quintile receives 14 percent (CFAF 867 billion). 17. Despite Guinea-Bissau almost achieving gender parity in primary education, girls benefit less from government spending overall. Almost half the students in public primary education were girls in 2010 but the share falls in secondary and higher education (Figure 4.8). In terms of government spending, girls from the richest quintile benefit from about CFAF 624 billion (43 percent of the total), while for boys the figure is CFAF 730 billion. 12 The internal efficiency coefficient is calculated as the ratio of the ideal number of student-years required to produce a number of graduates from a given student cohort in elementary/secondary education to the actual number of student-years spent to produce the same number of graduates, expressed as a percentage. 13 RESEN (2015). 14 World Bank (2018) Education PAD. 15 To assess average benefits, the current government expenditures by levels on 2013 were combined with the enrollment rate from 2010 ILAP. Benefits are also presented in a Lorenz-curve graph that cumulates benefits across wealth quintile. 76 Figure 4.10: Source of Public Spending for Education Figure 4.11: Distribution of total Spending by (Average over 2010–17) Sector (Average over 2010–17) 100% General Admin and Finance 29.7 80% Security 16.3 Education 12.3 60% Percent Health 9.9 40% Public works 6.6 20% Agriculture 3.0 0% Energy 0.8 2010 2011 2012 2013 2014 2015 2016 2017 0.0 10.0 20.0 30.0 Domestic External Percentage of total spending Source: MOF. Source: MOF. Figure 4.12: Guinea-Bissau and Peers: Education Figure 4.13: Benchmarking Primary Level Spending Outcomes 25 7 50 6 Difference from average primary completion rate MUS 20 40 More Overachiever 5 30 TCD efficient GIN s 15 4 CPV 20 Percent SYC MLI SLE 10 3 10 GNB MOZ STP BDI CIV COM NER 2 0 RWA COM GMB 5 COD MWI SLE BEN 1 -10 GNB CMR BDI CAR GHA MRT MDG SEN 0 0 -20 ETH BFA KEN -30 Underachiever CAR Less efficient GMB -40 % of government expenditures (left axis) -15.0 -10.0 -5.0 0.0 5.0 10.0 15.0 20.0 % of GDP (right axis) Difference from average government expenditure per student Source: World Development Indicators (Avg. 2010-2015) Source: World Development Indicators (average 2010–15) Figure 4.14: Trends and Required Education Expenditure (2010–20) 30 20 25 15 in CFAF billions 20 Percent 15 10 10 5 5 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Education expenditures ESP 2017-2019 costing Share of public expenditures (right axis) Share of GDP (right axis) Source: MEES, MOF 77 4.3 Education Spending: Trends and Links with Sector Priorities Alignment of the Budget with Government Priorities 18. Even though the education sector is among the priority sectors in the Government’s Strategic and Operational plan for 2015–20 (“Terra Ranka”), low budget allocations undermine this policy goal. The education sector is primarily funded by the government but its share of financing has fluctuated and shows a downward trend, dropping from 74 percent in 2010 to 69 percent in 2017 (Figure 4.10). On average, donor funding16 accounts for more than a third of the education spending, on average, over 2010- 2017.17 Total education spending remains low, amounting to only 2.2 percent of GDP or 12 percent of total public expenditure. As Figure 4.11 shows, education is the third largest sector after general administration and finance (30 percent of the budget) and security (16 percent). 19. Guinea-Bissau’s education spending is also low and underperforms at primary level compared to its peers. Education spending, both as a share of total government spending and GDP, is low compared to peer countries and the average for SSA and WAEMU countries (Figure 4.12). In fact, Guinea- Bissau spends less than countries with a similar or a lower GDP per capita level, including Sierra Leone, and The Gambia. In terms of the link between spending and outcomes, Guinea-Bissau also underperforms on primary completion rates (Figure 4.13). Maximizing financing for human capital – the World Bank Human Capital Project The newly launched World Bank Human Capital Project proposes that governments should invest more to deliver better health and education quality outcomes . This project represents a new effort to better understand the link between human capital investments and economic growth, and to accelerate financing for education and health. The human capital project targets better outcomes and the Bank aims to develop a new human capital index to rank countries based on the outcomes they are achieving with investments in health, education, and social services. Measuring outcomes will require a clear understanding of what constitutes inputs, outputs and a decision-making unit, and substantial data across a wide array of outcome indicators. In Guinea-Bissau, data on learning and health outcomes remain weak and somewhat outdated; hence the need for more and better data to measure progress towards achieving social sector priorities. In education, data will be needed on learning outcomes to measure the efficiency of the education sector at achieving better learning outcomes and improved allocation of public expenditures (World Development Report, 2018). The last (regional) assessment of learning outcomes in Guinea-Bissau was conducted in 2012. Additional data would be needed on school factors (teacher education and knowledge, pedagogical competencies, school grants and textbooks, electricity) and indicators of “value-added” to pupil learning by schools. In health, data on infant and under-five mortality, and stunting is part of the scope of the Human Capital Project. Data on health inputs (health workforce, infrastructure, and medical equipments) and outcomes (infant and under-five mortality, adult survival rate) would be required to measure the effectiveness of health sector investment in improving service coverage. The recently completed SDI will help shed light on health sector efficiency in Guinea-Bissau: productivity of health workers, absenteeism, and the competence of health care professionals. 16 The main development partners in Guinea-Bissau are: ADPP Humana People to People, the European Union (EU), Portugal, Plan International, UNICEF, the World Bank, and the World Food Program (WFP). There are also a handful of NGOs in the sector working at the central, regional, and school levels to improve service delivery and education outcomes. Information on donor spending by level of education is not available. 17 The assessment of the sector’s sources of funding is based on Government’s expenditures. An assessment of private expenditures was not possible since the MICS does not distinguish between private spending on private and public schools. 78 20. Although public education spending has increased over 2010–17, the sector remains largely underfunded. The preparation of the ten-year ESP for 2017–2518 provided an opportunity to tackle the challenges in the sector and channel funding to priority areas. Education spending almost doubled in real terms, from CFAF 8 billion in 2010 to CFAF 19 billion in 2015, but declined by a third in 2016. However, education spending picked up in 2017, reaching 2.2 percent of GDP (Figure 4.14). This volatility can be in part explained by political instability and high aid dependence. Education will need a substantial and sustained increase in funding to finance the interventions identified in the ESP. Resources are particularly needed to alleviate the shortage of qualified teachers, lack of publicly provided inputs within schools, and an infrastructure deficit that places serious constraints on both the quality and accessibility of education. Figure 4.15: Breakdown of Education Expenditure Figure 4.16: Guinea-Bissau and Peers: Education Expediture by Level Higher Vocational Pre-school Sierra Leone education and training 2% Alphabetisation scientific 2% 1% SSA research Burundi 9% Upper CAR secondary 15% Guinea-Bissau Comoros Basic education Gambia, The (EB1 and Lower EB2) 0 20 40 60 secondary 46% % of Government expenditures (EB3) 25% Expenditure on tertiary Expenditure on secondary Expenditure on primary Source: RESEN (2015). Source: World Development Indicators (Avg. 2010-2015). Expenditure by Education Level 21. The distribution of resources reflects that universal primary education has been prioritized but spending may be regressive.19 The share of primary education in public expenditures was at 46 percent in 2013 whereas secondary and post-secondary education represented 39 percent and 14 percent, respectively (Figure 4.15). The level of spending on primary education remains below GPE good practice recommendation of 50 percent.20 Spending by level is also disproportional to the enrollment distribution, with primary education receiving less than half of public resources but accommodating about 70 percent of total enrollment in public schools. In contrast, secondary education represents one-third of the students and receives 40 percent of the resources. Note, however, that one cannot make a definitive statement about whether spending is regressive or not without controlling for unit cost which is supposedly higher in post- primary education. The case for primary education is also based on equity considerations as it benefits all income groups while the share of students from the lowest quintile declines as the level of education rises. 18 The original ESP ran from 2016 to 2025, but due to the political crisis, the focus period of the plan fell to less than 10 years. 19 Since the budget nomenclature is not aligned to the Government’s education priorities and goals (budget codes for pre-school, primary, secondary, and tertiary are often combined) the expenditure review by level is based on the data available in the 2015 RESEN. The breakdown is based on the recurrent budget, information on the investment budget by level is not available. 20 https://www.globalpartnership.org 79 22. Given the overall low level of education spending, it is not surprising that pre-school and vocational training are largely underfunded. Figure 4.15 shows that pre-school education received only 2 percent of funding in 2013 even though it has proved to be effective in improving children’s cognitive skills and educational attainments.21 In fact, research has shown that investing in and increasing participation in early childhood development can improve educational efficiencies, notably reducing repetition, dropout and delayed entry. Likewise, only 2 percent is spent on technical and professional training, though skill acquisition is vital for an economy to compete and grow. 23. Benchmarking shows that Guinea-Bissau spends relatively more on secondary but less on post-secondary education than peer countries. Guinea-Bissau’s high share of spending on primary education can be attributed to the inclusion of EB3 expenditures under primary education. Its secondary education spending is also above the Sub-Saharan African average and puts Guinea-Bissau among countries with a high share of public secondary expenditures. This comes at the expense of lower spending on tertiary education, where Guinea-Bissau ranks the lowest (Figure 4.16). 24. There seems to be scope to prioritize existing education spending better . While a key issue is the inadequate funding, there seems to be some scope to reallocate resources towards primary education within the limited envelope, given the large number of students out of school and the low completion rate at primary level. Yet the amount spent per student increases after basic education, exacerbating inequities as spending on higher education levels tends to favor more students from well-off families. Expenditure by Economic Classification 25. With the bulk of the sector’s funding primarily going on salaries, Guinea-Bissau’s education spending is not in line with best practice benchmarks. The lion’s share of recurrent expenditure is accounted for by staff salaries (92 percent of the total) while non-salary recurrent expenditure accounts for only 8 percent (Figure 4.17). Even though the budget classification does not allow education expenditure to be mapped to economic functions by level, it can be assumed it reflects the spending pattern in primary education. This low spending on non-salary recurrent expenditures likely puts the country below benchmark suggested by the Education for All Fast-Track Initiative. 26. Low spending on critical educational inputs and investments raise concerns about the sector’s capacity to meet the needs of a growing student population while also achieving universal primary education. Spending on goods and services—which supports the purchase of textbooks, teacher training, curriculum development and school maintenance—is very low in Guinea-Bissau, less than 3 percent of recurrent expenditure. Spending on goods and services has declined in real terms for most of the past years (Figure 4.18). Construction of schools and classrooms is lagging behind (the average class size exceeded 100 students in 2013), further compounded by poor geographical distribution. 22 Most critical inputs, from textbooks, school material, training and school meals to the construction of classrooms, are funded by donors and non-governmental organizations (NGOs). Relying on external funding for these inputs is not only unsustainable, but renders the sector vulnerable to the often unpredictable nature of aid. 27. The sector’s investment expenditures are heavily aid dependent and thus volatile, although critical for achieving the ESP goals. Half of Guinea-Bissau’s development assistance is accounted for by three sectors, with the largest share taken by health (22 percent), followed by public works and housing (19 21 SDG 4 – the education goal – also calls for access to quality early childhood development, care and pre-primary education (SDG 4.2). 22 RESEN (2015) 80 percent) and education (14 percent) (Figure 4.19). The lower level of aid to the education sector in recent years reflects a global trend of stagnation in aid to basic education.23 Nonetheless, donor aid funds almost all of Guinea-Bissau’s investment in education (on average 99 percent).24 The upswing in 2013 and 2015 reflects a period of stability and government reforms while the drop in 2016 is mainly associated with the ending of several projects and less with the political instability at the time (Figure 4.20). 28. Aid management has been undermined by fragmentation of aid, limited project management capacity, and a lack of consolidated reporting. A limited number of donors support the sector, largely through small-scale projects due to the frequent political instability.25 Donor projects and programs constitute a strain on the country’s administrative capacity in terms of procurement, disbursement, and monitoring requirements.26 Aid disbursement is often unreliable, and there is no predictable and systematic reporting of aid. There is a coordination mechanism in place to improve aid coordination and effectiveness: a local group, chaired by the Minister of Education, meets several times during the year to share information and discuss strategic orientation and bottlenecks in the sector. Expenditure by Administrative Level 29. The education budget remains largely centralized, with resources at the subnational level limited to the collection of school fees from secondary education. Although the MOE has decentralized some administrative planning and monitoring functions to the regional directorates and staffed the offices, but fiscal decentralization is lagging behind, primarily due to capacity and fiscal constraints. Neither regional directorates nor schools receive any transfers from the central administration. Provisions are made on an ad hoc basis depending on resource availability and lack a strategic planning approach as well as clear distribution criteria. For example, the 2018 budget provides funding for the purchase of desks and chairs. 30. Education spending is regressive at the regional and local level . According to the 2010 ILAP household survey data (which does not include government spending), rural areas receive less than a quarter of education spending, even though nearly 60 percent the population lives in rural areas. 27 Resource distribution is biased towards the capital, Bissau, which absorbs over 60 percent of public resources but represents only 26 percent of the student population (Figure 4.21). In contrast, the student population in Oio receive less than 4 percent of available resources, although it represents 14 percent of the total student population. Education spending is also biased towards wealthier regions. Poor regions such as Gabu, Oio or Bolama/Bijagos have high student-teacher ratios but receive relatively less expenditure (Figure 4.22). Altogether, several challenges remain, including critical regional issues, such as the lack of clear criteria for distributing resources, problems retaining teachers in remote areas, and students in poorer regions receiving fewer learning opportunities. 23 https://www.brookings.edu/opinions/aid-to-basic-education-in-developing-countries-under-threat/ 24 All donor aid is classified under the investment budget though some parts in practice fund recurrent expenditures – however, information is not available. 25 The main education sector development partners in Guinea-Bissau are: ADPP Humana People to People, the European Union (EU), Portugal, Plan International, UNICEF, the World Bank, and the World Food Program (WFP). There are also a handful of NGOs in the sector working at the central, regional, and school levels to improve service delivery and education outcomes. 26 Information on off-budget is not available. 27 RESEN 2015. 81 Figure 4.17: Current Expenditure by Category Figure 4.18: Total Expenditure by Category (2010–15), % of total current expenditure (2010–17) 100% 80% 20 Percent of total spending 60% 40% 10 20% 0% 0 2010 2011 2012 2013 2014 2015 2016 2017 2010 2011 2012 2013 2014 2015 2016 2017 Salaries Goods and Services Salaries Investment Transfers Other current expenditures Goods and Services Transfers Source: MEF, BOOST. Source: MEF, BOOST. Figure 4.19: External Expenditure by Sector Figure 4.20: Education Investment by Source (2010-17) Other 7 10 Energy & Industry 2 Natural Resources 3 8 Justice 3 Fisheries & Marine Res. 3 in CFAF billions 6 Women, Family & Social… 5 Economy &Regional Integr. 7 4 Planning & Regional Integr, 7 Agriculture, Forestry and… 8 2 Education & Higher education 14 Public works & hoursing 19 0 Health 22 2010 2011 2012 2013 2014 2015 2016 2017 - 5 10 15 20 25 -2 Percent Domestic External Source: MEF, BOOST. Source: MEF, BOOST. Figure 4.21: Education Expenditure and Student Figure 4.22: Student-Teacher Ratio and Population by Region Poverty Level by Region 60 30 80 80 50 25 60 60 40 20 30 15 40 40 20 10 20 20 10 5 0 0 0 0 Share of expenditures consumed Share of expenditures consommed Share of group in the population Teacher/student ratio Linear (Share of group in the population) Poverty Source: RESEN (2015). Source: RESEN (2015), ILAP 2010. 82 4.4 Public Financial Management (PFM) in the Education Sector Overview of PFM Systems 31. Guinea-Bissau has partially aligned its PFM legal framework to WAEMU directives, but the framework has yet to be translated into specific processes or procedure manuals. It created a Committee (Comissão de Acompanhamento) to oversee the full internalization and implementation of all WAEMU directives. However, these processes have been delayed by the political instability and malfunctioning of the National Assembly until April-2018. In 2015, the National Assembly approved the new organic budget law and the Transparency Code, but some other important directives are still pending. The latter include the budget nomenclature, public accounting rules, the chart of accounts, and the Table of State Financial Operations (TOFE). The implementation of PFM arrangements in sectors rely on general PFM practices. The actual procedures in place are based on informal practices, spread out through the whole public administration, without any specific established procedures or procedure manuals. Figure 4.23: Financial Flows in the Education Sector 3 Note: The continuous lines stand for actual flow of funds, blue lines for in-budget flows, black lines for out-of-budget, red lines for informal payments, long-dashed lines for actual but irregular flows, and short-dashed lines for unrealized flows. 35. The education system has three main financial sources: public funds, private funds, and external funds (Figure 4.23). Public funds mainly consist of general tax revenues and also fishing licenses, timber exports and tourism revenues, which comprise the state budget. Private sources are mainly out-of- pocket (OOP) payments, but also informal payments from students. External funds come from international donors and organizations, which are often earmarked for specific programmatic expenditures. 37. School fees are believed to be an important source of internal revenue, but there is no much information about the actual amount collected and how they are used. There is no formally established mechanism for revenue management in the sector other than some established fees for schooling from secondary schools (Grade 7 and up). These are managed by the school director or treasurer and documented in registry books. In theory, 40-50 percent of fees are retained at the school level for operational expenses, while 10 percent are transferred to the DRE and 40 percent to the MOE and MEF. However, these procedures are neither fully established nor transparent. There is no financial reporting of the amount of school fees collected or how they are used. In practice, these fees are divided on a discretionary basis including with the central administration of the MOE, leaving almost no funds to meet schools’ needs for basic educational material and furniture. Recently, to promote transparency, the MOE 83 instructed all schools to open bank accounts for their internal revenues, but this practice has not been enforced. 38. Schools also receive in-kind support, such as textbooks, food for school meals, and chalk mainly from UNICEF and other donors. However, the distribution of these items is not well recorded and sometimes depends on the region and type of school, undercutting efforts towards quality school inputs. 28 Planning and Budgeting Process, and Medium-Term Perspective 40. The Ministry of Education’s sectoral plan is a “fundraising” tool rather than a planning and budgeting document. Staff from the MOE’s Gabinete de Estudos, Planificação e Avaliação do Sistema Educativo (GEPAS) are responsible for sectoral studies, planning activities and evaluation of the education system, but have limited tools, planning capacity and lack technical knowledge for developing planning activities in the education sector. On the basis of the Education Sector Plan, the MOE developed the first Triennial Budgeted Action Plan (Plano de Acção Trienal Orçamentado) for 2017–19 with external technical assistance. However, decentralized entities, such as the regional directorates and schools, were not consulted on the preparation of the plan. 38. The education budget has generally not been used as a tool to ensure sectoral priorities and guide expenditure priorities since 2015. None of the budgets since then have been approved by the National Assembly due to political instability. The government has been functioning based on monthly budgets, corresponding to the 1/12 of the 2015 budget. As a result, expenditure selection and prioritization processes have remained weak. 39. The budget preparation process follows a top-down approach with little involvement of sector ministries in the preparation process. There is a high concentration of PFM tasks at the MEF, while ministries and all other public entities have a high level autonomy in the management of internal revenue. The MEF is the sole steward of fiscal and budgetary policy and allocates fiscal space to the various ministerial departments, for submission to the council of ministers and the National Assembly. 29 Therefore, the budgeting process is more of a cosmetic process, tailored to fit the fiscal space imposed by the MEF and external partners, but which does not reflect the real needs of the education sector. 40. The organizational structure and limited capacity in the MOE further complicate the budget preparation process. The ministry functions based on an old organic law from the early 1990s. In 2015, the Council of Ministers elaborated and approved a new law defining its attributions and establishing how the ministry would be organized and function. However, this organic law has not yet been signed by the president or promulgated. During budget preparation, the MOE does not consult with decentralized entities, such as the regional directories or schools, to identify their needs. Further, the staff involved in planning and budgeting do not have the necessary skills and experience. For instance, the Administrative and Financial Directorate (DAF) lacks a dedicated budget officer as well as planning and budgeting IT tools. 