Ghana: Social Protection Assessment and Public Expenditure Review November, 2016 Social Protection and Labor and Poverty and Equity Global Practices Africa Region FOR OFFICAL USE ONLY Document of the World Bank ii Table of Contents ABBREVIATIONS AND ACRONYMS .................................................................................................. V ACKNOWLEDGMENTS ...................................................................................................................... VII MAIN MESSAGES................................................................................................................................ VIII EXECUTIVE SUMMARY ...................................................................................................................... IX I: CONTEXT AND PURPOSE.................................................................................................................. 1 A. CONTEXT ............................................................................................................................................................1 B. PURPOSE ............................................................................................................................................................2 II: ANALYSIS OF RISKS AND VULNERABILITIES ......................................................................... 5 A. VULNERABILITY AND POVERTY.................................................................................................................................5 1. Sources of Vulnerability ..............................................................................................................................5 2. Poverty ........................................................................................................................................................7 B. MAIN VULNERABLE GROUPS AND RISKS .................................................................................................................10 1. Poor Pregnant Women and Young Children .............................................................................................11 2. School-Age Children ..................................................................................................................................17 3. Youth.........................................................................................................................................................24 4. Working Poor and Underemployed ..........................................................................................................25 5. Elderly Poor ...............................................................................................................................................27 6. Poor Persons and Families Living with Disabilities and/or HIV/AIDS ........................................................28 III: CORE SOCIAL PROTECTION PROGRAMS TO ADDRESS THE RISKS ............................. 30 A. SOCIAL PROTECTION PROGRAMS ...........................................................................................................................30 1. Social Assistance .......................................................................................................................................30 2. Scholarships and Subsidies to Senior High Students .................................................................................44 3. Social Insurance ........................................................................................................................................47 4. Active Labor Market Programs .................................................................................................................48 B. ENERGY SUBSIDIES..............................................................................................................................................51 IV: ASSESSMENT OF THE SOCIAL PROTECTION SYSTEM ...................................................... 56 A. COVERAGE, GAPS, AND OVERLAPS.........................................................................................................................56 B. THE GENEROSITY OF BENEFITS ..............................................................................................................................60 C. TARGETING ACCURACY ........................................................................................................................................64 D. COST-EFFICIENCY................................................................................................................................................69 E. SUSTAINABILITY..................................................................................................................................................73 F. MONITORING AND EVALUATION............................................................................................................................75 G. INSTITUTIONAL ARRANGEMENTS AND COORDINATION MECHANISMS ...........................................................................77 1. Institutional Arrangements ............................................................................................................................77 2. Coordinating Mechanisms..............................................................................................................................79 V: SOCIAL PROTECTION PUBLIC EXPENDITURE AND MANAGEMENT ............................. 