41. The Government has recently introduced a medium-term perspective to budget planning with a new three-year budget covering 2018–20. This medium-term perspective was introduced in an effort to comply with the WAEMU directives. The medium-term budgetary framework is the tool from which the budget law is elaborated. The Ministry of Education, together with Health and Public Works developed Draft SDI 2018. 28 29Therefore, the budgeting process is more a cosmetic process, tailored to fit the fiscal space imposed by the Ministry of Finance and external partners, but does not reflect the real needs of the education sector. 84 the first pluriannual budget for 2018–20, but required external technical assistance due to a lack of ministry commitment and trained public servants. Comprehensiveness and Transparency of the Budget 42. The classification and coverage of the budget is not aligned with the education structure, hampering the tracking of budget allocations by education level . The budget comprises only five budget lines, in two main categories: Normal Operations (payroll, purchase of goods and services, and transfers) and Public Investment Program (domestic and foreign financing). A major impediment for the education sector is that the budget nomenclature is not aligned to the government’s education priorities and goals. The budget lines are not disaggregated by level of education (pre-primary, primary, secondary, and tertiary), which creates serious difficulties in terms of planning. 43. The budget contains limited information on performance and costing of service delivery . There is no service delivery performance information in the annual budgets or the budget supporting documentation, nor any evidence about performance audits or evaluations. GEPAS has elaborated some performance indicators with a 2013 baseline and estimates for these indicators. However, the targets have only been established for the years 2019, 2022, and 2025, and there is no information about the activities to be implemented and the relevant cost information for the different decisions, which would help ensure that the data sources and costing process are robust and relevant to the achievement of the objectives. Sticking to Sector Priorities: Budget implementation Budget Execution 44. The sector’s budget implementation performance shows that most categories were underspent, with the exception of externally funded investment (Table 4.4). The education budget was on average underspent by 8 percent between 2010 and 2017, but it has matched budget appropriations more closely in recent years. Not surprisingly, salaries (which averaged -5 percent underspent) were generally the best-performing type of expenditure. In contrast, even the limited resources allocated to goods and services were not fully used, reflecting the underfunding of important inputs and service delivery. The sector’s investment program has shifted from an underspend during 2010–12 to exceeding budget appropriations in 2013–17. This was driven by an increase in externally funded investments and primarily reflects the volatility of aid. Table 4.3: Budget Execution: Initial Budget versus Actual Spending (2010–17), Percentage 2010 2011 2012 2013 2014 2015 2016 2017 Avg. Salaries -4 -1 -3 -24 11 0 -8 -12 -5 Goods and services -29 -15 -73 -83 -7 -5 -63 -91 -46 Transfers -47 -46 -28 -70 -11 -21 -69 -55 -43 Other current expenditures -93 -44 -94 -72 -1 -61 Investment -64 -41 -41 14 24 90 6 78 8 Domestic -99 -58 -98 11 -61 External -52 -34 168 20 36 91 4 78 39 Total -37 -20 -23 -12 15 25 -9 -1 -8 Source: BOOST. 85 45. Lower-than-budgeted appropriations across all spending items undermine the functioning and effectiveness of the central administration and in particular the frontline service providers —the schools. Several factors might have contributed to the under-performance, including: (i) weaknesses in budget planning; (ii) delayed payments of committed expenses due to liquidity constraints in the MEF; (iii) poor procurement and management capacity; and (iv) poor control over expenditures. These shortcomings reflect the overall weaknesses of Guinea-Bissau’s PFM systems and are likely to have an adverse effect on education at the school level. Expenditure Management 46. Mismatches between the timing of payments and the availability of cash has contributed to the MOE’s low execution rates. The reinstatement and strengthening of the treasury management committee in 2017 greatly improved cash management. However, more remains to be done to build institutional capacity in the MOE, given the tendency for the MEF to expenditure decisions on an ad hoc basis with no strategic guidance. The absence of a system or criteria for determining spending and investment priorities has left it up to the Finance Minister to make these decisions, with only limited consultation with other ministries. Cash disbursement has thus remained unpredictable for the MOE, undermining its short-term planning and timely payments. It is also noteworthy that there is no established procedure for the commitment or cash management related to the use of internal revenues. 51. The MOE’s compliance with public procurement procedures has been low, which may have led to higher costs and poor-quality purchases. The Public Procurement Act (Law-decree No.5/2005), which is consistent with WAEMU directives, established competitive tendering as the preferred procurement method. In recent years, all procurement processes at the MOE have been externally financed and it has rarely prepared procurement plans. In 2017, the procurement directorate, staffed by only one person, participated in one procurement process. The MOE does not respect the principle of segregation of duties during budget execution. For instance, during the procurement of goods and services, the Administrative and Finance Directorate (DAF) notifies the MOE treasurer and performs the procurement itself through the ministry’s bank account and records the expenses. Non-compliance with public procurement raises concerns about the cost and quality of purchases, which in turn affects service delivery. The use of school fees or other public resources have been confined to direct contracting. 52. Guinea-Bissau’s integrated financial management system (SIGFIP) is not yet integrated, interconnected or decentralized, which undermines expenditure management . SIGFIP only works within the MEF while most expenditure is first documented on a paper-based approval system. Donor aid and public sector revenues also have to be manually inserted into the system. The absence of a fully functional financial management system at the MOE hampers the ministry’s access to real-time financial information that the DAF could use to administer programs effectively, formulate budgets, and expedite payment transactions. Delays in transactions and unreliable financial information ultimately affect frontline service delivery since scarce resources may be subject to discretion, impeding proper planning and implementation. Even more concerning, the absence of financial reporting on the use of resources renders the central administration less accountable to decentralized entities. Internal Control and Audit 53. The MOE’s internal control systems are weak, largely due to outdated laws, policies, and missing procedures. There is no evidence of any accounting system in place across central, regional, and school level. Each entity has its own accounting system, mainly paper-based registries, that in some cases are reviewed by the regional directorates. There are no established procedures covering the functioning of 86 the regional directorate and schools. In some cases, self-managed (autogestão) schools have started to develop procedure manuals, but there is no official document requesting these, or providing guidance about how to do it. School fees are sometimes kept in cash at the schools, or in the school directors’ homes, for security purposes. While the absence of a functioning control system affects key administrative functions of the ministry (from safeguarding the sector’s assets to keeping proper records, timely reporting and compliance with laws and regulations), the sector is more vulnerable to inefficiencies and corruption. 54. The effectiveness of internal audit is also undermined by the fact that the General Inspection has no legal base or criteria for auditing. The General Inspection (GI) of the MOE is charged with the internal audit function (internal regulation, the management of the inspectors, and the regulation of school accreditation). The organic law of the GI, however, remains underdeveloped. Its full functioning has also been constrained by technical factors, notably lack of capacity and lack of resources for inspections, limited use of information technology, and weak communication between the GI and other financial control entities (the Financial Inspection General of the MEF, the Ministry of Public Administration and the Anti-Corruption Entity). Most inspectors are based in urban centers and have visited schools on an ad hoc basis30 but there are no formally agreed standards based on which it evaluates schools and teachers. However, with the support of donors, the ministry has recently finalized National Quality Standards and recruited over 160 new, mid-career and energetic school inspectors who will be based regionally, close to the schools. Financial Reporting and External Controls 55. Financial reporting is generally deficient. Guinea-Bissau’s financial reporting system is weak due to limited coverage, untimeliness, and lack of reliable execution data. Nonetheless, budget execution monitoring has improved in 2017 with budget execution reports now being produced quarterly by the MEF, although there is no evidence of any financial reporting at the MOE. However, effective use of these reports is difficult given the data in SIGFIP do not reflect actual expenditure since the information is first processed manually and then entered into the system after significant delays. Even more concerning, the absence of financial reporting on the use of resources renders the central administration less accountable by decentralized entities. 56. Final payments are not monitored. The public accounts are closed at the engagement level and there is no institutionalized follow-up or monitoring of the final payments realized or any pending payments. In recent years there have been efforts to audit the public debt (other than the debt engaged with the external partners), but without much success. 57. Although some schools have been audited, external control has been very limited in the education sector. Created in 1992, the Court of Auditors (Tribunal de Contas; TdC) struggles to assert itself as a supreme entity for the external control of public finances. It is still governed by the Organic Law of 1992, while the adoption of a new organic law has been put on hold. 31 In 2017, the TdC audited some schools, mainly confined to the Bissau area due to mobility and budget constraints. The depth and scope of the audits have been limited because of the lack of specific regulations for the functioning of the schools, 30 The GI has performed some inspections but in a very limited area using private vehicles and laptops belonging to the inspectors and no budget for fuel or maintenance. The internal audit entities at the MOE and MEF and the SAI have begun to carry out selected audits at the regional level, but in the absence of specific established procedures, they cannot make any relevant recommendations. 31 The TdC is a member of two international collegiate organizations, notably, AFROSAI and INTOSAI; the norms used by TdC are similar but not identical to those of INTOSAI because of the realities in the country. 87 institutional arrangements, internal revenue management, and established financial reporting requirements that could be used as auditing criteria.32 4.5 Management of Human Resources 58. The Ministry of Education has staff at the national level and in the regional directorates, as well as school directors and teachers at the school level . In 2018, there were approximately 10,994 teaching and 1,052 non-teaching staff (including contract teachers) in the education sector. About 8,985 staff are on the MOE’s payroll, of which 89 percent are teachers. 59. While teachers constitute a key input into education and represent the largest cost element in the budget, the distribution of teachers across schools varies widely , which is a cause for concern from an equity and efficiency point of view. A recent study showed a large variation in student-teacher ratios across schools.33 The average ratio is 64.2 students per teacher whereas the bottom 10 percent of schools have ratios below 37 and the top 10 percent have ratios over 102. This is among one of the largest ranges of student-teacher ratios across schools in the world. The variation in ratios is also substantially larger within regions than between regions.34 While these large variations are worrying from an equity point of view, more work is needed to understand how far this affects educational outcomes. Given that teachers also absorb the lion’s share of the budget, an efficient teacher workforce is crucial. 59. The sector has a high staff turnover with most of the staff working in the private sector, with negative implications for performance. Political volatility has led to frequent staff changes in the ministry, impeding capacity building in the sector. There is a shortage of trained teachers. Guinea-Bissau hires about 2,000 teachers each year, but the teacher education institutions in the country only have the capacity to produce 1,000 graduates annually. There is no information about the qualifications of the other 1,000 teachers. The retention rate of teachers is very low. In fact, the education sector acts as a springboard for citizens to enter the public administration. Once hired, many of them never attend the schools to which they were appointed, while others abandon education to join another sector after the first year is over. Over 90 percent of the teachers employed by public schools also work in the private sector. This is also the case for non-teaching staff; most of civil servants in the ministry have a second job in the private sector. 60. Teacher performance has been a serious issue. Teacher absenteeism has been high. Teachers knowledge of subject matter is low – a learning assessment showed that 94 percent of teachers in grade 5 were unable to answer all the questions in Portuguese. 35 61. Delays in hiring process affect service delivery at the schools . Hiring is often not coordinated with the school year, with newly recruited teachers often arriving several months late in classroom. For example, the HR directorate at MOE was hiring and nominating teachers for the school year 2017/18 in April 2018. By the time these teachers start work, the school year will be over, and they will receive salaries without providing any effective work until the new school year. 32 The TdC has benefited from Pro PALOP –TL SAI project (an EU-UNDP partnership) to address its institutional challenges. The support aimed to strengthen effective control over the government’s public finances and to support institutional and civilian mechanisms th rough a South- South cooperation approach and peer learning. 33 Torsten Figueiredo Walter, Where are the teachers? The distribution of teachers across public primary schools in Guinea-Bissau, 2018. 34 Torsten Figueiredo Walter, Where are the teachers? The distribution of teachers across public primary schools in Guinea-Bissau, 2018. 88 62. Training plans for teachers are usually offered as benefits with no transparent criteria . Indeed, no competency-based training plan could be elaborated and implemented at the MOE, because the Ministry of Public Administration staff database: (i) does not reflect the current organizational structure of the MOE; (ii) the profiles of each position have not been defined; and (iii) the profiles of the staff who occupy those positions have not been developed. 89 Policy Recommendations – Health Technical Timeframe of feasibility policy impact Budget preparation and execution - Build planning and budgeting capacity in the Ministry of Education through providing training and Medium Medium term developing procedures manuals that document the mandates and responsibilities of each entity. - Ensure stability of the technical staff including Director Generals. - Improve budget transparency by: (i) making financial information (including the transfers of funds from High Short term MEF to the MOH and the decentralized units) and service delivery performance available to the public; (ii) documenting the distribution and use of fees collected from health centers. - Eliminate the practice of making expenditures using non-recorded internal revenue. High Short term - Improve payroll controls by: (i) procuring SIGRHAP for the MOH; (ii) updating the profiles of staff in the Medium Medium term MPA database; and (iii) addressing staff turnover and parallel work in the private sector. - Ensure value for money in procurement by: (i) preparing a procurement procedures manual, focusing on High Short term open and competitive procedures; and (ii)monitoring of results, and building staff capacity. - Reinforce the standards and procedures applied in internal audit by providing IGAS with the necessary High Short term budget to ensure its operational independence. Financial information technology systems - Work in coordination with MEF to connect the MOH to SIGFIP on a pilot basis. Medium Medium term Improve health service delivery -Improve the public resource tracking system. Strengthen regulations to eliminate informal payment, and Medium Medium term to rationalize the purchase and distribution of medicines. Human resource management - Strengthen health workforce policies, by defining clear career pathway for health workers, and revising the Medium Medium term remuneration policy to implement performance-based pay. 90 Chapter 5. Health1 Guinea-Bissau’s health sector faces persistent challenges related to the uneven and inadequate provision of health services, with most paid for by informal and out-of-pocket (OOP) payment. Both spending levels and outcomes are low—at or below those for its peers. Despite modest improvements in recent years, public financial management (PFM) systems remain weak. The country has adopted a national plan for the sector and prepared a three-year budget based on the plan. However, budget execution remains poor, flows of funds unaccounted for, and audit non-existent. Guinea-Bissau may want to: (i) build planning and budgeting capacity in the Ministry of Health; (ii) improve accounting and reporting and improve financial management information systems; (iii) track the use of public resources and eliminate informal payments; and (iv) strengthen its human resource management to improve the recruitment and management of health workers. 1. This chapter assesses key aspects of health financing and outcomes, as well as public financial management (PFM) systems in the health sector. Section 5.1 describes the organization and distribution of health care around the country. Sections 5.2 and 5.3 benchmark different aspects of the sector’s performance (including health care spending, delivery, and outcomes) against relevant peers. Section 5.4 examines challenges related to government health spending with a focus on PFM systems in the health sector and how they affect delivery and access to health care services in Guinea-Bissau. Finally, Section 5.5 concludes with discussion on policy recommendations. 5.1 The Health System in Guinea-Bissau 5.1.1 Health Service Delivery Organization 2. Health service delivery in Guinea-Bissau is structured around 11 sanitary regions, and organized into local, regional and central levels (MINSAP, 2018) . At the local level there are 114 sanitary areas, which are the primary locus for delivering primary health care (PHC) activities through 132 health centers. The regional level contains an administrative structure, the regional health directorates (DRS, Direção Regional de Saude), and technical units such drug warehouses, diagnostic centers, and secondary level regional hospitals. The central level is responsible for setting health policies, strategies, and regulations. This level includes the National Hospital, specialized hospitals, and satellite health centers. 3. There is a wide variation in the provision of health facilities across the country . For example, hospital bed density (beds per 1,000 population), varies from 0.56 to 2.29 and shows no association with under-five mortality rate (U5MR). The Autonomous Sector of Bissau (SAB) has around 46 percent of the total of hospital beds, followed by Cacheu region which accounts for 11 percent of the beds (Figure 5.1.a). Similarly, the concentration of health facilities (per 100,000 population) varies from 4.35 to 38.8, though the highest value (Bolama) is an outlier representing a small population. As for hospital beds, there is no link between facility density by population and mortality. However, there is a strong link between facility density by geographical area and U5MR (Figure 5.1.b). Oio/Farim has a low U5MR despite a low geographical density of facilities, perhaps reflecting the fact that much of Oio’s population lives in the city of Farim which has a significant concentration of clinics and thus small distances to travel to access health care. 1 Prepared by Edson Correia Araujo (Senior Economist, GHN04) with contirbutions from Nicolas Drossos (Consultant, GGO). 91 Figure 5.1a: Distribution of Health Facilities per 1,000 Population (2015) 50 Facilities Density (Units as Beds Density Type A/B Clinic Density Indicated) 40 30 20 10 0 Source: Republica da Guiné-Bissau, 2008a. Figure 5.1.b: Under-Five Mortality Rate and Figure 5.2: Evolution of the Health Workforce by Geographical Density of Facilities (Excluding Bissau) Category (1996–2016) 1500 1996 2007 2016 1000 500 0 Physicians Nurses Midwives Technical Support Other staff (aux.) Source: Republica da Guiné Bissau, 2008. Source: PNDS I (1997); PNDS II (2008); MINSAP (2016). 5.1.2 The Health Workforce 4. The country’s health system faces persistent challenges related to the inadequate supply and uneven distribution of health workers. The public health sector currently (officially) employs 2,173 workers, of whom 264 are physicians and 1,027 nurses (Figure 5.2). In 2016 there were 1.7 physicians and 11.5 health workers for every 10,000 inhabitants. In recent years, chronic political instability has had a noticeable impact on the health workforce, particularly in terms of the loss of skilled staff between 1996 and 2007. On balance, the size of the health workforce has been relatively stable over the last two decades, but Guinea-Bissau has made progress in upgrading auxiliary health personnel, and reducing the number of support staff in favor of training general nurses and physicians (MINSAP, 2017). In 2016, about 60 percent of all health workers were female, although women represented only 31 percent of physicians. 5. Health workers are concentrated in urban areas, leaving vast, remote regions without even a minimum health workforce. About 51 percent of all physicians, and 40 percent of all nurses are based in the SAB, home to just 25 percent of the country’s population. Populous regions such as Bafatá and Gabú suffer systematic disadvantage in the deployment of all types of health personnel in favor of the Bissau and neighboring Biombo areas. Medical specialists are all concentrated in the Bissau area – home to the country’s Level 3 hospital and a few private practices. When workloads are correlated with output indicators, they appear to be extremely low across the country, with most health centers hosting around 20 births and regional hospitals carrying out fewer than 15 Caesarean sections per month. 92 6. Scarcity of funds, political instability, and “commercialization from within”, represent the three key emerging forces shaping human resources for health in Guinea-Bissau. Government spending on staff salaries is noticeably small. The total wage bill for the health sector in 2015 was CFAF 216 million per month (about US$393,000) for the 2,173 health workers employed in the public sector. The state only covers health care-related salaries, with all other expenses (medicines, goods and services, investments) being partially covered by external funds and health facilities’ own revenues. 2 “Commercialization from within” means that most of the services and goods dispensed through the public national health system are paid for out-of-pocket (OOP), through a multiplicity of statutory legal charges and a myriad of illegal ones. Some of these charges are used to pay for health facility expenses, but the majority are pocketed by health workers. Given there seems to be no financing mechanism to recover facilities’ non -drug costs, revenues from charges are probably captured and managed by health workers acting as managers. 