83 A. SOCIAL PROTECTION PUBLIC EXPENDITURES ............................................................................................................83 1. Social Protection Spending .......................................................................................................................83 2. Social Protection Spending Trends ............................................................................................................86 3. Benchmarking of Social Protection Expenditures .....................................................................................86 B. SOCIAL SECTOR PUBLIC SPENDING AND SOURCES OF FINANCING .................................................................................87 1. Total Expenditures and Sources of Financing ...........................................................................................87 2. Social Spending and Sources of Financing ................................................................................................90 i 3. Financing Arrangements...........................................................................................................................96 C. BUDGET MANAGEMENT ......................................................................................................................................99 1. Formulation ..............................................................................................................................................99 2. Execution ................................................................................................................................................103 3. Auditing ..................................................................................................................................................105 4. Ongoing Reforms ....................................................................................................................................107 D. IMPLICATIONS FOR MANAGING SOCIAL PROTECTION EXPENDITURES ..........................................................................108 VI. CONCLUSIONS AND RECOMMENDATIONS ......................................................................... 111 A. CONSOLIDATING GAINS AND BUILDING A STRONGER PROGRAM PORTFOLIO ................................................................111 B. STRENGTHENING MANAGEMENT OF EXPENDITURES ................................................................................................114 C. EXPANDING WELL-TARGETED PROGRAMS AND RATIONALIZING POORLY TARGETED PROGRAMS ......................................116 1. Expanding LEAP and LIPW ......................................................................................................................116 2. Promoting graduation pathways and productive inclusion of the poor and vulnerable ........................117 3. Rationalizing Scholarships ......................................................................................................................118 4. Rationalizing Active Labor Market Programs .........................................................................................118 D. ENSURING SUSTAINABILITY OF EXISTING PROGRAMS ...............................................................................................120 1. Social Assistance and Scholarships Programs .........................................................................................120 2. National Health Insurance ......................................................................................................................120 3. Social Insurance ......................................................................................................................................121 GLOSSARY............................................................................................................................................. 122 APPENDIX A. ORGANOGRAM OF THE MINISTRY OF GENDER, CHILDREN, AND SOCIAL PROTECTION (MOGCSP) .................................................................................................................. 126 APPENDIX B. SUPPORTING TABLES ............................................................................................. 127 REFERENCES ........................................................................................................................................ 130 Boxes Box 3.1: Selected Youth Skill and Entrepreneurship Development NGO Programs ................... 49 Box 5.1: Budget Formulation and Execution in the District Assembly of Shai-Osu Doku ....... 102 Figures Figure 2.1: Poverty and Extreme Poverty by Region, 2012 ........................................................... 8 Figure 2.2: Sensitivity of Incidence of Poverty to Changes in the Poverty Line, 2012.................. 9 Figure 2.3: Trends in Mortality Rates ........................................................................................... 14 Figure 2.4: Under-Five Nutrition Indicators by Region ............................................................... 15 Figure 2.5 Total Health Expenditures and Out-of-Pocket Expenditures ...................................... 17 Figure 2.6: Primary and Junior High School Net Enrollment and Gender Ratios by Region, 2013 ....................................................................................................................................................... 20 Figure 2.7: Primary and Junior High School Net Enrollment Rates, by Quintile, 2012 .............. 21 Figure 2.8: Unemployment Rate per Age Group, 2012 ................................................................ 24 Figure 2.9: Poverty Incidence by Type of Employment of Head of the Household (Percent) .... 26 Figure 3.1: Registration in the National Health Insurance Scheme by Quintile ........................... 43 Figure 3.2: Fuel Price Adjustments (2008=100)........................................................................... 54 Figure 4.1: Gross Replacement Rates in Selected Sub-Saharan African Countries ..................... 64 Figure 4.2: Concentration Curves and Benefit Incidence for Apprenticeship and Scholarship Programs ....................................................................................................................................... 66 Figure 4.3: Concentration Curves and Benefit Incidence for Sources of Energy ......................... 67 Figure 4.4: Concentration Indexes for Selected Social Protection Programs ............................... 68 ii Figure 5.1: Distribution of Social Assistance Spending by Type of Program with and without spending on Scholarships for SHS, 2014 (Percent) ...................................................................... 85 Figure 5.2: Comparison of Spending on Social Protection and Energy Subsidies (Percent of GDP, Average 2012–14) ............................................................................................................... 85 Figure 5.3: Social Protection Real Expenditure, 2008–14 (GH¢ 2006) ....................................... 86 Figure 5.4: Total Revenues and Expenditures and Fiscal Balance ............................................... 88 Figure 5.5: Central Government Wage Bill, Interest Payments, Energy Subsidies, and Public Investment (Percent of GDP) ....................................................................................................... 88 Figure 5.6: External Financing of the Budget ............................................................................... 89 Figure 5.7: Trends in Total Public, Education, Health, and Social Assistance Real Spending .... 93 Figure 5.8: External Finance as a Share of Total Expenditures (Percent) .................................... 94 Figure 5.9: Deviation between Actual and Budgeted Revenues (GH¢ million)......................... 100 Figure 5.10: Difference between Actual and Budgeted Expenditures ........................................ 104 Tables Table ES.1: International Comparison of Social Assistance Expenditures ................................. xiii Table ES.2: Issues to Be Addressed in the Social Protection Programs ...................................... xvi Table 1.1: Selected Social Indicators .............................................................................................. 2 Table 2.1: Incidence of Poverty and Inequality, 2005 and 2012 .................................................... 8 Table 2.2: Incidence of Poverty by Age Group, 2012 .................................................................. 10 Table 2.3 Access to Basic Services, by Quintile, 2012 ................................................................ 10 Table 2.4: Reproductive Care Indicators (%) ............................................................................... 12 Table 2.5: Child Health Indicators (%) ......................................................................................... 13 Table 2.6: Low Birth Weight and Mortality Indicators, by Region.............................................. 14 Table 2.7: Comparisons of Under-Five Nutrition Indicators, Selected Sub-Saharan African Countries ....................................................................................................................................... 15 Table 2.8: Health Payments per Patient for Those Who Visited Health Facilities ...................... 16 Table 2.9: Living Status of Children (0–17 years) (% distribution) ............................................ 18 Table 2.10: Enrollment in Early Childhood Development Programs, 2013 ................................. 19 Table 2.11: Country Comparison of Net Primary School Enrollment (%)................................... 19 Table 2.12: Net Enrollment Rates and Children Who Do Not Attend School, 2013 ................... 20 Table 2.13: Reason for Never Attending School (Children 4-18 Years Old) (%) ........................ 22 Table 2.14: Annual Cost of Primary, Junior High School, and Senior High School Public Education per Student (GH¢)........................................................................................................ 23 Table 2.15: Apprenticeship Participation by Youth (15–25 years) .............................................. 25 Table 2.16: Rural Extreme Poor Households ............................................................................... 26 Table 2.17: Elderly Living Arrangements .................................................................................... 27 Table 2.18: SSNIT Pensioners ..................................................................................................... 27 Table 2.19: Composition of Vulnerable Families (%) .................................................................. 28 Table 3.1: Selected Social Protection Programs, 2014 ................................................................. 31 Table 3.2: SSNIT Membership and Pensioners ............................................................................ 47 Table 3.3: Electricity Tariff Structure, Effective October 2014 to March 2015........................... 52 Table 3.4: Energy Subsidies Reported in the MoF Budget ......................................................... 52 Table 3.5: Indicative Maximum Price of Fuels, Effective January 2015 ..................................... 54 Table 4.1: SP Program Scale by Target Group ............................................................................ 57 Table 4.2: Old-Age Pension Beneficiaries in Sub-Saharan African Middle-Income Countries .. 58 iii Table 4.3: Active Contributors to Pension Scheme in Sub-Saharan African Middle-Income Countries ....................................................................................................................................... 59 Table 4.4: Generosity of Programs in Developing Countries Relative to Average Consumption 61 Table 4.5: Simulated Impact of LEAP on Poverty Indicators ...................................................... 61 Table 4.6: Noncontributory Pensions in Sub-Saharan Africa Middle-Income Countries ............ 62 Table 4.7: Wage Setting in Labor-Intensive Public Works Programs .......................................... 62 Table 4.8: SSNIT Benefits ........................................................................................................... 63 Table 4.9: CGH indicator in the Poorest Quintile......................................................................... 65 Table 4.10: Concentration and Progressivity Indexes of Selected Programs ............................... 68 Table 4.11: Typical Operating Costs as Share of Total Program Cost (Percent) ......................... 70 Table 4.12: LIPW Administrative Costs ....................................................................................... 71 Table 4.13: Cost of the Ghana School Feeding Program.............................................................. 71 Table 4.14: Distribution of NHIS Expenses (Percent).................................................................. 72 Table 4.15: SSNIT Expenses ........................................................................................................ 72 Table 4.16: Payment Arrears ........................................................................................................ 