3 7. This system of informal charges appears to be so institutionalized among health professionals and users alike that it is considered to be the official way of supplementing small and irregular salaries. The health workers tend to see the additional charges as intimately related to their low and erratic remuneration, with many feeling “abandoned” by the state, which barely and irregularly pays for salaries, and leaves them fending for themselves to cover the facilities’ recurrent expenses. As no effective inspection system is in place for the health sector, the nature and extent of illegal charges seem to depend exclusively on the goodwill (and creativity) of the most senior officials in charge, their ability to enforce those charges, and consumers’ ability to pay the price. 8. The huge accountability deficit in the health sector reflects the overall weak governance in the public sector. There is widespread impunity at all levels of the national health system and few mechanisms are available to hold staff accountable for their actions. 4 Guinea-Bissau represents an extreme example of an ungoverned health workforce. In fact, with little supervision, health workers have gradually become the de facto operators of the health system. They run the system in their favor, charging under-the- table fees and setting quality standards according to their will, which has led to a private sector within the public health system.5 The prospects of a stable public health sector job, with the potential to engage in additional profit-making activities within the system, has led to an increased demand for training in healthcare work. This led to a rapid expansion of private sector training—widely perceived to be low quality. 5.2 Health Outcomes: Trends in Guinea-Bissau and Benchmarking 9. This section analyzes key trends in health outcomes in Guinea-Bissau and benchmarks them against its peers and regional comparators for a wide range of health outcome indicators (See Figures 5.3–5.10). It then conducts a benchmarking exercise to identify how far the observed changes in the epidemiological profile in Guinea-Bissau follow global trends or are unique to the country’s situation. 6 Relevant comparator countries were identified using the World Bank’s “ Find Your Friends” tool.7 The country’s structural peers include Burundi, the Central African Republic (CAR), Comoros, The Gambia, Sierra Leone, and Tajikistan. 2 Incomplete data on external assistance combine with absent information on paid user fees to obscure true financing levels. 3 Russo et al. (2017). 4 UNIOGBIS (2017) “Thematic report on the right to health in Guinea-Bissau. UN HR, Bissau, Guinea-Bissau.” 5 Russo et al. (2017). 6 Data sources include the Global Burden of Diseases and Injuries (2013) and the World Development Indicators (2016). 7 Structural peers are countries that share similar characteristics: Population size less than 10 million, real GDP per capita below US$600, high dependence on agriculture, below-average life expectancy, below-average trade with the world, and below-average government revenue. 93 Figure 5.3: Guinea-Bissau and Peers: Life Figure 5.4: Guinea-Bissau and Peers: Maternal Expectancy at Birth (1990–2015) Mortality Ratios (1990–2015) 65 3000 60 2500 MMR (per 100,000 live births) Life expectancy at birth, total 55 57 2000 55 50 53 1500 52 45 48 49 1000 40 500 35 0 1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015 GNB COM SLE GNB SSA GMB Source: World Development Indicators (2018). Source: World Development Indicators (2018) Note: GNB: Guinea-Bissau; COM: Comoros; GMB: The Gambia; BDI: Burundi; SLE: Sierra Leone; SSA: Sub-Saharan Africa Figure 5.5: Guinea-Bissau and Peers: Under-Five and Neonatal Mortality Rates (1990-2015) a. Under-Five Mortality Rates b. Neonatal Mortality Rates 300 70 Under-5 mortality (per 1,000 live Neonatal mortality rate (per 1,000 live 250 60 200 50 births) births) 150 40 100 30 50 20 1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015 GNB SSA GMB GNB SSA GMB SLE COM BDI SLE COM BDI Source: World Development Indicators (2018). Source: World Development Indicators (2018). Note: The annual estimates from WDI are partly based on a mathematical model. The MICS estimate is considered more accurate. Figure 5.6: Guinea-Bissau and Peers: Immunization Coverage Rates (1990-2015) a. Measles Immunization Coverage Rates b. DPT (Diphtheria, Pertussis, and Tetanus) Immunization Coverage Rates 100 120 Immunization, measles (% of children ages 12-23 Immunization, DPT (% of children ages 12-23 months) 90 100 80 80 months) 70 60 60 40 50 40 20 30 0 1995 2000 2005 2010 2015 1990 1990 1995 2000 2005 2010 2015 GNB SSA GMB GNB SSA GMB SLE COM BDI SLE COM BDI Source: World Development Indicators (2018). Source: World Development Indicators (2018). 94 Figure 5.7: Guinea-Bissau and Peers: Life Figure 5.8: Guinea-Bissau and Peers: Maternal Expectancy at Birth (1990–2015) Mortality Ratios (1990–2015) 65 3000 60 2500 MMR (per 100,000 live births) Life expectancy at birth, total 55 57 2000 55 50 53 1500 52 45 48 49 1000 40 500 35 0 1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015 GNB COM SLE GNB SSA GMB Source: World Development Indicators (2018). Source: World Development Indicators (2018) Note: GNB: Guinea-Bissau; COM: Comoros; GMB: The Gambia; BDI: Burundi; SLE: Sierra Leone; SSA: Sub-Saharan Africa Figure 5.9: Guinea-Bissau and Peers: Under-Five and Neonatal Mortality Rates (1990-2015) a. Under-Five Mortality Rates b. Neonatal Mortality Rates 300 70 Under-5 mortality (per 1,000 live Neonatal mortality rate (per 1,000 live 250 60 200 50 births) births) 150 40 100 30 50 20 1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015 GNB SSA GMB GNB SSA GMB SLE COM BDI SLE COM BDI Source: World Development Indicators (2018). Source: World Development Indicators (2018). Note: The annual estimates from WDI are partly based on a mathematical model. The MICS estimate is considered more accurate. Figure 5.10: Guinea-Bissau and Peers: Immunization Coverage Rates (1990-2015) a. Measles Immunization Coverage Rates b. DPT (Diphtheria, Pertussis, and Tetanus) Immunization Coverage Rates 100 120 Immunization, measles (% of children ages 12-23 Immunization, DPT (% of children ages 12-23 months) 90 100 80 80 months) 70 60 60 40 50 40 20 30 0 1995 2000 2005 2010 2015 1990 1990 1995 2000 2005 2010 2015 GNB SSA GMB GNB SSA GMB SLE COM BDI SLE COM BDI Source: World Development Indicators (2018). Source: World Development Indicators (2018). 95 10. Guinea-Bissau performs the same or worse than its peers on average (Figures 5.3–5.6). The country’s life expectancy for 2016 is 57.4 years, lower than the average of 62.5 years for LICs, and 60.4 for SSA. Over the last decade it has made progress in reducing infant mortality; the 2012 MDG progress report found that that Guinea-Bissau reduced its infant mortality rate (IMR) by 26 percent over 1990 –2010. However, both the IMR and the under-five mortality rate (U5MR) are among the highest in the world, 57.8 and 88.1 per 1,000 live births, respectively. The decline in neonatal mortality rate (NMR) is comparable to peer countries, but it remains significantly higher overall (Figure 5.6.b). The improvement in child health outcomes shows that targeted priority interventions can make a difference. These include: (i) relatively high coverage of immunizations; (ii) implementation of community-based interventions, including health promotion and prevention; and (iii) high coverage and use of long-lasting insecticide-treated nets (LLIN).8 11. Guinea-Bissau has one of the highest maternal mortality rates in the world . Recent evidence from the 2014 Multi-Indicators Cluster Survey (MICS),9 puts the MMR at 900 per 100,000 live births – the second highest in the world after Sierra Leone (1,360), and slightly higher than Chad (856) and South Sudan (789). The average MMR for LICs is 496, and 547 for Sub-Saharan Africa overall.10 The country failed to achieve the Millennium Development Goal (MDG) for the MMR and, at its current pace, it is unlikely to achieve the Sustainable Development Goal (SDG) for MMR by 2030. 11 Only 45 percent of deliveries take place within health facilities (UNICEF, 2015). On the supply side, challenges include an acute shortage of midwives, weak infrastructure, and limited availability of surgical services and medicines. On the demand side, barriers include under-the-table payments, poor perceptions of service quality, and cultural factors. Volumes of deliveries and outcomes are linked, with a threshold effect of around 500 deliveries per year for facility-based deliveries. Service delivery networks should be reformed to enable women to deliver in higher volume facilities, where proper access to better quality care can be ensured. 12. The country has a very high burden of malnutrition, which directly correlates with maternal and child mortality. According to a national food security assessment conducted by the World Food Program (WFP) in 2013, only 7 percent of the population was food secure. The level of food insecurity is particularly high in rural areas where more than 93 percent of the population may face food insecurity and require immediate assistance.12 Undernutrition is a major public health challenge in Guinea-Bissau, mainly attributable to food insecurity. The national prevalence of acute malnutrition stands at 6 percent, while the prevalence of stunting among children under five is 27.6 percent (UNICEF, 2015). Malnutrition is believed to be the single greatest cause of child mortality in developing countries, as it weakens a child’s ability to recover from a disease that would not kill a better-nourished child.13 13. Ten diseases account for over 70 percent of deaths in Guinea-Bissau (Table 5.1). Cardiovascular diseases are the single biggest cause of deaths (17 percent) and the third biggest cause of Disability Adjusted Life Years (DALYs) in the country, accounting for 7.9 percent. HIV/AIDS also accounts for a significantly higher proportion of deaths in Guinea-Bissau (9.9 percent) and DALYs (9.3 percent) than peer 8 For example, a recent evaluation by the National Institute of Public Health (INASA) found that over 2010-2013, the proportion of children under five and pregnant women who used LLIN increased from 36 to 94 percent, and from 32 to 91 percent, respectively (World Bank, 2016). 9 UNICEF (2015). 10 Note that these are all models estimated per 100,000 live births. 11 The MDG goal was to reduce MMR by 75 percent. The SDG 3 targets are: MMR of 70, U5MR of 25, and NMR of 12 by 2030. 12 The high level of food insecurity was partly attributed to a sharp drop in cashew nut prices in 2013. The Coup d'état in 2012 coincided with the beginning of the cashew nut season, which is the main source of income for more than 80 percent of rural households in Guinea-Bissau. 13 Benson and Shekar (2006). 96 countries. The other top causes of deaths include neonatal disorders, lower respiratory infections, neoplasms, tuberculosis, diarrheal diseases, and diabetes.14 14. In recent years, the incidence of non-communicable diseases (NCDs) grew considerably . This indicates that Guinea-Bissau is beginning to face a double burden of communicable diseases (infectious diseases) and NCDs, placing further pressure on its already weak health system. Since 1990, the proportion of deaths from communicable, maternal, neonatal, and nutritional diseases (CMNNDs) has declined across developing countries. However, while overall mortality has fallen since 1990 in Guinea-Bissau, mortality from CMNNDs has declined more slowly than in comparable countries, and mortality from NCDs and injuries has increased slightly, while it has fallen among its regional and economic peers. Table 5.1: Guinea-Bissau and Peers: Top 10 Causes of Death (2016) Top 10 causes of Deaths Guinea-Bissau Burundi The Gambia Comoros Sierra Leone 2016 Deaths Rank Deaths Rank Deaths Rank Deaths Rank Deaths Rank Cardiovascular diseases 17.0% 1 12.0% 2 14.0% 1 17.8% 1 11.6% 2 HIV/AIDS 9.9% 2 2.8% 10 7.9% 5 0.6% 22 3.8% 9 Neonatal disorders 8.3% 3 10.9% 3 13.0% 2 9.5% 5 10.0% 5 Lower respiratory 8.2% 4 8.9% 5 9.6% 3 9.7% 3 10.4% 3 infections Neoplasms 7.3% 5 6.0% 6 8.5% 4 11.1% 2 5.6% 6 Tuberculosis 6.4% 6 10.9% 4 4.1% 7 6.2% 6 3.7% 10 Diarrheal diseases 5.6% 7 13.0% 1 6.2% 6 9.6% 4 10.3% 4 Diabetes, urogenital, 4.9% 8 2.7% 11 3.9% 9 4.4% 7 4.0% 8 blood, endocrine diseases Unintentional injuries 3.8% 9 3.7% 9 3.9% 8 4.% 8 3.2% 12 Meningitis 3.0% 10 2.4% 12 3.3% 11 2.6% 11 3.2% 11 Malaria 2.3% 15 6.0% 7 2.6% 13 0.02% 32 14.7% 1 Nutritional deficiencies 1.6% 18 4.2% 8 2.6% 12 3.2% 9 1.9% 13 Other NCD 2.4% 14 1.4% 18 3.6% 10 1.5% 18 4.4% 7 Digestive diseases 2.2% 16 2.3% 13 1.9% 17 3.0% 10 1.8% 15 Source: Global Burden of Diseases, 2016 (accessed April 2018). 15. The burden of HIV in Guinea-Bissau is the highest in West Africa while its antiretroviral therapy coverage is one of the lowest (Figure 5.8). The HIV epidemic is pervasive, and women are disproportionately affected (2.1 versus 1.1 percent for men). Women made up 58.7 percent of the population over 15 years old with HIV in 2015. Factors contributing to the spread of HIV infection include limited knowledge of HIV prevention and risky sexual behavior, including multiple sexual partners and inconsistent use of condoms. Underlying socio-economic conditions also play a role in the spread of the infection through various pathways, including distorted risk perception among the poor and vulnerable. 5.3 Health Care Financing 17. Health spending in Guinea-Bissau is comparable with peers, but the composition of its spending is highly problematic. Guinea-Bissau spends a total of 6.9 percent of GDP on health care, close to those of peers (Figure 5.11). However, the country relies heavily on out-of-pocket (OOP) payments and donor resources to fund health services. Neither of these are desirable means of revenue collection: high 14 As in many developing countries, low quality medicines also remain a challenge. Globally, there are currently more deaths from patient safety incidents globally than from HIV or TB. 97 OOP payments expose households to financial shocks linked with ill health,15 while dependence on donor resources has implications for sustainable health financing as it impairs long-term planning.16 Figure 5.11: Guinea-Bissau and Peers: Health Figure 5.12: Guinea-Bissau and Peers: Government Spending per Capita Health Spending (2015), % total health spending 24 50 GMB Total health expenditure (% of GDP) Government health spending (% total) BDI SLE 40 SSA 19 GNB 30 14 LIC 20 COM COM BDI SLE 9 GNB LIC GMB 10 SSA 4 0 Source: World Development Indicators (2018). Source: World Development Indicators (2018) Figure 5.13: Guinea-Bissau and Peers: Out-of- Figure 5.14: Guinea-Bissau and Peers: Trends in Out- Pocket Payments (2015), % total health spending of-Pocket Payments (2000–15) 80 COM 100 Out-of-pocket expenditure (% of total Out-of-pocket expenditure (% of total 80 60 60 spemding) LIC SLE spending) 40 GNB SSA 40 GMB BDI 20 20 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 0 Source: World Development Indicators (2018). GNB SSA GMB SLE COM BDI Source: World Development Indicators (2018). Figure 5.15: Guinea-Bissau and Peers: External Health Spending, % total health spending 60 SLE External health expenditure (% of 50 BDI 40 total) GNB LIC 30 GMB 20 SSA COM 10 0 Source: World Development Indicators (2018). 15 Ke Xu et al. (2003) 16 Gottret and Schieber (2006). 98 18. Total government spending on health is, however, significantly lower than some of its regional and economic peers (Figure 5.12 and Table 5.2). In 2016, public health spending represented merely 31.3 percent of total health spending, less than Burundi (38.8 percent) and The Gambia (46.6 percent), but higher than that for Sierra Leone (9.0 percent) and Comoros (13.4 percent). Only 9.5 percent of general government expenditure has been allocated to health, far below Abuja Target of 15 percent. 19. Total health spending per capita in Guinea-Bissau is close to the LIC average but lower than the SSA average (Figure 5.9). In 2016, Guinea-Bissau spent US$39.5 per capita on health, while LICs spent US$35.3 per capita and the SSA average was US$84.9. Guinea-Bissau has seen an overall steady increase in per capita health spending, which more than doubled between 2000 and 2014, increasing from US$21.2 to US$43.9. This growth pattern in health spending has been in line with its West African peers, while the wider SSA region has seen a much faster growth rate, especially over the last ten years. 20. Households bear a high proportion of total health expenditures in Guinea-Bissau. On average, households spend 15 percent of their non-food expenditures on health care, but this masks wide variation across different income groups. In 2015, OOP payments represented 37.2 percent of total health expenditure, comparable to the SSA average of 36.3 percent and below the LIC average of 40.4 percent (Figure 5.13). The share of OOP payments in Guinea-Bissau has been higher than 45 percent of total health spending since 2000 (Figure 5.14). Table 5.2: Guinea-Bissau and Peers: Health Financing Indicators (2015) Country Total health expenditure Government OOP expenditure External Private health health health expenditure expenditure expenditure % of Per Per % total % gov. Per Per (% total) (% total) GDP capita capita exp. capita capita (US$) (PPP) (US$) (US$) Guinea-Bissau 6.9 39.5 100.3 31.3 9.5 14.7 37.2 31.6 37.2 Burundi 8.2 24.3 63.7 38.8 11.8 4.6 19.1 40.5 20.7 The Gambia 6.7 31.9 114.1 46.6 10.6 6.4 20.3 27.8 25.6 Sierra Leone 18.3 106.7 256.3 9.0 7.9 40.8 38.2 52.6 38.4 Comoros 8.0 58.6 121.0 13.4 3.8 43.8 74.8 10.1 76.5 SSA 5.4 84.9 198.7 34.7 N/A 30.7 36.3 12.9 52.1 LIC 6.0 35.3 94.5 20.3 N/A 14.3 40.4 31.4 47.4 Note: PPP: purchasing power parity. Source: World Development Indicators (2018). 26. OOP payments have not resulted in improved access to health services . The number of births assisted by a qualified health professional is low in regions with high OOP payments (Gabú and Bafatá), with the exception of Bissau (World Bank, 2016). Similarly, a larger share of pregnant women have four or more antenatal visits in the regions with relatively low OOP payments (Quinara and Oio). Neither are higher OOP payments associated with improved health outcomes: Oio, Quinara, and Biombo are the regions with the lowest neonatal and child mortality rates, and also have relatively lower OOP payments. In contrast, households in Gabú and Bafatá spend more on health care through OOP and face higher-than-average mortality rates.17 OOP payments for health care increase the absolute and extreme poverty headcount ratio Note, however, that this should be interpreted with caution since only two indicators were available for comparison, and there might 17 be several factors affecting health status across regions. 99 by 1.4 and 1.1 percentage points (World Bank, 2016). In other words, about 1.1 percent of the population (roughly 15,000 people) are driven into extreme poverty due to health care payments. 27. Guinea-Bissau receives a significant share of development assistance for health (Figure 5.15). Low income and Sub-Saharan African countries are increasingly relying on domestic resources for health spending. More than 30 percent of Guinea-Bissau’s total health spending is financed through external resources, about the same as the average for LICs (31.4 percent), but more than the Sub-Saharan African average of 12.9 percent. During the past 15 years, Guinea-Bissau has relied heavily on external resources, even though donor financing is more unpredictable and often managed outside the central government. 5.4 Public Financial Management (PFM) in the Health Sector 5.4.1 Overview of the PFM system 28. This section assesses public financial management (PFM) systems and practices in the health sector. The assessment also sheds light on the likely impact of PFM weaknesses on strategic allocation of resources and efficient provision of services. In terms of the overall PFM landscape, Guinea-Bissau has made modest progress, although political instability has proved to be the nation’s Achilles’ heel. According to the 2013 Public Expenditure and Financial Accountability (PEFA) assessment, some improvements were made in the areas of budget development and execution, despite subsequent reversals. In 2017, the reinstatement and strengthening of the Treasury Management Committee helped improve spending controls. 29. Similar to the case in other line ministries, PFM systems in the health sector are weak, partly due to chronic political and institutional instability . Guinea-Bissau has partially aligned its PFM legal framework with WAEMU directives, but the framework has yet to be translated into specific processes or procedure manuals. Although a Committee was formed tasked with these responsibilities, no significant progress has been made. The inability of the state to perform the most basic functions of a responsible and accountable PFM system has resulted in weak governance arrangements. For instance, significant salaryq payment delays and other grievances, followed by periods of strikes, have jeopardized the quality of services delivered in the health sector, notwithstanding some improvements in 2017. 30. The institutional arrangements and legal framework at the Ministry of Health are incomplete and weak. The organic law of the MOH – which defines its attributions and establishes how the ministry will be organized and function – has been approved by the Council of Ministers. However, the organic law has not yet been approved by the President or promulgated. PFM arrangements and policies in the health sector rely on the general PFM practices that exist at the Ministry of Economy and Finance (MEF). The actual procedures in place are largely based on informal practices, without any specific and formally established procedures or manuals. Similarly, legal advisory services at the MOH are weak and rely on an understaffed Legal Department, with the entire sector relying on one legal advisor for legal services. 31. In recent years, Guinea-Bissau has made some improvements in the area of strategic planning. This includes the elaboration of the National Plan for Health Improvement (Plano Nacional do Desenvolvimento Sanitário, PNDS), which is based on sector priorities, as presented in the government’s 2014 National Development Plan, Terra Ranka. However, the policies outlined in the five-year medium-term PNDS III (currently being finalized) are general, making it hard to link policy priorities with the budget. 100 Guinea-Bissau’s 1993 National Health Policy was revised in 2015, but it has not yet been approved by the Council of Ministers and not yet operational. 32. For the first time, a three-year budget (2018–20) has been prepared based on the PNDS III, in accordance with WAEMU directives. Nonetheless, the budget elaboration process has not yet been institutionalized in the health sector. The Administration and Finance Directorate (DAF) of the MOH is mainly responsible for the elaboration of the budget but it lacks dedicated and skilled staff. It is noteworthy that only a small portion of the health budget is allocated to operating expenses, goods and services, or investments. Most of these expenditures are partly covered by internal revenues and external donors at the central and decentralized levels, while investment is exclusively financed by development partners. 33. Internal revenues and expenditures are not reported in any financial statement or reports. The internal revenues of the health sector are not reported in annual budgets and nor are the expenditures paid out of such revenues. These financial transactions are monitored by neither the MOH nor the MEF and there is no established procedure or proper accounting system to record and report them. As a result, there is no reliable information about the overall budget and expenditure of the health sector. Only Entraide Médicale Internationale (EMI) – a non-governmental organization – is monitoring a certain number of health centers and hospitals, and giving incentives to medical staff based on performance criteria. 34. Fund transfers to and from the decentralized health entities are neither clear, not reported on state accounts. No information is available other than on the transfer of funds (Fundo de Maneio) to the regional directorates and/or healthcare centers from the central government. Salaries for health sector staff are paid directly by the MEF. Health centers and hospitals do not receive any funds for operating costs. They receive only marginal contributions from the Fundo de Maneio, mainly for food, which are insufficient to cover their operational expenses. Health centers and hospitals in turn make monthly transfers, amounting to 5 percent of the internal revenues they collect, to the regional health directorates. However, there is not much evidence about these transfers nor about how the regional directorates use them. Other funds enter the health sector directly at the decentralized level from donors, but they are not reported in the budget or any reports. The procedures are not fully established and transparent, but they are being monitored by EMI. 35. The accounting and reporting functions are not satisfactory . Data and reports are replete with errors and often inconsistent with each other. This is partly due to the concentration of IT systems within MEF and a lack of interconnection between the systems at the MEF, the MOH, and the Ministry of Public Administration (MPA). The MOH has a variety of other small systems, such as spreadsheets and two HR databases, but they have had only limited use beyond serving as a response to donor requirements. Guinea- Bissau’s accounting and reporting functions do not meet the following articles of the WAEMU Directive No.7: (i) Article 26, concerning control over internal revenues and related expenses; (ii) Article 27, related to the reconciliation of public accounts; (iii) Article 28, submission of annual state accounts to the SAI by end of a fiscal year; (iv) Article 57, the use of the Single Treasury Account for all government revenues and expenditures; and (v) Article 69, concerning public assets, particularly related to cost calculation and performance of the services provided. 