73 Table 4.17: NHIS–Expenditures for the Extreme Poor Exemptions ............................................ 74 Table 4.18: Arrears to SSNIT ....................................................................................................... 74 Table 4.19: MoGCSP Staffing, 2015 ............................................................................................ 78 Table 5.1: Social Protection Actual Expenditures, 2008–14 (GH¢ thousand) ............................. 84 Table 5.2: International Comparison of Social Assistance Expenditures ..................................... 86 Table 5.3: Budget Allocations from Discretionary Sources ......................................................... 90 Table 5.4: Education Expenditures and Financing (GH¢ million) ............................................... 91 Table 5.5: Health Expenditures and Financing (GH¢ million) ..................................................... 91 Table 5.6: MoGCSP Expenditures (GH¢ million) ........................................................................ 92 Table 5.7: Social Assistance and Scholarships Financing, 2014 .................................................. 93 Table 5.8: District Assembly Common Fund Disbursements ...................................................... 98 Table 5.9: District Assembly Common Fund by Type of Spending ............................................ 98 Table 6.1: Issues to Be Addressed in the Social Protection Portfolio ........................................ 113 Table 6.2: Simulated Cost of LEAP Expansion.......................................................................... 116 Table 6.3: Simulated Impact on Poverty Indicators of an Expended LEAP .............................. 117 TableB.1: Central Government Accounts ................................................................................... 127 (GH¢ million).............................................................................................................................. 127 TableB.3: Central Government Budget Allocation from Discretionary Sources ....................... 129 (GH¢ million).............................................................................................................................. 129 TableB.4: Budget Allocation to Social Assistance Programs from the Consolidated Fund ....... 129 iv Currency equivalents: US$1.00 = Ghana Cedi, GH¢ 2.9 (Average 2014) Government fiscal year: January 1–December 31 Abbreviations and Acronyms AF Additional Financing (World Bank) ALMP active labor market program ARIC Audit Report Implementation Committees CBT community-based targeting CGH Coady-Grosh-Hoddinott CLIC Community LEAP Implementation Committee COTVET Council on Technical Vocational Education and Training DA District Assembly DACF District Assembly Common Fund DED District Education Department DFID Department for International Development (United Kingdom) DP development partners DSDD District Social Development Department DSD Department of Social Development DSDO district social workers FAO Food and Agricultural Organization of the United Nations GAG Ghana Auditor General GAMA Greater Accra Metropolitan Area GES Ghana Education Service GETFund Ghana Education Trust Fund GHS Ghana Health Service GIFMIS Ghana Integrated Financial Management Information System Girls-PASS Girls Participatory Approaches for Student Success (DFID) GLSS Ghana Living Standards Survey GLST Ghana Luxembourg Social Trust GNHR Ghana National Household Registry GoG Government of Ghana GPRS Growth and Poverty Reduction Strategy GSFP Ghana School Feeding Program GSGDA Ghana Shared Growth and Development Agenda GSOP Ghana Social Opportunities Project GSS Ghana Statistical Service HH household(s) IGF internally generated funds ILO International Labor Organization IMF International Monetary Fund JHS junior high school LEAP Livelihood Empowerment Against Poverty Program LESDEP Local Enterprises and Skills Development Program LIPW Labor-Intensive Public Works program LLIN long-lasting insecticidal nets LMI lower-middle-income LMIC lower-middle-income country M&E monitoring and evaluation MDA Ministries, Departments, and Agencies v MDG Millennium Development Goal MELR Ministry of Employment and Labor Relations MESW Ministry of Employment and Social Welfare MICS Multiple Indicator Cluster Survey MLGRD Ministry of Local Government and Rural Development MIS management information system MMDA Municipal, Metropolitan and District Assemblies MoF Ministry of Finance MoGCSP Ministry of Gender, Children and Social Protection MTEF Medium-Term Expenditure Framework NAP National Apprentice Program NDPC National Development Planning Commission NGO nongovernmental organization NHIA National Health Insurance Authority NHIF National Health Insurance Fund NHIS National Health Insurance Scheme NPA National Petroleum Authority NPECLC National Program for the Elimination of the Worst Forms of Child Labor in Cocoa NPRA National Pensions Regulatory Authority NSPS National Social Protection Strategy NTS National Targeting System NYEP National Youth Employment Program OVC orphans and vulnerable children PBB program-based budgeting PEFA Public Expenditure and Financial Accountability PETS public expenditure tracking survey PMT proxy means test PURC Public Utilities Regulatory Commission PWD persons with disabilities Q quintile (from Q1, poorest, to Q5, richest) SHS senior high school SIT Social Inclusion Transfer SS Scholarships Secretariat SSNIT Social Security National Insurance Trust UCT unconditional cash transfer UNICEF United Nations Children’s Fund USAID United States Agency for International Development WFCL worst forms of child labor WFP World Food Programme WHO World Health Organization vi Acknowledgments This study was prepared by a World Bank team comprising José Silvério Marques, Iyioluwa Teleola Akinlawon, Christabel Ewuradjoa Dadzie, Maame Efua Ofosuhene, and Suleiman Namara, under the overall leadership of Maddalena Honorati (co-TTL) and Vasco Molini (co- TTL). José Silvério Marques and Maddalena Honorati are the main authors of the report, Christabel Ewuradjoa Dadzie coordinated the data collection and Iyioluwa Teleola