36. Payroll controls in the health sector are improving, but they remain very weak . Some steps have been taken toward effective control of human resources and the payroll system, with the development of the AXYS tool for the management of the HR database. However, this process has not yet been finalized, partly because the two databases, the payroll system and the personnel record database, have not been 101 integrated. There is no any formal auditing process of the payroll function at the MOH or the wider public administration, other than the AXYS check of the payroll. 37. Guinea-Bissau’s unreliable and unconnected financial management infor mation systems are unable to produce consolidate financial information. The MOH and the regional and decentralized health entities have no direct access to the systems used by other ministries. The existing systems are mostly centralized at the MEF, with one of them held by the MPA. In order to enter expenses into the MEF’s integrated financial management system (SIGFIP), the DAF specialist at the MOH has to physically move all the necessary documents and receipts to the MEF and manually enter them. Similarly, an HR specialist from MOH would have to go to the MPA to review and make any changes related to personal records in the MPA’s personnel management database (SIGRHAP). The management unit within the Directorate General for Health Administration (DGASS) uses a relatively better performing financial management information system. However, the procurement and HR modules have not been used. 38. Weak human resource management is one of the key impediments to improving economic and financial governance, as well as the efficiency of health service delivery. HR management systems and capacity at the MOH are largely unsatisfactory because of underdeveloped HR arrangements; a lack of health workforce policies and competency-based training plans; an unclear career pathway for different groups of health workers; lack of an efficient remuneration policy; and weaknesses in the SIGRHAP database. In addition, civil servants in the HR division do not have the necessary skills and experience. Poor hiring criteria and high staff turnover undermine the quality of the delivery of health care services. The HR Directorate at the MOH is not able to: (i) supervise individual departments and ensure that managers adhere to employment and labor regulations; (ii) manage the employment and placement of health care staff; and (iii) meet a multitude of objectives regarding hiring, employee relations, training and organizational structure. 5.4.2 Linking Service Delivery Bottlenecks to PFM Arrangements 32. This section uses the “fishbone” framework to identify specific health financing and PFM weaknesses that are linked with proximate causes of health service delivery bottlenecks . As noted above, health service delivery in Guinea-Bissau is affected by health financing and existing PFM systems. The “fishbone” approach18 is used in health systems research to identify causal chains for a variety of service delivery constraints. Data for the analysis were collected through the service delivery indicator (SDI) instrument, a health facility review instrument, and interviews with relevant officials in the health and finance ministries, selected health facilities in five regions of the country, and other development partners supporting the health sector. 33. From the information gathered, we identified several persistent challenges faced by the country’s health system: (i) low public health spending; (ii) poor infrastructure; (iii) inadequate clinical and 18 The fishbone diagram visually maps out the major underlying factors that contribute to a problem or outcome. To populate the fishbone diagrams the research team used a combination of sources. A review of the literature enabled identifying certain health financing and PFM issues that could potentially constrain service delivery. This was complemented with information sourced from interviews with key informants. The interviews were conducted using semi-structured questionnaires, after receiving the informant’s permission. Information contained in the interviews was initially hand-coded line-by-line by the study investigators. For each of these interviews, separate matrices for hospitals and primary health centers was constructed compiling respondent reports on service delivery constraints, their health financing and PFM causes, and other causes. Individual matrices were discussed by the research team and through these deliberations, and iterative consultative process, the matrices were refined and elements rearranged. The research team then identified certain service delivery constraints around which the fishbone(s) were developed. 102 managerial training systems; (iv) malfunctioning referral systems; and (v) non-operational health- information systems. Of these, three service delivery areas stood out, based on their financing and service delivery implications as well as their reported importance: (i) inadequate health sector resource allocation and operational budget; (ii) the budget preparation process; and (iii) internal and external controls. The Challenge of Resource Allocation: Linking Resources with Sector Priorities 34. The health sector accounted for only 7–8 percent of the total government budget over 2010- 17 and even this is not fully executed. The health budget is hardly enough to support the government priorities elaborated in National Health Policy 2017 and the PNDS III. Guinea-Bissau’s current level of health spending also falls far short of the 15 percent Abuja Declaration target the country has committed to. Moreover, the MOH has not been able to execute its budget entirely, sometimes spending less than half of its annual budget appropriation. Budget execution at the MOH fell as low as 50 percent in 2015 and 42 percent in 2016, lowering actual health expenditure to 3.5 percent of the state budget. Government health spending is not only low but also hugely skewed towards staff costs. Roughly 90 percent of the health budget is spent on personnel costs, more than half of which is used to pay salaries alone. Figure 5.16: Financial Flows in the Health Sector Note: The continuous lines stand for actual flow of funds, blue lines for in-budget flows, black lines for out-of-budget, red lines for informal payments, long-dashed lines for actual but irregular flows, and short-dashed lines unrealized flows. 35. The only category with high budget execution rates was personnel costs . While this does not necessarily reflect any prioritization of health sector staffing, the high execution rates have benefited from established payment mechanisms. In other areas, where such mechanisms are lacking, low execution rates indicate lack of capacity in strategic planning and implementing health policy actions in the medium and long term.19 Low budget execution rates reveal weaknesses of the PFM system in terms of preparing reliable budgets, elaborating cash flow projections, and transparency in intragovernmental transfers. However, in 19 For example, in 2015, the MOH allocated 1 percent of its budget to investments and approximately half of the funds were allocated to purchasing goods and services. 103 2017, some improvements have been observed in health budget execution, raising the executed health expenditures to 70 percent of the overall state budget. The Planning and Budgeting Process 39. The budgeting process in the MOH is more a “fundraising” tool than a planning and budgeting document. Most PFM tasks are concentrated in the MEF, while line ministries and other public entities have a high degree of autonomy over the management of internal revenue. The MEF is the sole steward of the fiscal and budgetary policy, and allocates fiscal space to the ministries. The budget preparation process thus follows a top-down approach. There are no staff in the MOH responsible for budget preparation and elaboration. Overall, the budgeting process is very weak and does not reflect the real needs of the sector given: (i) it is largely tailored to the fit the fiscal space imposed by MEF; (ii) there is limited consultation within MOH and almost no contribution from decentralized entities; (iii) in some cases, the programs in the budget are not closely tied to the sector’s real needs or target indicators; (iv) there are no established procedures for elaborating rolling plans and budgets; and (v) the budget lacks various scenarios, responding to pessimistic and optimistic cases. 36. The weak fiscal framework has resulted in very limited flow of funds to the health sector, stimulating out-of-pocket (OOP) and informal payments, and undermining service provision . The health system in Guinea-Bissau is funded from three main financial sources: public funds, private funds, and external funds (Figure 5.16). Public funds mainly include tax revenues, fishing licenses, and timber exports. Private sources are mainly OOP payments, but also informal payments from patients. External funds come from international donors and organizations, which may be earmarked for specific programmatic expenditures, in and out of the budget. Internal and External Controls: Transparency 37. Internal controls in the health sector are very weak. Internal control functions include, among others, safeguarding assets, ensuring proper records, good-quality and timely reporting, and compliance with applicable laws and regulations as well as with internal policies. Internal controls remain weak mainly because they rely on outdated policies, unofficial organic laws, and some informally established procedures within the MOH and decentralized units. The internal control mechanisms for non-salary expenditures rely on organizational arrangements that exist in the expenditure chain, which cannot substantially eliminate or reduce the risk of mismanagement and/or corruption. The chain of budget execution starts in the sector ministry which prepares the expenditure/payment process before settlement is referred to the MEF in hard copy and the related funds are entered into SIGFIP. 38. The effectiveness of internal audit is undermined by the fact that the General Inspection at the MOH has no legal base or criteria for auditing. The General Inspection for Health Affairs (Inspeção- Geral para Assuntos da Saúde; IGAS) is responsible for internal audit functions in the health sector, which include: internal regulation, a procedures manual, the career of the inspectors, the regulation of health care centers. There are four inspectors at the central level, who specialize in HR management, pharmacies, clinics, and sanitation, and 3-4 inspectors in the regions. However, regulations remain outdated while internal audit plans rarely executed, partly because of a high turnover of inspectors. Although IGAS is autonomous, it is mostly dysfunctional due to capacity limitations and lack of resources, limited information technology, and 104 weak communication between the GI and other control entities (the Court of Auditors, the Financial Inspection General-MEF, the MPA, and the Anti-Corruption Entity).20 39. Financial reporting in the MOH is incomplete and largely unsatisfactory . Financial reports do not include the internal revenue and related expenditures of various line ministries, including the MOH. According to Art. 80 of the Directive 07, the state's general accounts should be able to produce the financial statements of the state, including the balance sheet, the income statement, the financial operations table of the Treasury (TOFE), the cash flow chart; however, these have never been achieved. In fact, financial reports are drawn up at the level of engagement (settlement), two levels before the payment of expenses which in many cases never happens or is realized with significant delays. Tracking delayed payments is a cumbersome exercise conducted by the MOH DAF and is not performed on a regular basis. The TOFE presents information on budget execution, but it is not used as a decision-making tool by line ministries. 40. The Supreme Audit Institution’s contribution to health sector performance, transparency and proper use of the allocated funds, and safeguarding of assets has been limited. The Supreme Audit Institution (TdC – Tribunal de Contas) lacks the technical capacity and the necessary logistics, and has yet to establish itself as a supreme entity for the external control of public finances. The activities of the TdC remain primarily oriented towards the prior control of the legality and the budgetary coverage of the acts and contracts of the State. In general, the external control by the TdC or by the National Assembly is very weak. The National Assembly has no any finance committee able to review external audit reports. The audits carried out by the TdC cover a very small percentage of the total government spending. 20 The GI has performed some inspections but in a very limited radius with private vehicles and laptops belonging to the inspectors and no budget for fuel and maintenance. The internal audit entities at MEN and MEF and the SAI began to carry out some selected audits at the regional level, but in the absence of specific established procedures, they cannot make any relevant recommendation. 105 Potential Priority Actions – Security Sector Measure Technical Timeframe of feasibility policy impact Deepen analysis of retirement and pensions - Obtain understanding, through an actuarial study, of how security High Short term personnel retirement affects pensions and fiscal space. - Explore options for downsizing security personnel approaching High Medium term retirement and make space for other spending. - Determine the role of gratuities of military and security personnel as High Short term well as of veterans in the context of a budget for demobilization and reintegration. Identify priorities to guide government spending on security - Elicit a security sector reform (SSR) strategy based on objectives for High Medium term security provision set against a realistic budget envelope. - Rebalance personnel, operations, and capital spending to ensure that Medium Medium term SSR objectives are met. Improve the efficiency of public financial management - Conduct a PFM review focused on payroll and personnel High Short term management to obtain understanding how to improve public spending and accountability in all phases of the budget process. 106 Chapter 6: Security Sector1 The security sector has played a central role in the formation of Guinea-Bissau after independence, with the country still facing continued political instability and international threats from money laundering to drug trafficking. Attempts at security sector reform remain a work in progress, despite significant international support. On average, security spending accounts for about 15 percent of the national budget, with the lion’s share representing wages and salaries. The current security forces are rapidly ageing with many personnel already over retirement age, and a top-heavy hierarchical structure. In order to rebalance its security and development priorities, the government should (i) address the outstanding issue of retirement, pensions and Liberation War veterans to explore downsizing the security forces and make adjustments to the rank structure; and (ii) rebalance its sector budget away from salaries in order to modernize and professionalize the sector. 1. This chapter provides a rapid assessment of the fiscal affordability of the security sector (military and internal security) in Guinea-Bissau. It uses a tried-and-tested approach that has been applied in over 20 countries. World Bank staff undertook missions in December 2017 and May 2018, reviewed relevant materials,2 and worked with the International Security Sector Advisory Team (ISSAT), which has both the relevant sector and country expertise. However, the present exercise does not represent an exhaustive, in-depth review of the security sector. Volatile political circumstances did not allow for a full assessment, working with government, of sector policy options, and their cost implications. The present PER does not also assess systems and controls over the use of public finances, nor review criminal justice institutions (other than the Public Order Police and the National Guard 3) or the intelligence services. The analysis sets up a macroeconomic foundation for potential security sector reforms (SSRs), and prepares the ground for a future in-depth and more comprehensive assessment of the security sector. 2. The chapter is organized as follows. Section 6.1 outlines the main historical and sectoral challenges pertaining to security and justice institutions in Guinea-Bissau, including various phases of its security sector reform efforts. Section 6.2 presents an overview of security and justice institutions, with an emphasis on the composition of security spending by economic and functional classification. Section 6.3 examines the affordability of security spending against revenue and growth projections as well as benchmark trends in security expenditures against other fragile and neighboring countries. Finally, Section 6.4 concludes with a discussion of options for policy reforms. 6.1 Historical and Sectoral Context 6.1.1 Historical Context 3. Guinea-Bissau’s pathway to state building has been characterized by instability, with chronic armed and political violence starting long before independence in 1974 . The cycle of power contestation has prevented the consolidation of political, security, economic, and administrative structures 1 Prepared by Bernard Harborne (Lead Technical Specialist, GSUGL) with contirbutions from Andry Ralijaona (Consultant, GSUGL); Paul Maximilian Bisca (Consultant, GTFMR), and Lala Anicia (ISSAT). 2 This includes the recent comprehensive analysis of UNIOGBIS and ISS, Relaunching Defence and Security Sector Reforms in Guinea-Bissau, May 2018. https://issafrica.org/research/books-and-other-publications/relaunching-defence-and-security-sector-reforms-in-guinea-bissau. 3 The Judiciary Police which is the main organ of criminal investigation falls under the aegis of the Ministry of Justice. The same applies to the national Interpol office and the Transnational Crime Unit, meaning none of these police organs were covered in the study. 107 needed to ensure the delivery of public services and guarantee the safety and security of the population. Repeated military coups, fragmented elites, and high levels of rent seeking have long impeded institution building and key reforms. Deep-seated political and institutional fragility has been manifested in a fractured social contract between the state and society. The state is incapable of performing even basic functions and has limited presence in much of the country. 4. Guinea-Bissau inherited weak institutions with significant ramifications for its development trajectory. Unlike other Portuguese colonies, it was not considered fit for settlement and, prior to independence, the colonial administration had made no significant investment in the country beyond what was required to extract natural resources.4 Other than a small local elite, Bissau-Guineans received virtually no provision of health care, education, or social services, and were unable to access any aspect of the public administration.5 At the time of independence, Guinea-Bissau’s literacy rate was 1 percent, and the country had only 60 kilometers of paved roads. The Liberation War, which was one of the bloodiest in Africa, did not automatically deliver improved livelihoods and peace. Instead, the country has been beset by one episode of political instability after another, starting in 1980 when the first elected president was ousted in the country’s first military coup—leading to the peaceful separation from Cape Verde. Box 6.1: International Sanctions In 2012 the EU adopted targeted restrictive measures against certain persons, entities, and bodies who aimed to block a peaceful political process. This measure focuses on actors who played a leading role in the mutiny of April 2010, and in the coup d’état of April 12, 2012, and who continued to try to undermine the rule of law and the primacy of civilian power. Sanctions included a travel ban and the freezing of funds and economic resources of the individuals and entities concerned. The UN Security Council also imposed a travel ban on individuals seeking to prevent the restoration of constitutional order. These sanctions are still in place, given the prevailing level of political instability in the country, and they have helped restrict attempts at further unconstitutional changes of power. Source: EU (Guinea-Bissau Sanctions) Order 2012, Statutory Document No. 0476/12, Security Council Report, 06/04/2017, 5. Guinea-Bissau also experienced periods of exacerbated violence during the military confrontations of the 1998–99 civil war, and after the 2012 coup (Figure 6.1). The exploitation of ethnicity and factionalism for political ends led to multiple assassinations, political crises, and four successful coups since independence, preventing the consolidation of state institutions. Since the election in late 2014, the armed forces have abstained from intervening in politics. This is likely due to a combination of deterrence by the Economic Community of West African States (ECOWAS) mission stationed in Bissau (ECOMIB) and the imposition of sanctions by the European Union and the United Nations after the 2012 coup (Box 6.1).6 The current military leadership has also promoted a non-interventionist ethos. 6. However, political volatility has persisted, requiring mediation by ECOWAS, which resulted in the signing of the Conakry Agreement in October 2016 as the basis for return to stability . The basic provisions of the Agreement include commitments toward an inclusive government, sharing of power amongst the parties and participation of civil society. Most importantly, it also signaled the need for defense, security and justice sector reform.7 4 World Bank (2016) Guinea-Bissau Systematic Country Diagnostic (SCD). 5 http://documents.worldbank.org/curated/en/100721467968248103/pdf/106725-CSD-P155168-IDA-SecM2016-0127-IFC-SecM2016-0078- MIGA-SecM2016-0076-Box396273B-PUBLIC-disclosed-7-5-16.pdf 6 Decision 2012/237/CFSP, 03/05/2012, UN Security Council Resolution 2048, 18/05/2012. 7 UNSG Statement, 16 December 2016, SG/SM/18371-AFR/3507 108 Figure 6.1: Political Violence and Protest in Guinea-Bissau (1998–2018) Source: ACLED database, 2018 Figure 6.2: Political Instability in Guinea-Bissau and Impact on GDP (1971–2014) Source: World Bank (2016) SCD. Figure 6.3: Governance Indicators (2006, 2011, Figure 6.4: Ibrahim Index of Governance (2012–16) 2016) 100 80 60 40 20 0 2012 2013 2014 2015 2016 OVERALL GOVERNANCE SAFETY & RULE OF LAW RULE OF LAW NATIONAL SECURITY Source: World Bank (2018) World Governance Indicators. Source: Ibrahim Governance Index. 109 7. The failure by key political actors to comply with these terms led to ECOWAS imposing additional sanctions on persons or organizations impeding the process for ending the crisis .8 The lack of progress in the implementation of the Conakry Agreement was accompanied by a rise in social and political protest, occasionally met by the authorities with punitive crowd-control measures. However, in April 2018, the president appointed a new consensus prime minister who formed a government that resulted from a political consensus reached between the two major parties in the parliament, PAIGC and PRS, and also endorsed by ECOWAS. 6.1.2 Sector Context and Sources of Insecurity Military Involvement in Domestic Politics 8. Guinea-Bissau is one of the most coup-prone and politically unstable countries in the world. Between 1999 and 2009, Guinea-Bissau experienced a change of government every year and in the last 18 months the country has had four governments.9 Four successful coups have been recorded in independent Guinea-Bissau, with another 16 attempted, plotted, or alleged. Among the previous five chiefs of staff of the armed forces, three were murdered and one was forcibly exiled by rival officers. Guinea-Bissau has one of the highest incidence of attempted or successful coups recorded. 9. The costs of political and economic mismanagement, notably from the 2012 coup, have been severe, causing a sharp decline in donor support and further suppressing the fledgling private sector (Figure 6.2). The military-backed government mismanaged the 2012/13 cashew campaigns, aggravating income losses among farmers and domestic traders. Timber smuggling increased when border controls were eased, and corruption spread. The immediate suspension of donor programs in response to the coup dried up external financing. Public investment ground to a halt, current spending had to be cut, and arrears increased.10 In a similar vein, in 2015, some donors withdrew support to force a resolution of the political gridlock and a reversal of the illegal bank bailout, resulting in a suspension of direct budget support. Total budget and project support plunged from 12 percent of GDP in 2014 to 6 percent in 2017. 6.1.3 Security and Justice Institutions 10. Relations between the political elite and the military continue to be fragile. State fragility and recurrent instability have hampered repeated efforts to foster the rule of law and to build public institutions on a foundation of good governance that will be respected by the security sector. Weak oversight and accountability within the security structures have also persisted, together with capacity gaps within the parliament, and weak collaborative links between civil society and the main security stakeholders. 11 Parliament has not held an official session between 2015 and mid-April 2018 due to the political gridlock. 11. The political impasse has undermined security and justice institutions. Although the security sector accounts for 35 percent of the government’s wage bill and about 15 percent of the national budget, 12 inefficiencies in spending, lack of government commitment, fragmented elites, and rent-seeking activities 8 ECOWAS Decision A/DEC.2.01/2018 (05/02/2018). 9 World Bank (2017) Country Partnership Framework (FY18-FY21) 10 World Bank (2016) Systematic Country Diagnostics. 11 https://issat.dcaf.ch/Learn/Resource-Library/Country-Profiles/Guinea-Bissau-SSR-Background-Note 12 World Bank (2015) Country Economic Memorandum (CEM). 110 undermine the functioning of the sector. The primary challenges for Guinea-Bissau’s security sector relate to the value for money of security spending, and a chronic lack of effective governance and accountability. 12. The internal security agencies have very limited presence outside the capital and a few small urban hubs.13 Actual policing and provision of justice continue to be left to non-state, informal, or customary authorities. Where a public police presence exists, poor understanding of mandates and roles, and the absence of other formal actors along the criminal justice chain, means that the police often end up meting out justice and performing duties outside their remit. Police is absent in most of the territory, border control management remains weak, and operational capabilities to patrol the Bijagós Archipelago are insufficient. As a result, Guinea-Bissau has been unable to combat transnational criminal networks involved in trafficking drugs, timber, and arms as well as people-smuggling operations across the sub-region. 13. For these reasons, safety and rule of law in Guinea-Bissau remain problematic with a rank of 45/54, and a score of 32.6/100 in the Ibrahim Index of African Governance (Figure 6.4).14 These low figures reflect poor regulatory frameworks and weak institutional capacity for enforcement, as well as problems in access to justice, the judicial process, judicial independence, and property rights. In addition, there is widespread impunity over high-profile assassinations, as well as criminality, including organized crime and lower-level offenses. However, despite deteriorating scores for personal safety, Guinea-Bissau’s scores for the rule of law, accountability, and national security have been improving. National security in particular has benefitted from reduced government involvement in internal and external armed conflicts, including in the border areas of Cassamance, Senegal. 14. The World Governance Indicators confirm these trends (Figure 6.3). Despite its low performance in the area of governance and accountability, Guinea-Bissau has maintained somewhat higher scores in “voice and accountability”, and a degree of political stability related to the absence of violent extremism. On the other hand, its scores for government effectiveness and regulatory quality have deteriorated, partly due to the high turnover of governments and the virtual parliamentary freeze in recent years. Its capacity to control corruption also appears to have visibly diminished —or corruption has increased significantly without any commensurate improvements in the country’s capacity to address it. 15. The criminal justice system has limited capacity to perform the most basic functions of a responsible and accountable justice system. Guinea-Bissau’s justice institutions are struggling to provide legal aid, and the protection of women, and to investigate, prosecute, and effectively sentence crimes of sexual and gender-based violence. As a result, a high percentage of the population relies on customary justice. In addition, at the central level, the justice sector faces many obstacles. Cases of politicians suspected of embezzling public money, or members of the security forces accused of involvement in violent crimes, if investigated at all, usually end up in judicial stand-offs. Police and magistrates are often the targets of both corruption and intimidation, and the resulting impunity and the absence of the rule of law have had a deleterious effect on judicial development beyond the criminal domain. 16. An important indicator of the degree of impunity in Guinea-Bissau is the endemic nature of sexual and gender based violence, and domestic violence. Over the last decade, the approval of Law 6/2014 improved the criminalization of domestic violence. Despite this progress, legal gaps remain around early and forced marriage, levirate marriage,15 and polygamy and child prostitution. The lockdown of the 13 https://issat.dcaf.ch/Learn/Resource-Library/Country-Profiles/Guinea-Bissau-SSR-Background-Note 14 http://mo.ibrahim.foundation/iiag/ 15 Under levirate marriage when a man dies, his brother would be obliged to marry his wife. 111 legislature in this last political cycle has prevented the passing of any further legislation and the implementation of the law is still falling short. 6.1.4 Regional and International Context 17. Despite being a small country that has been at relative peace for nearly two decades, Guinea- Bissau confronts several historical as well as more recent challenges related to its geographical and political position. Data on drug seizures or rates of crime and violence, are limited but a number of the key trends are outlined below. Armed Rebellion 18. The Movement of Democratic Forces of Casamance (MFDC) —the longest-running active separatist movement in Sub-Saharan Africa—resides along Guinea-Bissau’s northern border. Since 1982, MFDC rebels have been fighting for independence for the southwestern region of Senegal between The Gambia and Guinea-Bissau, known as the Casamance. The MFDC is currently effectively in a ceasefire with the government of Senegal after fighting a long campaign to gain independence, using Guinea-Bissau as a sanctuary. During peaks in the confrontation, Guinea-Bissau was affected by inflows of refugees along this border, reinforced by the ethnic closeness of the populations. There are considerable unattended movements of people and goods across the border, including allegedly illegal small arms and light weapons trade. The violence associated with the MFDC in the past has diminished considerably in recent years. Drug Trafficking 19. West Africa has also become an attractive destination for transnational organized crime . Drug cartels have collaborated with local partners to turn the region into a significant transit route to Europe and North America for illicit drugs produced in South America and Asia (Figure 6.5). West Africa has long produced cannabis, mainly for local consumption, but is now also becoming a producer and exporter of synthetic drugs such as amphetamine-type stimulants.16 At least 50 tons of cocaine are transferred from the Andean countries to West Africa every year, on their eventual way northwards to Europe. In 2016, roughly two thirds of the drugs that originated in Colombia, Peru, and Bolivia passed through Cape Verde, Mali, Guinea-Bissau, Ghana, Benin, Togo, and Nigeria on their way to Western Europe. This figure is increasing, and West Africa’s smuggling output has already doubled since 2008. 17 20. The growth in drug trafficking comes as the region is emerging from years of political conflict and, in some countries, prolonged violence. This instability has left a legacy of fragile state institutions and weak criminal justice systems that are vulnerable to infiltration and corruption by organized crime and are hard pressed to keep up with the quickly adapting skills of the traffickers. Guinea-Bissau’s vulnerability to drug trafficking is enabled by several factors, including: (i) a considerable portion of its territory located offshore; (ii) extreme poverty and high unemployment; (iii) inability to monitor the shipping traffic of its 88 islands; (iv) ineffective customs and law enforcement; (v) political instability; and (vi) the involvement of the military and well-connected elites in trafficking accompanied by widespread impunity.18 16 http://www.wacommissionondrugs.org/WACD_report_June_2014_english.pdf 17 https://encyclopediageopolitica.com/2018/04/26/the-geopolitics-of-narcotics-the-west-africa-corridor/ 18 Former Rear Admiral Jose Americo Bubo Na Tchuto was sentenced to imprisonment in a New York court, US for drug-trafficking. Pierson and Raymond, Reuters, October 4, 2016 112 Figure 6.5: Drug Trade Routes to Western Africa Money Laundering 21. Guinea-Bissau has yet to achieve full compliance with international standards and agreements against money laundering and financing of terrorism . There is poor supervision of the formal financial sector and the Financial Intelligence Unit is only partially functional, owing to its lack of resources, analytical staff, and technical equipment. From 2015 to 2018 six cases were investigated and submitted to the Attorney General’s Office, but only one process proceeded to the legal accusation phase. The country has not developed adequate legal provisions for the conduct of customer due-diligence procedures,19 and has yet to fully incorporate into national legislation and implement the Anti-Money Laundering Uniform Law, a requirement for members of the West African Economic and Monetary Union (WAEMU) adopted in July 2015. Child Labor, Sexual Exploitation, and Trafficking 22. The country serves both as a source of and a destination for trafficked children . Children tend to be employed in forced begging, forced agricultural labor and mining, domestic work, and commercial sexual exploitation. Despite the adoption of anti-trafficking laws in 2011, there is little capacity to investigate cases and enforce the law, and inadequate strategies to protect children or to prevent their trafficking. UNICEF estimates that 200 children in Guinea-Bissau are trafficked each month, with the majority of victims 19 Article 26 of National Assembly Resolution No. 4 of 2004 stipulates that if a bank suspects money laundering it must obtain a declaration of all properties and assets from the subject and notify the Attorney General, who must then appoint a judge to invest igate. The bank’s solicitation of an asset list from its client could amount to informing the subject of an ongoing investigation. 113 being boys who are attracted by the promise of religious studies ( talibé) and who end up being trafficked by their religious instructors.20 Girls tend to be engaged in commercial sexual exploitation in the Bijagós, on the mainland, and in Senegal.21 To counteract these trends, in 2016, the Government approved a Code of Conduct against sexual exploitation in tourism and the National Committee for Prevention and Combat of Human Trafficking developed proposals to support the reintegration of repatriated talibé children.22 Regional Integration 23. On the other hand, Guinea-Bissau’s integration in regional organizations had a significant impact on the country’s recent developments. In particular, membership in ECOWAS has helped generate a certain degree of stability, including more recently through the deployment of the ECOWAS Security Mission in Guinea-Bissau (ECOMIB) to support the transition back to constitutional order. 6.1.5 Security Sector Reform 24. Despite a series of attempts at SSR over the years, there has not been a successful and comprehensive process that addresses the efficiency, effectiveness, governance and accountability of the security forces. Successive attempts to establish a viable pension system have either met with military resistance to their implementation or failed to harness enough resources for successful completion. Piecemeal approaches to reform fostered by external partners, without strategic and operational coordination, have also contributed to these failures. Portugal, Brazil, Angola, the United Kingdom, the United States, Spain and France have all engaged with this issue to different degrees over time. Most of the external support provided so far has focused on policy development, assistance with legislative frameworks, training and capacity building, and some infrastructure rehabilitation. 25. Multilateral institutions have also provided substantial support but the different phases of assistance have not yet had any transformational impact. Among multilateral partners, the EU, UN, African Union, ECOWAS and the Community of Portuguese Speaking Countries have played an active role in terms of political engagement and enhancing the technical support provided to the defense and security forces. In December 2007, Guinea-Bissau was placed on the agenda of the UN Peacebuilding Commission, to help raise international attention to the political and economic fragility, as well as to the volatile security situation of the country. There have been three broad phases of SSR engagement over a 12-year period. Inception (2006–08) 26. The 2006 National Strategy for Reorganization and Modernization of defense and security forces laid the foundation for SSR in Guinea-Bissau.23 The strategy provided a diagnosis of the defense and security sector, contextualizing the rationale for SSR and establishing its objectives and the options for implementation. These included: - Resize the sector on the basis of the country's needs and real economic capacity (i.e. a security sector fit for purpose) - Modernize the security sector pursuant to the mission assigned to it by the government 20 https://www.unicef.org/wcaro/english/Child_Trafficking_Study_Guinea-Bissau_June_2010.pdf 21 https://www.state.gov/j/tip/rls/tiprpt/countries/2015/243448.htm 22 https://www.state.gov/j/tip/rls/tiprpt/countries/2015/243448.htm 23 Republic of Guinea-Bissau, “Interministerial Committee for the Reorganisation and Modernisation of the Defence and Security Sector, Reorganisation and Modernisation of the Defence and Security Sector- Strategy Paper“, October 2006. 114 - Clarify the status of the Combatentes da Liberdade da Pátria (war veterans) and restore their dignity - Reinforce the sector's participation in the consolidation of sub-regional security - Reinforce the judiciary and make it more efficient - Mobilize national and international resources - Involve civil society and the population at large in implementing the reform. 27. The strategy to resize the security sector on the basis of the country’s needs and real economic capacity comprised: (i) setting a new mission for the armed forces, (ii) resizing it according to objective criteria; and (iii) a comprehensive review taking into account vital national interests, the preventive and dissuasive role of defense, economic imperatives, geostrategic information, and the development vision. The strategy established a five-year process of gradual reductions to reach a total of 3,440 personnel divided among the army (50 percent of the total, i.e. 1,720 personnel), navy (35 percent, i.e. 1,204 personnel), and the air force (15 percent, i.e. 516 personnel). 28. The strategy went beyond mere size reduction; it also encompassed the need to modernize the armed forces through better equipment, and service and living conditions. The retirement of excess personnel, or personnel reaching the end of their careers, was to be accompanied by improved retirement conditions and economic and social integration. 29. The strategy called for a sustainable solution to the case of the Liberation War ex-combatants who have regularly contested the existing pension scheme. The strategy acknowledged the need to establish a systematic and reliable database of ex-combatants, as their number was increasing rather than naturally diminishing. Part of the problem was related to the interpretation of the criteria for determining who was an ex-combatant. Estimates in the strategy were that there were between 7,000 and 10,000 in 2006, more than those at the end of the war. To resolve the situation, a new census was required, and there was a need to revise and update the legislation on ex-combatants to remove ambiguities, bring their prerogatives in line with the current economic and social context, clarify the rights of their relatives, and distinguish between Liberation War ex-combatants and retirees from the current armed forces.24 In short, the strategy required a new pension framework. 30. The strategy was accompanied by a three-year implementation plan, overseen by an inter- ministerial committee chaired by the prime minister, a steering committee, and a permanent secretariat. Yet this organizational framework for implementation,25 monitoring, and evaluation was insufficient to advance the reform. Progress was stifled by a combination of limited financial resources and heightened resistance to change within the security sector, which culminated in the assassination of the Army Chief of Staff, General Tagme Na Waie, and President Nino Vieira in March 2009. Impetus (2008–12) 31. In 2008, the EU started a program of support to SSR, drawing on the 2006 strategy. EU assistance led to the development of a legal framework,26 including a Police Service Code of Conduct. The 24 There was also a wide disparity amongst different grades existed, especially between former high-ranking officers. 25 The United Kingdom was the main partner during this period, rendering significant support to this initiative, inspired by its positive experience with assisting SSR in Sierra Leone. 26 Law of Military Service, Law of Military Condition, Law setting-out the Framework for the Organization of the Armed Forces, Law of National Defence and of the Armed Forces, Military Statute of the Armed Forces, Strategic Concept of National Defence, Military Strategic Concept, Armed Forces Missions, Organic Law of the General Staff of the Armed Forces, Organic Law of the Army, Organic Law of the Navy, Organic Law of the Air Force, Code of Military Justice, Bylaw of Military Discipline, Bylaw of Military Insignias, Military Programming Law. 115 legislation was approved in 2010, but the program came to a halt during that year in the wake of renewed political instability. In the context of this program—which also aimed to support the demobilization, reintegration and pension agenda—a biometric census of the armed forces, the security forces, and the ex- combatants was also conducted. This led to some improvements, including the introduction of salary payments through the banking system to prevent double disbursements and payment to “ghost” members. 32. ECOWAS attempted to re-energize the SSR process in April 2009. It hosted a roundtable on the Restructuring and Modernization of the Defense and Security Sector, with the central objective of identifying constraints to the speedy implementation of the SSR program. The rehabilitation of barracks and the implementation of the pension fund for the armed forces were at the heart of the discussions. A new element was the proposal for a communication plan for greater citizen engagement, given that the public were not consulted during the drafting of the 2006 strategy. Accordingly, in 2010, a roadmap with priorities (including demobilization and reintegration, and the pension fund) was adopted by the Military Component of the ECOWAS-CPLP Working Group to facilitate resource mobilization and assist the production of guidelines to expedite the dispatch of technical teams to support the take-off of the activities. Stagnation (2012 Onwards) 33. The central problem of pensions and personnel management remained unresolved in the aftermath of the coup in 2012 until elections were held later in 2014. ECOWAS played a proactive role in this phase of SSR. It helped the government develop a draft law for the pension fund and a gratuity scheme to address lingering issues in the armed forces (including aging personnel, poor capacity, bloated ranks, and irregular and illegal recruitment since 1998). The program was intended to finance barracks rehabilitation and the proposed pension fund, demobilization, and recruitment, at an estimated cost of US$63 million.27 According to the Special Pension Fund, members of the defense and security forces were to be demobilized annually over a five-year period (2015-2019).28 The Ministry of Defence estimated that 500 would be demobilized per year, totaling circa EUR 24 million. 29 However, post-electoral instability hindered any progress. 34. The pensions and gratuity decree law was approved in 2015 by the government, as a way to provide momentum for the restructuring and potential downsizing of the army . Around its approval, the law galvanized intense debate about the criteria, mechanisms, and proportionality of payments of retirement pensions and lump sums. Beyond that, discussions extended to issues pertaining to the historical legitimacy of the military, the ethical fabric of public institutions, the values underlying civil-military relations, and the importance of SSR to rebuild trust and sustain transformation. 30 35. Ultimately, the different proposals for the pension fund31 were considered to be misaligned with the reality of the country, and the gratuities were contrasted with the extreme poverty of most of the population. The proposed incentives included a “gratuity” estimated at some US$58 million, in addition to pension annuities of US$25 million. This fact contributed to the further alienation of important constituencies from the common historical narrative of the Liberation War, which the military claimed to 27 ECOWAS was also able to rehabilitate 4 of the 12-selected barracks and identified 2,282 military, police, and national guard personnel (about half of the total number of the armed forces) for placement within the pension scheme. 28 Republic of Guinea-Bissau, Decree-Law on the Special Pension Fund, February 2015. 29 Meeting with MoD Officials, Bissau May 2018. 30 Ibid. 31 Habibu Bappah (2017). “Why Peace Fails in Guinea-Bissau.” http://library.fes.de/pdf-files/bueros/fes-pscc/14166.pdf 116 embody. One of the strategic issues remains the clarification of the status of veteran and freedom fighter, as a pre-condition for their “dignified and sustainable” demobilization or retirement.32 Figure 6.6: Distribution of Personnel by Sector a. Distribution of Public Servants, by Sector Education, 37% Internal Security, 17% Health, 9% Other, 3% Defense, 18% Administration, 12% Source: Ministry of Public Administration and MEF. b. Distribution of the Armed Forces c. Distribution of Interior Forces General staff of the armed forces Navy Army Air Force Others Military Tribunal Source: Ministry of Economy and Finance Source: Ministry of Economy and Finance 6.2. Evolution of Guinea-Bissau’s Security Sector 6.2.1. Legal and Policy Framework 36. Guinea-Bissau’s chronic instability at the political and military level, is largely associated with the precarious state of its defense, security and justice institutions. In recent years, the legal framework of the armed forces has been partially updated, but implementation is lagging. The various attempts at SSR described above have resulted in new policy and legal frameworks guiding the security forces. In 2009, the internal security forces were restructured from eight to four agencies, under the management and oversight of the Ministry of Interior and the Ministry of Justice. The intelligence services were also affected by modifications in the legislation, with the State Information Services being shifted from the president’s to the prime minister’s authority. However, although there is a parliamentary defense and security commission to oversee the armed forces and the internal security forces, the legislation did not set up an external oversight body for the intelligence services. Annex VIII shows the organigram of the Ministry of Interior. 32 Ibid. 117 37. Law 11/2011 was passed on the organization of the armed forces, orientating the framework and institutional structure of the defense apparatus (Figures 6.6.a–c). Other revised legislation includes the compulsory military service law, the military statute, the national defense and armed forces law, and the statute of the ex-combatants of the Liberation War. However, the legislation has not been consistently applied. For example, delays in the implementation of the compulsory military service law due to lack of material conditions have meant that the first recruitment of soldiers since 1992 only took place in December 2017. However, this was a one-off initiative lacking any plan for future incorporation, given the absence of predictability of financial inflows to sustain the process. 33 38. Since the 2010/2011 changes, the operationalization of the then approved legal frameworks have now demonstrated a need for revision to conform with current security and justice challenges , and to incorporate the lessons learned in the last few years. In particular, the Ministry of Defense and the Armed Forces have advanced on a series of revisions of legal and policy frameworks, including institutional structure in the recent times; however, these are yet to be legally approved. Annex VIII show the organigrams of the Armed Forces and the Ministry of Defense, respectively. 39. Overall and across the public sector, there are several legislations, stemming from the colonial era, which need to be updated to reflect the current situation, including regional and international processes. A prioritization of the latter is central for core development of each sector, representing a foundation stone for planning, management, and internal and external monitoring, as well as oversight of policy and budget allocation and implementation. Inconsistencies between different legal documents make it difficult to obtain a full and reliable diagram of the security and justice sector institutional architecture. For the security sector, in particular, budgetary allocations will remain unpredictable if, for example, ministerial structures continue to depend on the discretion of each incumbent Government. Annex VIII shows the institutional framework of the security and justice sector. 6.2.2. Overview of Defense and Internal Security Institutions 40. Guinea-Bissau’s security sector accounts for 8,246 of the total public sector workforce of 23,461 (35 percent). 34 The armed forces are overseen by the Ministry of Defense, the Ministry of Interior oversees the security forces (including the Police, Immigration and Fire Services), and the Ministry of Justice oversees the judiciary institutions. The armed forces number 4,342 personnel: 1,944 army soldiers (44.8 percent of the total); 1,738 people at the General Staff headquarters (40 percent); 310 in the navy; and 116 in the air force. The remaining 70 are distributed among support/administrative functions. The security services consist of 3,904 men and women, 79.8 percent of which are in the police (3,116 officers); this figure also includes an unknown number in the National Guard, which has military status. The police force headcount gives a ratio of 598 inhabitants per police officer, which compares with the UN benchmark of 450 inhabitants per police officer. 35 The rest are distributed among the Immigration Service (248 officers), 33 Meeting with MoD and General Staff Officials, Bissau May 2018. 34 All the personnel information regarding the security sector are extracted from the SIGRHAP database of the Ministry of Economy and Finance and the Quadro de Pessoal database of the Ministry of Public Administration 35 The United Nations recommends on average one police officer for every 450 citizens. However, a number of caveats related to context need to be taken into consideration in deciding on the ratio, including the country's geography and road network; the country's policing model; police mobility and communications capacity; the resilience/contribution of informal policing mechanisms; the country's crime rate; and the fiscal space to recruit and sustain a professional police service. 118 the National Civil Protection and Fire Service, a General Secretariat of 188 employees, and other administration services with 125 staff. Box 6.2: International Comparisons of the Ratio of Supervisors to Subordinates In modern armed forces, the ratio of senior officers to subordinates is ordinarily within the range of 1 major general (or lieutenant general) to every 10,000 to 15,000 soldiers . There are wide variations across countries, reflecting the fact that the appointment of higher ranks can sometimes be made for political reasons without affecting the adequacy of the institution for its intended purpose. In Switzerland, for example, active armed forces number 20,950 while the reserve forces number 14,427. There is only one post of general, and it is vacant in peacetime (Military Balance 2017, International Institute for Strategic Studies). In contrast, the Madagascan army has 222 generals (less than 1% of high ranking officers) and 2,900 field officers (10%) in an army of 30,000. Source : NATO and Etat-major des Forces Armées Malgaches (2018). Defense: Structure and demographics 41. The Ministry of Defense payroll taken in February 2018 from the databases in the MEF and MPA, identified 1,099 individuals (25 percent of the total armed forces) who do not have complete salary records. They have an ID number and are assigned a salary line, but do not have a rank, and nor are they assigned to a position. Of these, 575 (52.3 percent) belong to the army, 432 (39.3 percent) to the General Staff, 82 to the navy, and 6 to the air force while 4 could not be attached to any department. The remaining 75 percent gives a large enough sample of the armed forces population to provide a high level of confidence in the statistics for the Armed Forces. 42. The hierarchical structure of the armed forces displays a marked imbalance, indicating structural and public finance inefficiencies (Figures 6.7–6.10). Based on the verified records of 2,966 personnel, the armed forces display an inverted pyramid structure, denoting a failure to plan human resources, hinting at periodic promotions and showing minimal or very limited new recruitment. The sample had 46 high-ranking officers, 923 field officers (31.1 percent) and 1,031 subalterns (35 percent) for an army of 414 private soldiers (14 percent), well out of line with international norms (Box 6.2). The composition alone signals that the structure of the armed forces has not been a government priority, or that internal difficulties have hampered the implementation of the nation’s security policies. Such top-heaviness36 is a sign of a sector that is not highly functional or representative of a country’s needs. 43. It is difficult to provide a definitive count of the number of women in the armed forces as the gender field in the data sample was not updated for 2,878 people .37 Only 135 workers are identified as female, of whom 14 are military personnel. Their ranks include 1 brigadier general, 3 field officers, and 8 first lieutenants. The remaining female personnel work in the military tribunal and (the majority) in support services as assistants, janitors, and cooks. The average age of the 135 identified women is 54, with the youngest being 35. 44. The average monthly salary in the armed forces amounts to CFAF 158,368 (about US$280), with 61 percent of the wage bill going to the top third of the ranks, while the lower ranks (33 percent of the headcount in the sample) receive 9 percent of the compensation. The highest base military salary is 36 The number of generals and admirals has increased from 1 in 1991 to 37 in 2017, UNIOGBIS and ISS, ibid 37 The database was accessed as of February 21, 2018. 119 for a general and amounts to CFAF 560,000 per month (US$950), putting them in the same salary bracket as the prime minister, while a colonel’s salary of CFAF 370,000 (US$630), falls into the same bracket as a government minister. The lowest salary in the army is the same as for the rest of the public services, at CFAF 26,000. After including a 20 percent allowance, as for the rest of the civil service, the highest total compensation is CFAF 672,000 (US$1,100). The lowest base salary is CFAF 24,000, or CFAF 28,800 including the allowance (US$50). Figure 6.7: Distribution of Army Ranks Figure 6.8: Army Payroll, by Rank Source: MEF and World Bank Staff estimates. Source: MEF and World Bank Staff estimates. Figure 6.9: Army Personnel, by Age Figure 6.10: Reverse Cumulative Distribution of Army Personnel, by Age (2018) 100% 98% 16% 15% 81% 14% 12% 12% 11% 67% 40% will be over 60 9% years old in 5 years 5% Average Age 3% 13% 2% 48.7 1% 0% 1% 0% 1% Source: MEF and World Bank Staff estimates. Source: MEF and World Bank Staff estimates. 45. Salary distributions show no significant difference between the armed forces and the rest of the public service. The salary spread for the armed forces is 22, indicating that the highest salary is 22 times the lowest salary. The spread for the general civil service is 27, or 22, if the president’s salary is excluded. Figure 6.8 shows how the army payroll is distributed among the categories of ranks for the sample of 2,966 military staff. 46. The absence of a sector and human resources policy has led to the current average age of the armed forces reaching 48.7 years old (Figure 6.9). The average age of the sample of 2,966 is 50.8 years old. The mode is 38 years old, confirming that the contingent is mature. Almost 25 percent of the sample has currently reached the official retirement age of 60 (Figure 6.10). An additional 15 percent of the sample population will reach retirement age in five years, increasing the total share of those over retirement age to 120 40 percent of the sample if there is no change in policy. When the 1,099 incomplete records are included, 20 percent of armed forces personnel are already over retirement age, with the figure rising to 32 percent in five year time. Figure 6.11: Distribution of Police Officers, by Rank Figure 6.12: Police Payroll, by Rank Source: MEF and World Bank Staff estimates Source: MEF and World Bank Staff estimates. Figure 6.13: Police, by Age Figure 6.14: Reverse Cumulative Distribution of Police by Age (2018) 19% 100% 94% 16% 82% 12% 63% 12% 10% 47% 9% will be over 24% 60 years 5% 5% old in 5 Average Age 45.3 5%years 1% Source: MEF and World Bank Staff estimates. Source: MEF and World Bank Staff estimates. Internal Security: Structure and demographics 47. For the security forces, the hierarchy is closer to a pyramidal structure, but with a significant body of commanding officers (equivalent to the armed forces’ “subalterns” and “field officers”). See Figures 6.11 and 6.12. World Bank staff analyzed a dataset consisting of 3,862 uniformed police officers with operating duties, 98.9 percent of the whole Ministry of Interior, excluding 42 in support and managerial roles. In the sample, there were 16 high-ranking officers, 587 “field officers”, and 751 “subalterns” with the non-commissioned officers accounting for 1,903 of the payroll. 48. Due to incomplete records in the Ministry of Interior database extract, it is difficult to provide an estimate of the gender distribution of the police force . A few observations can be made from the available data: 218 personnel are identified as female, comprising 21 field officers from captain to lieutenant colonel, 25 subalterns, 95 non-commissioned officers, and 73 constables. The remaining four female 121 employees work as an adviser (high-ranking manager), head of unit, senior technician or non- administrative staff. 49. Salaries in the public order force vary considerably between the lowest and highest ranks, and are also supplemented by 20 percent allowances on top of base salaries . The lowest base salary is CFAF 23,667 per month for “new” officers and the highest is CFAF 400,000 per month, for the highest - ranking officer (Brigadeiro Geral). Adding in the 20 percent allowances, the total compensation becomes respectively CFAF 28,400 to CFAF 480,000, resulting in a salary range spread of 15.9. The spread is narrower than in the military, but it remains wide. Figures 6.11 and 6.12 show that the 15.6 percent of the total strength who are high-ranking and field officers in the police receive 42 percent of the total compensation. About 65 percent are either non-commissioned officers or lower ranks and they account for 31 percent of the total wage bill. 50. The average age in the police force is 45.3 years old for the 3,116 staff sampled (Figure 6.13). Currently 14 percent (441 personnel) are above the retirement age of 60, and in 5 years, another 10 percent will have reached retirement age, 24 percent of the current personnel (Figure 6.14). 6.3. Affordability and Sustainability 51. The section assesses the affordability of the Government’s security spending and the ability of the sector to be affordable over time, i.e. its sustainability . Extensive analysis around the world on such questions of security affordability38 has found no hard-and-fast answers. These are questions that need to be considered by each finance ministry together with their security counterparts on a country-by- country basis. Experience suggests, however, that there are four tests to apply in considering the affordability question: i) Security spending within the parameters of national revenues : in some countries, such as Afghanistan and Somalia,39 expenditures far outweigh total revenues and therefore are not affordable without significant external support, which in turn is not sustainable. ii) International benchmarks: there are no international standards or prescriptions for military spending; for policing there is a general UN standard of 1 police officer per 500 inhabitants. However, it can be useful to compare spending levels with other countries of similar economic size and structure, facing similar sets of challenges or within the neighboring region, also known as benchmarking.40 iii) Fit-for-purpose: a related question whether the resources being used for a sector are being used effectively. If finite public resources are being wasted then that is not sustainable over time. The question for finance ministries is whether resources are being aligned with policy objectives, which in turn meet human and national security needs. This test is particularly critical in fragile states given the strong correlation between security and development: the safer a population, the more likely an economy will grow. There is a widespread consensus that the countries most affected by conflict, instability, and displacement have fared the worst on poverty reduction. Although the 38 Harborne et al, ibid. 39 Harborne et al, ibid. 40 It should be noted that due to the availability of data, benchmarking for military spending is much easier than that for police and criminal justice institutions. 122 World Bank is not able to advise on defense policy,41 it can take any changes in government policy in the security sector, such as greater investment in equipment, training or recruitment of women, and cost such policies over the medium-term. iv) Trade-offs: finally, with finite public resources, spending on security will offset spending on other sectors, a country needs to calibrate its security expenditure against the demands of those other sectors. Insights into the budgetary trade-offs between security spending, on the one hand, and Guinea-Bissau’s key development priorities (including infrastructure, education and health), on the other, is a critical step forward toward ensuring that resources are channeled to their most productive uses. Figure 6.15: Guinea-Bissau Expenditure by Sector (US$) 120.0 100.0 79.1 2015 2016 2017 80.0 60.0 40.0 5.0 18.9 20.6 20.0 7.3 6.2 0.3 1.3 0.3 1.3 1.4 0.2 0.2 0.0 Source: MEF. 6.3.1. Security Spending in Guinea-Bissau 52. Government spending on the security sector averaged 2.5 percent of GDP (or 16 percent of total government expenditure) over 2015–17. If the tumultuous period before 2015 is included, average security spending would increase to 3 percent of GDP over 2010–17. As a share of the national budget, total government security spending slightly decreased to 15.1 percent in 2017 and averaged 13.9 percent between 2014 and 2017 (Figure 6.15). Is this level of spending affordable? To help answer this question, this section: (i) describes the costs of the security sector; (ii) benchmarks security spending in Guinea-Bissau against other fragile countries and regional peers; and (iii) assesses the affordability and sustainability of the security spending against revenue and growth projections. 6.3.2. Composition of Security Spending 53. The main cost drivers of spending have been personnel and related operations expenditure. 42 Since 2015, personnel expenses have, on average, made up 68.5 percent of total government security spending. During the same period, operational expenditure has accounted for 30.5 percent of total 41 According to World Bank guidelines set out in 1991, staff do not have the competence or mandate to advise client governments on policy matters regarding military spending. 42 The Minister of State of Economy and Finance and the Chief of Defense Staff of the Armed Forces explained to the World Bank staff that their main objectives were to avert any frustration and resentment in the armed forces, and to ensure that the forces would stay tranquil. As far as the interior security services are concerned, given the scope agreed by the Government for this PER, the WB staff has not carried out a more granular analysis yet on the adequacy of the nature and volume of the expenses they incur to accomplish their duties and mission statement. 123 spending. Capital expenditure on security equipment and infrastructure is near to nil.43 Total security sector personnel expenses have averaged CFAF 11.66 billion or US$19.76 million over 2015 –17 (for both the Ministry of Defense and Ministry of Interior). Tables 6.1 and 6.2 break down the official categorization of spending by these two ministries from 2014 to 2017. Personnel expenses averaged CFAF 6.65 billion (US$11.27 million or 65 percent), over 2015–17 for the Ministry of Defense, while they averaged CFAF 5.01 billion (US$8.5 million or 74 percent) for the Ministry of Interior. Table 6.1: Decomposition of Defense Spending (CFAF millions) Expenditure 2014 2015 2016 2017 1. Personnel cost 6,479 6,256 6,499 7,185 3. Purchase of goods and services 3,970 3,911 2,765 3,869 4 Transfers 72 178 65 115 5 Other current expenditure 76 7 Investment expenditure (actual) - 28 5.4 - 7 Investment expenditure (tax exempt) - - - - 10 Public investment (external grants) 3,007 - - - 11 Public investment (external loans) - - - - Total 13,604 10,373 9,336 11,169 Source: MEF. Table 6.2: Decomposition of Internal Security Spending (CFAF millions) Expenditure 2014 2015 2016 2017 2. Personnel cost 5,859 5,515 4,536 4,992 3. Purchase of goods and services 1,001 1,433 1,377 2,396 5 Other current expenditure 50 7 Investment expenditure (actual) - 5 - 7 Investment expenditure (tax exempt) - 1.6 10 Public investment (external grants) 47.6 57 32.7 14.6 11 Public investment (external loans) - - - - Total 6,957 7,011 5,946 7,402 Source: MEF. 54. Operations expenses include both “purchase of goods and services” and “other current expenditures”, totaling CFAF 5.26 billion or US$8.9 million on average over 2015 –17. The Ministry of Defense has spent CFAF 3.51 billion (or US$5.96 million – 34 percent) while Ministry of Interior spent CFAF 1.74 billion (or US$2.94 million – 26 percent) on these categories. “Other current expenditure” has been negligible during the period. Capital expenses on equipment and infrastructure are almost nil for both ministries. The Ministry of Interior received some external grants of CFAF 35 million or US$59,000 reflecting the freeze in financial support for the armed forces among the international community.44 43 Inflation rates are integrated in the estimates and are based on IMF Article IV report December 2017. 44 With the possible exception of China. 124 Figure 6.16: Armed Forces Personnel: Regional and Figure 6.17: Relative Size of Armed Forces: Peer Countries Regional and Peer Countries % of the total labor force 60,000 2.50 2012 2013 2014 2015 50,000 2.00 40,000 1.50 30,000 1.00 20,000 10,000 0.50 0 0.00 2012 2013 2014 2015 2016 Source: World Development Indicators (2018). Source: SIPRI (2017), World Development Indicators (2018). Figure 6.18: Total Military Expenditure: Regional and Figure 6.19: Relative Military Expenditure: Peer Countries, constant 2015 US$ Regional and Peer Countries, % of GDP 4.0% 600 2012 2013 2014 2015 2012 2013 2014 2015 3.5% 500 3.0% 400 2.5% 300 2.0% 1.5% 200 1.0% 100 0.5% 0 0.0% Source: SIPRI (2017). Source: SIPRI (2017), World Development Indicators (2018). Figure 6.20: Relative Military Expenditure, % of Government Spending 20% 2012 2013 2014 2015 15% 10% 5% 0% Source: SIPRI (2017), World Development Indicators (2018). 125 55. While this review does not constitute a full assessment of operations expenses, consultations indicated that some of this spending was on fuel and food rations but it is assumed that operations expenses also comprise other personnel expenses. This would explain in part the findings in Chapter 2 on the public sector wage bill, which highlights how official statistics do not provide a full picture of total government employee compensation. Official figures significantly underestimate the wage bill; other compensation items not included in official figures are estimated to reach an average of 4 percent of GDP. Due to lack of accurate and more disaggregated data, this analysis has not calculated how much of the “unofficial compensation” could be apportioned to the security sector. This component of security spending comprises also personal allowances, emoluments, and other benefits associated with the military class. 56. The imbalance of personnel, operations, and capital expenses indicates a sector that is not functioning as effectively as it could. In general, the composition of security spending in similar countries falls within the following ranges: 54–60 percent for personnel expenses, 20–27 percent for operational expense, and 10 percent for capital investments.45 The security sector budget in Guinea-Bissau suggests that it is not designed to meet current security threats. The section that follows discusses how spending trends in Guinea-Bissau compare with neighboring countries. 6.3.3. International Benchmarking 57. This section compares key security sector trends for Guinea-Bissau with its structural (Burundi, Comoros, and Sierra Leone) and regional peers (Benin, Cote d’Ivoire, Ghana, Guinea, Liberia, and Senegal). The analysis focuses primarily on international benchmarking, available only for the armed forces. It covers (i) the size of the armed forces; (ii) the percentage of the military in the labor force; and (iii) military expenditure in absolute terms, as well as in terms shares of GDP and government spending. Data sources include: (i) The Stockholm International Peace Research Institute (SIPRI) database on military expenditures; (ii) International Institute for Strategic Studies (IISS) “Military Balance 2017”; (iii) World Development Indicators (2018); and (iv) World Bank staff analysis. 58. Compared to its neighbors and peers, Guinea-Bissau’s military is larger only than Liberia’s in absolute numbers of armed force personnel (Figure 6.16). But as a share of the labor force, Guinea- Bissau’s army is second only to Burundi, which highlights the significant role of the security sector in providing livelihoods in a constrained job market and as a compensation network for those in uniform and their families (Figure 6.17). While Guinea-Bissau is below the average for countries in the World Bank’s list of fragile and conflict-affected states (FCS), the figure is still high for a country that does not face significant external threats or active internal conflict. Measured in constant US dollars and based on official data only, Guinea-Bissau appears to be one of the lowest-spending countries for its armed forces compared to neighbors and peers, at levels similar to Liberia’s (Figure 6.18). Military spending accounted for a very high share of the national budget in 2012 (Figure 6.20), at 20 percent, partly due to increased military expenditure during the 2012 military coup. There has since been a dramatic decrease, to 6 percent in 2015. 6.3.4. Affordability Scenarios 59. Affordability analysis compares projections of domestic revenues with estimates of future costs. This analysis considers two scenarios: (i) the baseline or status quo scenario, which serves as the 45 Military Spending in Developing Countries: How Much is too Much? Brian Macdonald, 1997 126 reference for future scenario iterations; and (ii) a retirement savings scenario that takes into account the findings above concerning ageing military personnel and potential retirees. Figure 6.21: Guinea-Bissau Security Figure 6.21: Projected Domestic Revenue versus Expenditure: Baseline Security Sector Expenditure 8.6 8.8 8.8 8.8 8.8 261.4 8.2 8.3 7.9 8.0 236.9 213.2 million $ Million $ 191.0 155.4 143.7 142.9 20.8 21.3 21.8 22.4 22.4 22.4 22.4 20.0 20.3 7.9 8.0 8.2 5.9 7.8 6.3 6.2 20.0 20.3 20.8 15.2 15.6 16.0 15.9 2015 2016 2017 2018 2019 2020 2021 2022 2023 2015 2016 2017 2018 2019 2020 2021 Personnel Expenses Operating Expenses Personnel Expenses Operating Expenses Capital Expenses Capital Expenses Domestic Revenue Low Source: IMF, MEF. Source: IMF, MEF. The Baseline – Security Structure Remains Unchanged 60. The baseline scenario (no retirement and no recruitment) takes the current security sector configuration and projects it into the future, adjusting for inflation. This projection only accounts for uniformed personnel: (i) 4,065 personnel in the armed forces; and (ii) 3,862 officers for the police force. The analysis considers the purchase of goods and services and other current expenditure as operations expenses. 59. If the total headcount of 7,927 remains unchanged, with the same levels of personnel, capital, and operating expenditures, security spending is marginally affordable. While security spending would average about US$30 million (Figure 6.21), domestic revenues are projected to increase to US$260 million within three years (Figure 6.22). Assuming the projected domestic revenues of $191 million in 2018 are realized, the security sector expenditure is likely be to be marginally affordable starting from 2020. Under this hypothesis, the ratio between security sector spending and projected domestic revenues would be 14 percent, excluding non-uniformed personnel. Table 6.3: Number of Retirees by Institution and Year Year Police Armed Forces Total 2018 559 755 1,314 2019 81 60 141 2020 73 93 166 2021 75 61 136 2022 96 119 215 2023 74 107 181 TOTAL 958 1,195 2,153 Source: MEF, MPA. 127 Figure 6.23: Cumulative Difference in Personnel and Operations Expenses: Baseline and Retirement Scenarios (2018–23) 25.0 20.0 Personnel Expenses: Baseline ($46.3 M in "savings) 15.0 Million $ Personnel Expenses: Retirement 10.0 Operations Expenses: Baseline ($15.5 M in "savings) 5.0 Operations Expenses: Retirement 0.0 2015 2016 2017 2018 2019 2020 2021 2022 2023 Source: MEF, World Bank Staff estimates. The Retirement Savings Scenario 60. The retirement savings scenario assumes security personnel will retire in accordance with the existing labor law (at the age of 60, no exceptions), and with no additional recruitment taking place. These calculations are based on the cohort of uniformed personnel in the sector (4,065 military and 3,862 policemen). Once retirement begins, the military headcount would decrease to 3,301 in 2018, and to 2,870 by 2023. Meanwhile, the police force headcount would decrease to 3,303 in 2018 and 2,904 by 2023. Table 6.3 below shows the numbers retiring for each institution over the next five years. An effective retirement policy from 2018 onwards would thus immediately decrease total headcount by 1,314, which represents 16.5 percent of the uniformed personnel. In five years, a total of 2,153 of uniformed personnel (27 percent) would have retired. 61. The fiscal implications could potentially save the government a cumulative total of US$28.4 million in military personnel costs and $17.9 million in police personnel costs over the next 5 years (Figure 6.23). The wage bill for those military personnel due to retire in 2018 represents a payroll amount of CFAF 2.26 billion (US$3.83 million) in 2018; for the police it amounts to CFAF 1.31 billion or US$2.22 million. The combined savings from the government wage bill would total US$46.3 million over five years. In addition, Guinea-Bissau could potentially save an additional US$15.5 million in goods and services expenditure, in other words those operational expenses directly related to personnel. 6.4. Recommendations 59. The military and security forces have been at the heart of Guinea-Bissau’s development. As envisaged by the 2016 Conakry Accord, robust security sector reform is necessary to ensure that security provision favors development and poverty reduction. As the history of repeated attempts at reform indicates, this is a challenging area of government policy. Despite international partner renewed efforts to support reforms in 2015 the process stifled up to this day given political instability. Regardless of its centrality to the short and long-term stability, security and development of Guinea-Bissau, the issue 128 appears to have disappeared from the agenda. Therefore, the overall challenge remains unaddressed, and in particular the most immediate subject of demobilization, pensions and retirement follows on, hindered by three major constraints: (i) promises on pensions and gratuities made to veterans and potential retirees and their families that have been resisted; (ii) the cost of overaged security personnel; and (iii) the need for equity in the provision of entitlements that are commensurate to the generally poor living standards of the remainder of population.46 60. This reform deadlock points to a number of overarching social and political challenges but even from this brief assessment it is clear that only by strengthening the security-development nexus will these challenges be resolved. Figure 6.2 is a telling illustration of the “stop-go equilibrium” in which Guinea Bissau is trapped, unable to achieve greater poverty reduction and shared prosperity due to the onset of political crisis, much of it associated with the military. Over the last 15 years, some key partners have tried to assist respective governments in their SSR efforts; these have ultimately not succeeded. While not providing any magic solutions, one of the key findings of this review is that development actors also have to engage in supporting state reform, and that such process, whilst unsettled, requires joint preparation ahead of the upcoming November 2018 elections, if policy and implementation benefits are to be yielded in the post-electoral period. More comprehensive work will have to be undertaken, but even the outline provided here offers future pointers for the reform process: i) Affordability: In 2015, the government presented a US$270 million SSR plan to donors in Brussels.47 This chapter suggests that it has sufficient domestic resources to leverage international support to finance the necessary reforms, particularly in the area of pensions. For example, the work presented here shows retirement could bring potential savings of up to $46.3 million over five years that could be devoted to such efforts. ii) Financing instruments: further expenditure and PFM analytics could provide a clearer connection between development financing, such as budget support, and the SSR process. Policy actions can be linked to such support, as in The Gambia and Central African Republic. In turn, experience from elsewhere can be applied to the Guinea-Bissau context such as use of pooled funding arrangements, trust funds, and recruiting a fiduciary agent to leverage finance for such a reform process. iii) Fit-for-purpose: while not comprehensive, the context analysis above outlines a number of challenges Guinea-Bissau faces, ranging from drug to human trafficking. SSR could include a rebalancing of expenditure towards greater investment, such as the financing of a coastguard, as well as the training needed to develop a modern set of professional and accountable security institutions. Recent evidence48 suggests that forces are motivated by professional employment and remuneration through salaries. For this reason, more analysis will be needed to understand the full picture of government compensation and the potential role of gratuities as obstacles to an effective SSR that can turn the security sector into a public service benefiting all citizens. 46 Habibu Bappay, Why Peace Fails in Guinea-Bissau, Friedrich-Ebert-Stiftung, 2017 47 UNIOGBIS and ISS, https://issafrica.org/research/books-and-other-publications/relaunching-defence-and-security-sector-reforms-in- guinea-bissau 48 Holger Albrecht & Ferdinand Eibl, How to Keep Officers in Barracks: Causes, Agents, and Types of Military Coups; International Studies Quarterly, May 2018. 129 The government could therefore consider the following recommendations: 6.4.1. Fiscal Sustainability and Pensions for Security and Defense 62. While Guinea-Bissau’s security sector spending is high compared to peer countries in the region, it is forecast to fall as a share of projected revenues. This would allow the Government the necessary fiscal space to address the outstanding issue of pensions and the Liberation War ex-combatants. With about 32 percent of the military and 24 percent of internal security forces reaching retirement in five years, this remains a priority fiscal and sector policy issue. Even though projections show fiscal revenues outweigh security sector spending until 2021, the government should carry out an actuarial study to assess the impact of personnel retirement on its ability to provide pensions and explore options for downsizing and new recruitment to rebalance the rank structure. 6.4.2. Policy-Based Budgeting for Security 61. The economic classification of spending in Guinea-Bissau’s armed forces shows that the military and police spend most of their resources on personnel costs. This is consistent with many “light infantry” models in Africa; but the lack of capital expenditure reveals little focus on long-term planning and modeling of a force structure that is fit for purpose. Consequently, the government should consider aligning its budget with a clear SSR strategy that takes into account projections of future needs against a coherent set of security objectives aimed at modernizing and professionalizing its institutions, including a greater emphasis on capital investment and training. 6.4.3. A Public Financial Management Review of Security 62. These efforts could be underpinned by a public financial management (PFM) review to shed light on the budget process, from planning to execution and audit . Such a process can be undertaken as a joint effort by the MEF, the security ministries and the World Bank. Given the priorities discussed with partners since 2012, such a PFM review should focus on payroll and personnel management along with strengthening systems for oversight, accountability and transparency of expenditures. 130 References Bappah, H. (2017). ‘Why Peace Fails in Guinea-Bissau.’ Available at http://library.fes.de/pdf-files/bueros/fes- pscc/14166.pdf Benson T and Shekar M (2006). Trends and Issues in Child Undernutrition. Chapter 8. In: Jamison DT, Feachem RG, Makgoba MW, et al., editors. Disease and Mortality in Sub-Saharan Africa. 2nd edition. Washington (DC). Brian S. M. (1997). ‘Military Spending in Developing Countries: how much is too much? Ottawa: Carleton University Press, 1997, 220 pp. CONFEMEN. (2012). ‘Program in Support of Education Systems of CONFEMEN Countries (Programme d’Appui aux Systems Educatifs des Pays de la CONFEMEN).’ Dakar, Senegal. Flynn, S., and Mario, P. (2014). ‘Prevention and management of government expenditure arrears.’ Technical Notes and Manuals 14/01. Washington, DC: International Monetary Fund. Gaspar, V., Jaramillo, and P. Wingender. 2016. ‘Tax capacity and growth: Is there a tipping point?’ IMF Working Paper 16/234, International Monetary Fund, Washington, D.C. Gottret P and Schieber G (2006). Health Financing Revisited: A Practitioner's Guide. Washington, DC: World Bank. Harborne, B., Dorotinsky, W., and Bisca, P. M. (2017). ‘Securing development: Public Finance and the Security Sector.’ Washington, DC: World Bank. Holger, A., and Ferdinand E. (2017). ‘How to keep officers in barracks: Causes, agents, and types of military coups. International Studies Quarterly, forthcoming. IMF. (2015) ‘Article IV consultation and request for a three-year arrangement under the ECF.’ IMF Country Report No. 15/194. Washington, DC: International Monetary Fund. IMF. (2017) ‘Guinea-Bissau: Article IV consultation and fourth review under the ECF arrangement, and financing assurances review.’ IMF Country Report No. 17/380. Washington, DC: International Monetary Fund. K Xu, DB Evans, K Kawabata, et al. (2003): Household catastrophic health expenditure: a multi-country analysis. The Lancet. Vol(362):111-117. Lienert, I. (2010). ‘Role of the legislature in budget processes.’ IMF Technical Notes and Manuals 10/04. Washington, DC: International Monetary Fund. Ministry of Economy and Finance (2015). ‘Guinea-Bissau: Brief overview of the main taxes collected by the Directorate-General of Taxes and Duties (DGCI).’ Ministry of National Education (2015). ‘Republic of Guinea-Bissau: Relatório da situação do sistema educativo (Report on the situation of the education system).’ August 2015. Ministry of National Education. (2015). ‘Republic of Guinea-Bissau: Education Management Information System (EMIS 2014-15).’ Bissau, Guinea-Bissau. MINSAP. (2017). ‘Plano Nacional de Desenvolvimento Sanitário 1997-2001.’ Ministério da Saúde Pública, Governo da República da Guiné-Bissau. Bissau, Guinea-Bissau. National Statistical Institute (INE). (2014). ‘The Guinea-Bissau Multiple Indicator Cluster Survey (MICS5).’ Bissau, Guinea-Bissau. Republic of Guinea-Bissau. (2006). ‘Interministerial Committee for the Reorganisation and Modernisation of the Defence and Security Sector, Reorganisation and Modernisation of the Defense and Security Sector-Strategy Paper“, October 2006. Republic of Guinea-Bissau. (2010). ‘Inquerito Ligeiro de Avaliacao da Pobreza (ILAP) 2010 household survey data.’ Bissau, Guinea-Bissau. Republic of Guinea-Bissau. (2014). ‘Programa de Estabilização e Desenvolvimento.’ Bissau, Guinea-Bissau. Republica da Guinea-Bissau (2008a). Plano Nacional de Desenvolvimento Sanitario. Ministerio da Saude Publica. Bissau, Guinea-Bissau. 131 Republica da Guinea-Bissau (2014). Plan Strategique de Nutrition 2015-2019. Ministerio da Saude Publica. Bissau, Guinea-Bissau. Republica de Guinea-Bissau (2008b). Orientações Estratégicas para o Desenvolvimento dos RHS da Guiné-Bissau - 2008-2017. Ministerio da Saude Publica. Bissau, Guinea-Bissau. Russo, G., Pavignani, E., Guerriro, C., Neves, C. (2017). ‘Can we halt health workforce deterioration in failed state s? Insights from Guinea-Bissau on the nature, persistence and evolution of its HRH crisis.’ Human Resources for Health Journal 15:12. UNICEF (2018, forthcoming). ‘Study on Out-of-School Children in Guinea-Bissau.’ United Nations, New York. UNICEF. (2015). ‘Republic of Guinea-Bissau: Multi Indicators Cluster Survey (MICS).’ Bissau, Guinea-Bissau. UNIOGBIS and ISS. (2018). ‘Relaunching defense and security sector reforms in Guinea -Bissau, May 2018.’ Available at https://issafrica.org/research/books-and-other-publications/relaunching-defence-and-security- sector-reforms-in-guinea-bissau. UNIOGBIS, 2017. ‘Thematic report on the right to health in Guinea-Bissau.’ United Nations Integrated Peacebuilding Office in Guinea-Bissau, Bissau, Guinea-Bissau. UNIOGBIS. (2017). ‘Thematic report on the right to health in Guinea-Bissau. UN HR, Bissau, Guinea-Bissau.’ United Nations Integrated Peacebuilding Office in Guinea-Bissau. Bissau, Guinea-Bissau. United Nations. (2017). ‘World Population Prospects’. Report No. ESA/P/WP/248. United Nations, New York. UNODC. (2013). ‘Transnational organized crime in West Africa: A threat assessment.’ February 2013. World Bank (2009). ‘Republic of Guinea-Bissau: Public Expenditure Management and Financial Accountability Review (PEMFAR).’ World Bank, Washington, D.C. World Bank (2013). ‘Republic of Guinea-Bissau: Public Expenditure Financial Accountability (PEFA).’ World Bank, Washington, D.C. World Bank (2016). ‘Guinea-Bissau Systematic Country Diagnostic (SCD): Turning challenges into opportunities for poverty reduction and inclusive growth.’ Report No. 106725-GB. Washington, DC: World Bank. World Bank (2018). ‘Quality for Education all project.’ Draft Project Appraisal Document, 2018. World Bank, Washington, D.C. World Bank (August 2015). ‘Estimating returns to education in Guinea-Bissau with 2010 household survey data (ILAP 2010).’ World Bank, Washington, D.C. World Bank. (2011). ‘World Development Report: Conflict, Security and Development.’ World Bank Group Flagship Report. World Bank, Washington, D.C. World Bank. (2015). ‘Republic of Guinea-Bissau Country Economic Memorandum: Terra Ranca! A Fresh Start!’ Report No. 58296-GW. World Bank. (2016). ‘Republic of Guinea-Bissau: Guinea-Bissau Health Sector Diagnostic. June 2016. World Bank, Washington, D.C. World Bank. (2016). Pension system in Guinea-Bissau: key challenges and prospects for the future. Unpublished report (November 2016). World Bank, Washington, D.C. World Bank. (2016). Pension system in Guinea-Bissau: key challenges and prospects for the future. Unpublished report (November 2016). World Bank, Washington, D.C. World Bank. (2017). ‘CPIA Africa: Assessing Africa’s policies and institutions.’ World Bank, Washington, D.C. World Bank. (2017). ‘Republic of Guinea-Bissau Public Expenditure Analysis.’ Unpublished report. World Bank, Washington, D.C. World Bank. (2018). ‘Doing Business: Reforming to create jobs.’ World Bank Group Flagship Report. World Bank, Washington, D.C. World Development Report (2018). ‘Learning to realize education’s promise.’ World Bank Group Flagship Report. World Bank, Washington, D.C. 132 Annex I: Policy recommendations of World Bank (2017) Public Expenditure Analysis Revenue Recommendation Period* Improve - Carry out analytical work that may focus on assessing: (i) the impact of Medium term revenue shifting from indirect to direct taxation; (ii) the trade-offs for the Government mobilization to introduce the VAT and the appropriate design and implementation; and (iii) mechanism to smoothen volatility of revenues. - Accelerate capacity strengthening at the tax and customs departments Short term through on the job training to carry out inspections. - Formalize terms of reference for the tax exemption committee at the MEF. Short term - Strengthen monitoring to improve compliance to law mandating businesses Medium term to strengthen accounting and financial reporting - Focus on low hanging fruits to boost revenue in a short time, notably by: (i) Short term strengthening the customs depart and restoring the full capacity of the custom information system ASYCUDA and (iii) introducing an amnesty for certain taxpayers to pay outstanding tax arrears. - Continue modernizing the legislation with a focus on (i) establishing a simple transparent regime for taxing small and medium-sized business, (ii) prepare the tax and custom administration’s staff on the new upcoming regulations, (iii) develop the legislation for property taxation, and (iv) harmonize regulations with WAEMU standards. Expenditure Recommendation Improve the - Carry out analytical work that may focus on assessing (i) expenditure tracking Medium term quality of and impact of spending, (ii) funding and performance of public enterprises (iii) spending review expenditure under select budget categories and ensure correct classification, and (iv) the impact of the increase in the minimum wage. - Consolidate public sector personnel/HR database and assign clear Medium term responsibilities for all HR and benefit related adjustments - Improve the quality of public investment by strengthening the investment Medium term database, developing a matrix of short term expenditure priorities, improving the system for formal project appraisal and evaluation and integrating infrastructure maintenance needs in the recurrent budget. PFM Recommendation Improve - Institute government led PFM coordination unit to lead reform effort across Short term public the main categories. financial - Adopt and promulgate WAEMU related legislation Short term management - Improve budget preparation process reflecting negotiated priorities and use Short term it determined spending - Improve data reliability and increase transparency through regular and Short term comprehensive reports on budget execution. - Strengthen public procurement transparency Short term - Ensure full use of SIGFIP for entire expenditure chain, including commitments Medium term and payments. - Strengthen internal and external controls through provision of sufficient Medium term budget to National Finance Inspectorate and Tribunal de Contas. - Strengthen macroeconomic statistic department at MEF to improve quality Medium term and reporting of fiscal statistics. 133 Annex II: Definition of Peer Countries The PER compares Guinea-Bissau’s performance and structural features against relevant (structural) peers identified using a standard benchmarking tool. The team identified Guinea-Bissau’s structural peers using the World Bank’s ‘Find Your Friends’ tool. To the extent that data is available, the PER benchmarks Guinea - Bissau against these peers in addition to the standard Low Income Country averages and SSA averages. Where appropriate and possible, the PER also benchmarks Guinea-Bissau against its WAEMU counterparts (Benin, Burkina Faso, Cote d’Ivoire, Mali, Niger, Senegal, and Togo), which the Government often looks to for policy inspiration. A database – comprising indicators on public spending and outcomes – has been created for this PER using data from government, field surveys, the World Bank, the IMF, and others. The BOOST dataset, which reorganizes Guinea-Bissau’s budget data and presents them in a user-friendly format, has been produced and used for the expenditure assessment in this report. For the education and health chapters, the team also collected data/information from visits to schools and health facilities in five regions of Guinea-Bissau (Oio, Cacheu, Bafata, Quinara, and the Autonomous Sector of Bissau (SAB)). Structural Peers The team selected countries with similar economic characteristics to Guinea-Bissau. The following criteria were set: 1. Population size of less than 10 million. 2. GDP per capita below $600. 3. High dependence on agriculture. 4. Below SSA average life expectancy. 5. Below SSA average trade with the world. 6. Below SSA average government revenue. These criteria deliver the following group of countries: Central African Republic (CAR), Comoros, The Gambia, Sierra Leone, and Tajikistan. Region Population Landlocked/Island Credit GDP Inflation Investment (million) /Small State rating growth to GDP ratio 2010-16 Burundi SSA 9.88 Landlocked fragile Not rated 2.39 8.2 25.2 state Central African SSA 4.73 Landlocked fragile - -2.26 12.1 12.9 Republic state Comoros SSA 0.85 Island, small fragile - 2.37 0.28 19.49 state Gambia, The SSA 2.12 Small fragile state Not rated 2.86 5.68 21.05 Sierra Leone SSA 6.56 Fragile state Not rated 5.21 8.25 23.16 Tajikistan EAP 8.74 Small fragile state Not rated 6.91 6.79 19.27 Guinea-Bissau SSA 1.8 Small fragile state - 3.89 1.78 6.16 134 Annex III: Lessons from Successful Cases Rwanda: Rwanda made signficant progress in raising tax revenue Burkina Faso: Key Tax Revenue 20 through improving tax adminstation and tax policy measures. Over Trends Trade tax the past decade, tax revenue has increased by about 4 ppts, Goods and services tax reaching about 16 percent of GDP in 2016. This increase mainly Direct tax reflected improvements in tax adminstration, including collection Percent of GDP of local taxes at the central level, improvments in auditing procedures, and closer scrutiny of large taxpayers. Tax policy 10 measures have also played a role, which included VAT on mobile airtime, royality taxes on mining, and taxes for special petroleum and infrastructure funds (IMF, 2017). The country has also strengthened taxpayer registeries and electronic filing, and reduced outstanding arrears more efficiently. In addition, the Rwanda 0 Revenue Authority formulated royality taxes and revised its Fixed 2000 2002 2004 2006 2008 2010 2012 2014 2016 Asset Tax legislation which instituted new rates and a property valuation system based on market prices. On VAT, it rationalized Source: IMF. various incentive schemes to include only strategic sectors (e.g. 25 Liberia: Key Tax Revenue Trends exports, manufacturing, energy, informaton and communications technology, construction, and agriculture). Trade tax Burkina Faso: Tax revenue surged from about 11.5 percent of GDP 20 Goods and services tax Direct tax in 2008 to a little over 16 percent in 2013. These gains were the combined outcome of improved tax adminstration and tax reform 15 measures. The government made continued efforts to modernize Percent of GDP tax adminstration, raise awareness of tax compliance among economic actors, and foster efforts to combat evasion and 10 corruption. Revenue mobilization also benefited from tax reform measures to broaden the the coverage of VAT and introduce a special tax on transactions related to mining stocks, among other 5 measures. The increased revenue can also be traced to the phasing out tax holidays for some mining companies and the increase in the royalty rate from a single 3 percent rate to a sliding scale of 3-5 0 2000 2002 2004 2006 2008 2010 2012 2014 2016 percent depending on global prices (with $1,500/ounce being the trigger level for the higher rate, and US$1,300/ounce for the lower Source: IMF. rate). Rwanda: Key Tax Revenue Trends Liberia: The tax-to-GDP ratio in Liberia increased by 6 ppts of GDP during 2005–07 and 4 ppts between 2008 and 2012. The increase Direct tax 15 Goods and services tax in revenue principally reflected improved tax and customs Trade tax administration, as well as tax policy measures. The revenue administration measures included: establishing a taxpayer service 10 Percent of GDP center to reduce compliance costs; computerizing taxpayer registration, filing and payments; conducting audits of large taxpayers, and strengthening the large taxpayers’ office. In addition, Liberia took capacity enhancing measures in the key 5 revenue agencies, including hiring international experts. Liberia’s tax policy reforms involved raising excise taxes on alcoholic beverages, cigarettes, cell phone usage, and airlines and hotels. The goods and services tax (GST) was applied to duty-paid value (i.e. 0 2006 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017(p) after customs has been applied) and the scope of the GST was extended to include restaurant services. Source: IMF. Source: IMF and World Bank Staff assessment. 135 Annex IV: Data Tables Table 1B.1: Key Revenue Trends (2010–17), Percent of GDP 2010 2011 2012 2013 2014 2015 2016 2017 Revenues and grants 20.6 17.6 14.8 11.9 19.4 20.2 16.4 17.4 o/w grants 9.9 7.3 5.6 3.9 7.8 6.6 4.0 4.6 Tax revenues 7.9 7.8 7.7 6.8 8.5 10.0 9.9 10.8 Direct revenues 2.3 2.3 2.5 2.1 2.2 2.8 3.4 2.8 Corporate tax 1.4 1.5 1.6 1.3 1.3 1.5 2.1 1.8 Personal Income tax 0.2 0.3 0.2 0.3 0.2 0.5 0.4 0.4 Property contribution 0.6 0.5 0.5 0.5 0.6 0.7 0.7 0.6 Other direct taxes 0.0 0.1 0.1 0.1 0.0 0.1 0.1 0.1 Indirect taxes 5.6 5.5 5.3 4.7 6.3 7.2 6.5 8.0 Tax on imports 1.4 1.3 1.4 1.2 2.1 1.8 1.5 2.1 Extraordinary tax on cashew 0.5 0.7 0.6 0.7 0.7 0.9 0.9 0.9 Imported consumer goods 0.5 0.4 0.4 0.4 0.5 0.7 0.6 0.7 General sales tax 2.6 2.5 2.4 2.0 2.5 3.2 2.9 3.8 Other indirect taxes 0.5 0.5 0.5 0.5 0.5 0.7 0.5 0.6 Non-tax revenues 2.8 2.5 1.5 1.1 3.2 3.6 2.6 2.4 Fishing license, fines, etc. 1.6 1.2 1.3 1.1 2.0 2.5 1.7 1.0 Financial compensation 1.2 1.3 0.0 0.0 1.1 1.0 0.8 0.8 Sales of goods and services 0.1 0.0 0.2 0.1 0.1 0.1 0.1 0.6 Table 1B.2: Decomposition of Public Spending (2010–17) Percent of total spending Percent of GDP 2010– 2014–17 2010 2011 2012 2013 2014 2015 2016 2017 13 Current expenditures 56.1 65.5 71.4 66.2 61.3 62.2 67.0 62.4 10.9 13.47 Wages and salaries 23.5 27.8 28.9 32.5 26.6 22.1 22.2 22.0 4.7 5.1 Goods and services 9.1 8.4 14.5 11.8 9.4 11.8 8.5 13.4 1.8 2.3 Transfers 14.2 15.9 16.8 16.8 13.0 13.5 22.2 15.6 2.7 3.5 Interest payments 0.7 0.2 0.5 0.2 1.4 3.4 2.2 2.8 0.1 0.5 Other expenditures 8.6 13.3 10.7 4.8 10.9 11.5 11.9 8.5 1.6 2.3 Capital Expenditures 42.0 34.0 26.2 33.8 34.6 34.7 30.8 35.7 5.8 7.4 Domestic 0.5 0.2 1.4 2.6 1.4 2.5 2.0 1.4 0.2 0.4 External 41.4 33.8 24.8 31.2 33.2 32.1 28.8 34.3 5.6 7.0 Special Treasury Acc. 1.9 0.5 2.4 0.0 4.1 3.2 2.2 1.9 0.2 0.6 Table 1B.3: Economic Classification of Current Expenditure (2010–17) 2010– 2014– 2010 2011 2012 2013 2014 2015 2016 2017 13 17 Percent of current expenditure Average Wages and salaries 42 42 40 49 43 36 33 35 43 37 Goods and services 16 13 20 18 15 19 13 22 17 17 Transfers 25 24 24 25 21 22 33 25 25 25 Interest payments 1.2 0.3 0.7 0.4 2.4 5.4 3.3 4.5 0.6 4.9 136 Other 15 20 15 7 18 18 18 14 14 17 Table 1B.4: Composition of Transfers and Subsidies by Beneficiary Entity (2010–17) 2010–13 2014– 2010 2011 2012 2013 2014 2015 2016 2017 17 Percent of GDP Average Pensions 1.3 1.0 1.2 1.2 1.2 1.0 1.2 1.1 1.2 1.1 Embassies, SOEs, etc. 0.5 0.5 0.5 0.3 0.7 0.7 2.5 0.5 0.5 1.1 National Assembly 0.5 0.5 0.5 0.5 0.6 0.7 0.5 0.4 0.5 0.6 Private Individuals 0.5 0.4 0.5 0.4 0.4 0.3 0.3 0.2 0.5 0.3 Other transfers 0.1 0.1 0.1 0.0 0.1 0.4 0.1 1.0 0.1 0.4 Total 2.9 2.6 2.8 2.5 3.0 3.1 4.6 3.3 2.7 3.5 Table 1B.5: Composition of Other Current Expenditures (2010–17) 2010– 2014– 2010 2011 2012 2013 2014 2015 2016 2017 13 17 Percent of GDP Average Other common expenditures 0.6 0.8 0.8 0.3 1.4 1.1 1.4 0.5 0.6 1.1 Tax collection incentives 1.0 1.2 1.0 0.4 1.1 1.5 1.0 1.3 0.9 1.2 Other current expenditures 0.2 0.1 0.0 0.0 0.0 - - - 0.1 0.0 Total 1.8 2.1 1.8 0.7 2.5 2.6 2.4 1.8 1.6 2.3 Table 1B.6: Composition of Purchases of Goods and Services by Sector (2010-17) 2010 2011 2012 2013 2014 2015 2016 2017 Percent of total purchases of goods and services Administration 60.4 55.4 45.8 46.6 49.1 55.3 55.9 60.3 Finance and economy 41.0 32.8 28.9 25.5 29.8 40.6 36.3 44.8 Presidency 9.7 9.6 8.6 13.5 9.6 5.9 9.7 4.8 Parliament 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Justice 1.7 4.0 1.5 0.7 1.3 1.4 1.5 1.8 Other 8.0 9.0 6.7 6.8 8.3 7.4 8.4 8.8 Defense and Security 19.5 31.1 47.8 50.0 44.9 32.3 34.5 30.5 Defense 15.4 27.0 42.8 45.5 35.9 23.6 23.1 18.7 Interior Security 4.0 4.1 5.0 4.4 9.0 8.7 11.5 11.8 Social 18.4 9.8 4.8 2.5 3.0 9.8 5.8 6.8 Education 10.7 5.5 1.0 0.5 0.5 4.9 1.4 0.4 Health 7.6 3.7 3.7 2.0 2.1 4.5 3.9 6.1 Other 0.1 0.6 0.1 0.0 0.4 0.4 0.4 0.3 Economic 1.8 3.7 1.6 1.0 3.1 2.6 3.8 2.4 Agriculture and rural dev. 0.2 0.4 0.2 0.1 0.3 0.4 0.9 0.6 Energy 0.0 0.3 0.1 0.1 0.5 0.2 0.3 0.3 Public works 0.4 0.5 0.3 0.1 0.2 0.2 0.1 0.1 Commerce, tourism, and art 0.4 1.1 0.4 0.3 0.8 0.7 1.0 0.5 Transport and telecoms 0.0 0.3 0.1 0.1 0.3 0.2 0.3 0.1 Environment 0.0 0.6 0.2 0.1 0.2 0.1 0.2 0.2 Other 0.8 0.5 0.2 0.2 0.7 0.7 1.0 0.6 137 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Table 1B.7: Budget Execution Deviations (2010–17) Percent of budget appropriations Avg. Avg. 2010 2011 2012 2013 2014 2015 2016 2017 2010–13 2014–17 Wages and salaries -3.4 2.5 2.9 -4.9 2.0 0.8 2.5 2.2 -0.7 1.9 Goods and services -17.3 -5.1 13.0 -30.5 -2.7 17.0 -15.1 18.4 -10.0 4.4 Transfers -14.2 1.6 -1.6 -17.5 4.4 18.4 97.7 6.2 -7.9 28.6 Other current expenditure 26.1 50.6 -7.5 -55.3 117.9 140.0 151.7 58.2 3.5 117.0 Interest payments -87.8 -96.5 61.5 -64.0 -20.1 37.5 -8.0 58.6 -46.7 17.0 Investments -14.6 -21.7 -56.9 -34.9 33.5 13.4 1.7 11.3 -32.0 15.0 Special treasury account -49.1 -82.4 -15.9 -100.0 304.9 34.6 -6.1 -34.3 -61.9 74.8 Total -14.1 -10.7 -25.0 -25.9 22.1 19.8 21.0 9.1 -18.9 18.0 Table 1B.8: Composition of Total Expenditure by Sector (2010–17) Avg. Avg. 2010 2011 2012 2013 2014 2015 2016 2017 2010–13 2014–17 Percent of total spending Percent of GDP Administration 43.7 44.9 44.3 32.9 37.7 45.2 50.3 45.1 7.1 9.7 Finance and economy 28.7 32.3 29.6 18.4 25.0 31.0 39.0 34.8 4.7 7.0 Presidency 1.3 1.3 1.5 1.9 1.1 0.8 1.1 1.1 0.3 0.2 Parliament 2.6 3.2 3.0 3.4 2.6 3.1 2.7 2.2 0.5 0.6 Justice 3.3 3.2 3.4 3.6 2.9 5.6 2.3 2.1 0.6 0.7 Other 7.7 5.0 6.8 5.6 6.1 4.7 5.2 5.0 1.1 1.1 Defense and security 15.9 18.5 19.4 21.0 17.4 12.4 10.8 12.1 3.1 2.9 Defense 11.4 12.4 13.3 13.5 11.5 7.4 6.6 7.2 2.1 1.8 Interior security 4.5 6.1 6.1 7.5 5.9 5.0 4.2 4.8 1.0 1.1 Social 20.1 19.7 25.0 26.0 26.2 26.4 22.7 28.6 3.8 5.7 Education 9.3 11.6 11.9 14.4 13.4 13.6 10.1 12.5 2.0 2.7 Health 9.2 6.9 11.8 9.5 9.4 11.2 11.3 14.2 1.6 2.5 Other 1.6 1.2 1.2 2.1 3.4 1.6 1.4 1.8 0.3 0.4 Economic 20.3 16.9 11.4 20.2 18.7 16.0 16.2 14.2 2.9 3.6 Agriculture and rural devt 5.1 5.6 2.0 4.6 1.5 2.1 1.5 1.9 0.7 0.4 Energy 0.0 0.1 0.1 0.1 0.4 1.5 3.6 0.7 0.0 0.3 Public works 7.1 3.9 3.8 9.7 11.0 5.4 5.5 6.6 1.0 1.6 Commerce, tourism and art 0.8 0.7 0.7 0.6 0.6 0.6 0.8 0.5 0.1 0.1 Transport and telecoms 0.0 0.1 0.2 0.2 0.2 0.5 0.3 0.1 0.0 0.1 Environment 0.1 1.3 0.5 0.2 0.3 0.9 0.7 0.5 0.1 0.1 Other 7.2 5.3 4.0 4.8 4.6 5.1 3.7 4.0 0.9 1.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 17.0 20.7 Table 1B.9: Composition of Domestic Expenditure by Sector (2010–17) Avg. Avg. 2010 2011 2012 2013 2014 2015 2016 2017 2010–13 2014–17 Percent of total spending Percent of GDP 138 Administration 54.9 53.8 52.2 46.9 50.3 52.2 61.1 59.1 5.8 7.9 Finance & economy 33.9 36.7 32.6 25.8 32.0 35.0 45.1 43.8 3.6 5.5 Presidency 2.3 1.9 2.0 2.8 1.6 1.3 1.6 1.6 0.3 0.2 Parliament 4.5 4.8 4.0 4.9 3.9 5.0 3.9 3.2 0.5 0.6 Justice 2.7 3.0 4.5 5.2 3.7 3.4 3.0 2.9 0.4 0.5 Other 11.5 7.3 9.1 8.2 9.0 7.4 7.5 7.6 1.0 1.1 Defense and security 25.1 25.3 25.6 30.5 22.4 19.9 15.7 18.6 3.0 2.7 Defense 17.7 16.8 17.4 19.6 13.6 11.9 9.6 11.0 2.0 1.7 Interior security 7.4 8.5 8.2 10.9 8.8 8.0 6.1 7.3 1.0 1.1 Social 17.6 17.2 16.1 16.6 19.0 17.1 15.9 16.3 1.9 2.4 Education 12.2 12.4 10.7 10.7 13.3 12.3 11.3 12.0 1.3 1.7 Health 5.2 4.5 5.4 5.9 5.5 4.5 4.4 4.2 0.6 0.7 Other 0.2 0.3 0.1 0.0 0.2 0.3 0.2 0.1 0.0 0.0 Economic 2.4 3.7 6.0 5.9 8.4 10.8 7.3 6.3 0.5 1.2 Agric. & rural dev. 1.1 0.8 0.6 0.7 0.6 0.9 0.5 0.7 0.1 0.1 Energy 0.0 0.1 0.1 0.2 0.2 0.9 1.2 0.2 0.0 0.1 Public works 0.4 0.6 0.3 3.6 0.3 1.8 0.3 0.8 0.1 0.1 Commerce. tourism & art 0.5 0.9 0.9 0.7 0.8 0.8 1.1 0.8 0.1 0.1 Transport & telecom. 0.0 0.2 0.3 0.3 0.3 0.2 0.2 0.1 0.0 0.0 Environment 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 Other 0.4 1.1 3.7 0.4 6.0 6.2 4.1 3.6 0.2 0.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 11.2 14.3 Table 1B.6: Composition of Capital Expenditure by Sector (2010–17) Avg. Avg. 2010 2011 2012 2013 2014 2015 2016 2017 2010–13 2014–17 Percent of total spending Percent of GDP Administration 26.6 26.8 21.4 2.1 11.5 21.5 20.7 19.8 1.2 1.4 Finance & economy 19.3 22.8 21.1 2.1 9.6 11.0 18.6 19.2 1.0 1.1 Presidency 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Parliament 0.0 0.0 0.1 0.0 0.0 0.1 0.0 0.2 0.0 0.0 Justice 4.5 3.6 0.1 0.0 1.4 10.1 0.9 0.3 0.1 0.2 Other 2.9 0.4 0.1 0.0 0.5 0.3 1.2 0.0 0.1 0.0 Defense and security 4.0 5.7 1.0 0.0 7.5 0.2 0.1 0.0 0.2 0.2 Defense 3.3 4.2 1.0 0.0 7.3 0.1 0.0 0.0 0.1 0.1 Interior security 0.7 1.6 0.0 0.0 0.1 0.1 0.1 0.0 0.0 0.0 Social 24.2 24.5 50.8 43.6 42.0 46.2 39.9 50.1 2.0 3.3 Education 5.7 9.9 15.3 21.4 13.3 17.3 7.8 13.1 0.7 1.0 Health 15.0 11.4 31.1 16.1 19.2 24.7 28.1 32.2 1.0 1.89 Other 3.5 3.1 4.5 6.1 9.5 4.2 4.0 4.8 0.2 0.4 Economic 45.2 43.0 26.7 54.3 39.0 32.1 39.3 30.1 2.5 2.46 Agric. & rural dev. 10.8 14.8 6.0 12.2 3.4 5.2 3.8 4.5 0.7 0.3 Energy 0.0 0.0 0.0 0.0 0.8 4.0 11.3 1.6 0.0 0.3 Public works 16.3 10.8 13.8 28.1 31.5 15.1 17.6 18.2 1.0 1.5 Commerce. tourism & art 1.3 0.3 0.3 0.3 0.2 0.4 0.4 0.0 0.0 0.0 Transport & telecom. 0.0 0.0 0.1 0.0 0.3 1.1 0.7 0.0 0.0 0.0 Environment 0.2 3.7 1.5 0.4 0.7 2.3 2.0 1.2 0.1 0.1 Other 16.6 13.3 5.1 13.3 2.1 4.0 3.3 4.7 0.8 0.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 5.8 7.4 139 Annex V: Staff Numbers per Ministry Table 2A.1: Staff Numbers per Ministry (2017) MINISTERIO 2017 Assembleia Nacional Popular 148 Presidência da República 101 Presidência do Conselho de Ministros 230 Supremo Tribunal da Justiça 154 Tribunal de Contas 68 Ministério da Justiça 335 Ministério Público 233 Ministério dos Negócios Estrangeiros, da Cooperação Internacional e das Comunidades 279 Ministério da Defesa Nacional e dos Combatentes da Liberdade da Patria 4.358 Ministerio do Interior 4.362 Ministério da Administração Territorial 132 Ministério das Finanças 659 Ministério da Função Pública, Trabalho e Modernização do Estado 135 Ministério da Economia, do Plano e da Integração Regional 136 Ministerio da Educacao e do Ensino Superior 8.685 Ministerio da Comunicaçao Social 182 Ministério da Saúde Pública 1.974 Ministério da Mulher, Familia e Coesão Social e Luta contra a Pobreza 65 Ministerio de Obras Publicas, Habitacao e Urbanismo 254 Ministério da Energia e dos Recursos Naturais 146 Ministério da Agricultura e Desenvolvimento Rural 487 Ministerio do Comercio i Industria 127 Ministério do Turismo e Artesanato 52 Ministerio das Pescas 86 Secretaria de Estado de Juventude, Cultura e dos Desportos 164 23.552 140 Annex VI: Wage Bill Projection Model Forecasts for the evolution of the wage bill from 2018 to 2023 were developed using a Microsoft Excel- based simulation model. In addition to establishing a baseline projection to inform an understanding of the size and potential trajectory of the wage bill over the medium term, the model was used to analyze how the wage bill might evolve under different assumptions regarding GDP growth and policies governing hiring and compensation (i.e., to conduct a sensitivity analysis around the baseline projection) and to estimate the fiscal impact of potential options for reform. The simulation model reconstructed the wage bill in 2017 using individual employee data and wage tables governing compensation. Employee data for individual workers, including information regarding their grade and wage level were obtained for the month of December 2017 from the Ministry of Economy and Finance’s integrated human resource management database (SIGRHAP) and the Ministry of Public Administration’s database. These two databases were then joined into a single database using individual civil servant ID numbers. In conjunction with the wage tables governing central government compensation, the data were used to create distributions for the number of workers by age and their associated mean wages by age. Once age-specific headcounts are multiplied by mean wages and summed, the distributions sourced from the database accounted for roughly 95 percent of the budget total for 2017 for wages and salaries. The workforce was disaggregated into four groups defined primarily by the wage tables that govern their compensation. The disaggregation of the workforce was necessary for two reasons: (i) significant heterogeneity exists across key sectors of the government in the distribution of their workers and their mean wages by age; and (ii) sector-specific options for policy reform cannot be simulated unless the workforce is disaggregated in the model. The four groups were (i) Civil Service (which includes employees working across the whole civil service who are not specialists in one of the other three categories); (ii) Justice (judges and other judicial system employees); (iii) Health (frontline workers within the Ministry of Health, e.g. doctors and nurses), and (iv) Police/Defense (which includes workers paid using the wage tables for the Police and the Army). Table 2B.1: Number of Workers and Mean Wages, by Employment Group (2017) Employment Group Number of workers Mean annual wage (CFA) Annual wage bill (CFA) Civil Service 12.015 1.122.603 13.488.075.045 Health 1.671 1.349.262 2.254.616.802 Justice 755 4.089.277 3.087.404.135 Police/Defense 7.404 1.533.881 11.356.854.924 Total 21.845 1.381.870 30.186.950.906 Source: IMF staff estimates. Headcounts and for each employment group were scaled up to match the total civil service headcount and the budget total for 2017 for wages and salaries. The number of workers by age was scaled up while the associated mean wages for each age group was kept constant, given the absence of data needed to assess the relative wage expenditures attributable to workers paid outside of SIGRHAP (which includes defense personnel paid in cash). The scaled distributions, therefore, capture all wages and salaries paid by the budget. The distributions of workers by age and their associated mean wages were then projected forward from 2018 to 2023. The distribution of workers by age for each employment group was projected on the basis of (i) an input variable representing the number of new hires in each year of the simulation period; and (ii) assumptions regarding the mean and distribution of the age of retirement and the mean and distribution of the age of entry of new entrants into the workforce. The distribution of mean wages by age was projected on the basis of (i) an input variable representing the exogenous annual rate of increase in wages (i.e. an increase in wage levels as determined by salary tables); (ii) a second input variable intended to capture the value of wage increases due to 141 promotions; and (iii) a third input variable intended to capture notch salary increases (i.e. the increase in wages associated with workers being elevated to higher point-value within their grade in the wage table; something which has so far only been implemented for health workers, with plans to implement for all civil servants). The resulting distributions were then used to compute total wage spending in each year of the simulation period and to express wage spending relative to nominal GDP using the macroeconomic forecast developed by the IMF. The structure of the model meant it could be used to simulate a range of policies governing employment and compensation under varying assumptions for macroeconomic conditions . Collectively, the inputs and assumptions enable the model to simulate the impact on wage spending of policies on replacing retiring workers with new recruits, the hiring of workers into newly created positions, the average age of new hires, changes in the wage tables, the effective introduction of notch increases across the whole civil service, changes in the progression of workers along the point-values of their pay grade, and changes in the retirement age. The model also captured the impact of wage drift on the evolution of the wage bill over the projection period. The term “wage drift” refers to aspects of public employment policy – such as longevity-related increases in compensation (that is, the “notch” adjustment), promotion patterns and policies, and practices regarding the reclassification of employment positions – that drive wage growth independently of broader decisions regarding the overall size of the workforce or wage levels. Conceptually, wage drift is intended to capture all factors that create upward pressure on wage spending independent of changes in the total number of workers or in their respective wage table. Estimating the impact of wage drift on wage spending relied on the rate of increase of the average wage of workers who remained in the workforce between 2014 and 2017, a period where there was no official wage increase. This was calculated using SIGRHAP data from 2014 to 2017. The unified database created for this study included granular-level data on each worker’s grade and point-value. Reform options were evaluated relative to a baseline that reflects assumptions regarding the evolution of the workforce and compensation. Specifically, the baseline projection presumed the authorities’ current policies continue to be applied across the projection period. In terms of their fiscal impact, the two most important policies are the following: (i) wage levels as expressed in salary tables remain unchanged; and (ii) the awarding of a notch increase every three years for frontline health workers not yet at the highest point for their grade. Wage increases were simulated using inflation and GDP projections sourced from the macroeconomic forecast developed by the IMF. In line with historical data, the impact of wage drift was assumed at 3.87 percent for all years of the projection period (i.e., mean salaries were adjusted upward by 3.87 percent for everyone in every year irrespective of decisions regarding overall wage increases). In the absence of stated policies governing employment, the baseline scenario assumed that the number of central government workers would grow in line with the current rate of growth for the working-age population using demographic projections developed by the United Nations. 142 Annex VII: Public Sector Wage Tables Table 2C.1: Civil Service Grupos Categoria Cl. Indice Salario Base Subsidio Iliquido Grupo 0 Presidente da Republica P1 0P01 375 750.000 150.000 900.000 Presidente da Assembleia P2 0P02 300 600.000 120.000 720.000 Primatura P3 0P03 281 562.000 112.400 674.400 Ministros P5 0P05 188 376.000 75.200 451.200 Sec. De Estado; Ass. Juridico P6 0P06 169 338.000 67.600 405.600 Grupo I Sec. Geral; Contador Chefe A 1A01 129 257.500 51.500 309.000 Inspector Geral B 1B01 116 232.780 46.556 279.336 Dir. Geral; Cont. Esp.; Secretarias de (0P03) C 1C01 103 206.000 41.200 247.200 Pr. Inst.; Dir. Ensino Sup.; Dir. Hops. Central; Governador C 1C01 103 206.000 41.200 247.200 Chefe de Gabinete D 1D05 100 199.820 39.964 239.784 Assesor E 1E01 75 150.380 30.076 180.456 Grupo II Dir. Serviço 1ª; Tecnicos (BO); Cont. Verif. A 2A01 55 109.180 21.836 131.016 Dir. Serviço 2ª; Cont. Ver. 2ª B 2B01 47 94.760 18.952 113.712 Del. Reg. Ensino; Del. Reg. Saude B 2B01 47 94.760 18.952 113.712 Dir. Serviço 3ª; Tecnicos; Prof (B); Sec./Escr. C 2C01 41 81.900 16.380 98.280 Dir. E. Secondario; Del. Sub-Reg. Ensino; Pres. Sector C 2C01 41 81.900 16.380 98.280 Grupo III Chefe Rep.; Dir. EB; Tec. (C,D,E) A 3A01 35 69.300 13.860 83.160 Secretarias de Min. & SE A 3A01 35 69.300 13.860 83.160 Tec. Sup. 2ª Secretaria de Min. & SE; Secret. D. Geral B 3B01 30 60.900 12.180 73.080 Tec. Sup. Estag. C 3C01 25 50.400 10.080 60.480 Tec. Medio D 3D01 22 44.100 8.820 52.920 Chefe Secção E 3E01 19 37.800 7.560 45.360 1º Oficial, Oficial de Deligencia F 3F01 16 31.500 6.300 37.800 2º Oficial G 3G01 14 27.300 5.460 32.760 3ª Oficial H 3H01 13 26.000 5.200 31.200 Aspirante I 3I01 13 26.000 5.200 31.200 Escrituario Dact. J 3J01 13 26.000 5.200 31.200 Grupo IV Pessoal Não Adm. A 4A01 13 26.000 5.200 31.200 Pessoal Não Adm. B 4B01 13 26.000 5.200 31.200 143 Table 2C.2: Justice Categoria Indice Salario Base Subsidio Iliquido Pres. do STJ, Proc. Gral Rep., Pres. Trib. Contas 6A01 472,08 944.160 188.832 1.132.992 Vice-P. STJ, Vice-P.G.R., Vice-P. T.C. 6B01 395,83 791.665 158.333 949.998 Juiz Cons. STJ, P.G.A., Juiz Cons. T.C., Dir. Nac. PJ 6C01 302,08 604.165 120.833 724.998 Juiz Desemb., P.R., D.N.A.PJ 6D01 264,58 529.165 105.833 634.998 Juiz de direito & Delegados > 3 anos, Insp. Coord. Niv. 6E01 227,08 454.165 90.833 544.998 Juiz de direito & Delegados <= 3 anos, D.C. Prision. 6F01 189,58 379.165 75.833 454.998 Dir. Do DIC, Dir. Do LPC, Dir. Do DCATE, DGPGR 6G01 187,50 375.000 75.000 450.000 Perito sup. Crim. Niv. III 6H01 167,92 335.835 67.167 403.002 Sec. Judicial & Sec. Tecnicos no Tribunal Sup. 6I01 145,83 291.665 58.333 349.998 Sub-inspector de Nivel III 6I01 145,83 291.665 58.333 349.998 Secretarios Judiciais, Secretarios Tecnicos 6J01 125,00 250.000 50.000 300.000 Juiz sectorial & Delg. Sectorial sem licença 6J01 125,00 250.000 50.000 300.000 Agente de Inv. Crim. De Niv. I, Chefe G. Prisional 6K01 114,58 229.165 45.833 274.998 Escrivao de Directo, Tecnicos Principais 6L01 104,17 208.335 41.667 250.002 Adjunto Escrivao de Direito, Tecnico Adjunto 6N01 83,33 166.665 33.333 199.998 Agente de Inv. Crim. De Niv. II, Guarda Pr. Nivel I 6N01 83,33 166.665 33.333 199.998 Oficiais de Deligencia, Escrituarios Judiciais 6O01 62,50 125.000 25.000 150.000 Agente de Inv. Crim de Niv. III, Guarda Pr. Niv. II 6O01 62,50 125.000 25.000 150.000 Chefe de serviços de segurança interna 6P01. 54,17 108.335 21.667 130.002 Pessoal de segurança interna de nivel III 6Q01 50,00 100.000 20.000 120.000 Table 2C.3: Defense and Security DESCRIÇAO CATEGORIA INDICE SALARIO BASE SUBSIDIOS ILIQUIDO GENERAL AA01 280 560.000 112.000 672.000 TENENTE-GENERAL AB01 250 500.000 100.000 600.000 MAJOR-GENERAL AC01 220 440.000 88.000 528.000 BRIGADEIRO-GENERAL AD01 200 400.000 80.000 480.000 CORONEL BA01 185 370.000 74.000 444.000 TENENTE-CORONEL BB01 175 350.000 70.000 420.000 MAJOR BC01 135 270.000 54.000 324.000 CAPITAO CA01 82 164.000 32.800 196.800 TENENTE CB01 61 122.000 24.400 146.400 ALFER ES CC01 51 102.000 20.400 122.400 SARGENTO MOR DA01 30 60.000 12.000 72.000 1° SARGENTO DB01 27 54.000 10.800 64.800 SARGENTO DC01 26 52.000 10.400 62.400 FURIEL DD01 25 50.000 10.000 60.000 NOUVEAU DE01 20 40.000 8.000 48.000 CABO EA01 14 28.000 5.600 33.600 SOLDADO EB01 13 26.000 5.200 31.200 NOUVEAU EC01 12 24.000 4.800 28.800 144 Table 2C.4: Health 1 2 3 4 5 6 Categoria cod0 cod1 Indice Iliquido Indice Iliquido Indice Iliquido Indice Iliquido Indice Iliquido Indice Iliquido Medico Esp.Hosp. Pr. 2A05 5A01 81,4 195.288 83,4 200.232 86,5 207.648 0,0 0,0 0,0 Medico Esp. Hosp. 2B05 5B01 75,2 180.456 77,3 185.400 80,3 192.816 0,0 0,0 0,0 Ass. Clin. Geral/Equi 2C05 5C01 69,0 165.624 71,1 170.568 74,2 177.984 0,0 0,0 0,0 Medico 2D05 5D01 61,8 148.320 64,9 155.736 68,0 163.152 0,0 0,0 0,0 Tec. Sup. Eq. Medico 2E05 5E01 61,8 148.320 64,9 155.736 68,0 163.152 0,0 0,0 0,0 Enfirmeiro Superior 2F05 5F01 61,8 148.320 64,9 155.736 68,0 163.152 0,0 0,0 0,0 Tec. Sup. A. Soc. 2G05 5G01 46,4 111.240 49,4 118.656 52,5 126.072 0,0 0,0 0,0 Enfirmeiro Professor 2H05 5H01 41,0 98.280 43,1 103.320 44,3 106.296 46,4 111.240 48,4 116.184 0,0 Tec. Dt. Especialista 2I05 5I01 45,3 108.768 48,4 116.184 0,0 0,0 0,0 0,0 Enf./Pat./Especialist 2J05 5J01 39,9 95.760 41,0 98.280 43,1 103.320 44,3 106.296 46,4 111.240 48,4 116.184 Enf. Monitor 2K05 5K01 36,8 88.200 37,8 90.720 39,9 95.760 42,0 100.800 43,3 103.824 45,3 108.768 Tec Dt. 2L05 5L01 34,7 83.160 36,8 88.200 37,8 90.720 39,9 95.760 42,0 100.800 0,0 Enf./Pat./Geral 2M05 5M01 34,7 83.160 36,8 88.200 37,8 90.720 39,9 95.760 42,0 100.800 43,3 103.824 Auxiliar 12esc 2N05 5N01 22,1 52.920 23,1 55.440 25,2 60.480 27,3 65.520 29,4 70.560 0,0 145 Annex VIII: Organigrams of the Security Sector Organigram of the Ministry of Interior 146 Organigram of the Armed Forces 147 Organigram of the Ministry of Defense 148 Security and Institutional Framework 149