Akinlawon provided excellent research assistance. The team would like to thank the government officials working on social protection programs for their help, as well as the partners and other stakeholders that are supporting the strengthening of the social protection system in Ghana. The team would particularly like to thank the Honorable Nana Oye Lithur, Minister of Gender, Children and Social Protection (MoGCSP), Mr. Kwesi Armo-Himbson, Chief Director of MoGCSP, and Mr. Mawutor Ablo, Director for Policy Planning and Monitoring and Evaluation (PPME), for their invaluable support. The team is extremely grateful for an in depth technical review of the report carried out during a two-day workshop in Prampram, Ghana, in January 2016 and subsequent rounds of comments by a Technical Committee comprised of Robert Austin (Ministry of Local Government and Rural Development), Sauda Ahmed (Ministry of Finance), Mary Mpereh (National Development Planning Commission), Emma Ofori Agyemang (Ministry of Employment and Labor Relations), Richard Adjetey (Ministry of Gender, Children and Social Protection), Mawutor Ablo (Ministry of Gender, Children and Social Protection), Isaac Osei Akoto (Institute of Statistical Social and Economic Research), Eric Oduro Osae (Institute of Local Government Studies), Desmond Duametu (Ministry of Local Government and Rural Development), Collins Akuamoah (National Health Insurance Authority), Young Opare (Ghana School Feeding Program), David Osei-Wusu (Institute of Local Government Studies), Baaba Otoo (Lead Consultant–Social Protection Policy). Different social protection stakeholders that participated in the technical validation workshop held in Accra in October 2015 played a key role in the successful completion of the report and provided important information, guidance, and recommendations on rationalizing social protection expenditure in Ghana. The report benefitted greatly from comments by Sarah Hague (UNICEF Ghana), Stephen Younger (Ithaca College), and World Bank colleagues, including Dilek Aykut, Theo Braimah Awanzam, Aline Coudouel, Errol Graham, Kathy Lindert, Victoria Monchuk, Manuel Salazar, and Anita Schwarz. We thank Kathleen Beegle (Program Leader, World Bank) for valuable advice in her capacity as Program Leader, Pablo Fajnzylber (Practice Manager, Poverty and Equity Global Practice), and Stefano Paternostro (Practice Manager, Social Protection and Labor Global Practice) for the overall leadership and support. We thank Nancy Morrison for editing the report. The Social Protection Assessment and Expenditure Review was prepared based on the team’s visit to Ghana from February 25 to March 13, 2015, desk research, and continuous follow-up by the members of the team based in Ghana. The report reflects the information received from program managers up to May 2015. vii Main Messages The expansion of key targeted social assistance programs to reach all the extreme poor should be prioritized. Ghana has a comprehensive social protection system to address the main life cycle risks and vulnerabilities, but its coverage, while expanding, is still limited, especially among the poor. Currently, only 25 percent of the extreme poor are covered by LEAP, the country flagship cash transfer program, about 10 percent of rural household are covered by LIPW, and about 31 percent of people in the poorest quintile are registered with a valid NHIS card. These three targeted social assistance programs (LEAP, LIPW, and NHIS exemptions) need to be scaled up to increase the coverage of the poor. Contributory pension coverage through the SSNIT is also low, with less than 8 percent of the elderly receiving a pension. Efficiency gains can be obtained by reducing overlaps in programs, rebalancing the portfolio of programs, and reallocating resources to well targeted social protection programs. The government of Ghana needs to address overlapping scholarships and student subsidies to senior high schools and improve their targeting to the poor, for example. Continuation of efforts to reduce inappropriately targeted and regressive energy subsidies could help create fiscal space for expanding targeted social assistance programs. Spending on electricity and fuel subsidies absorbed more than twice the amount of public resources allocated to social assistance programs in 2014, for instance. More investment is needed to promote sustainable livelihood activities for social assistance beneficiaries who have the capacity to work and potential to graduate, and to facilitate the access of youth to more productive jobs. Minimal initiatives exist to promote the inclusion of the poor and vulnerable in productive activities, as a medium-term strategy to reduce their dependency on public transfers both in rural and urban areas. Effective labor intermediation services in Ghana should be created, including: (i) labor market information systems; (ii) job search assistance; (iii) career counseling; and (iv) special programs for the unemployed, including second chance education programs. The social protection policy under development should clearly define the scope of social protection, set specific short-, medium-, and long-term goals to be achieved, and provide the elements for the development of a medium-term program to achieve the goals. Developing a consistent and agreed definition of social protection is essential, as different stakeholders are still working with very different understandings of which programs social protection comprises. Currently, there are no clear, rallying goals for the social protection system that can be monitored. Once clear goals are established, policies and programs should be geared and monitored toward achieving those goals. The social protection medium-term program and action plan should be supported by a budget and monitoring and evaluation (M&E) framework. The medium-term action plan should set the institutional and administration arrangements of different levels of government to enhance the coordination of the social protection system. The functional responsibilities of different levels of government should be better defined and intergovernmental financial transfers should be geared toward providing MMDAs with the resources to effectively discharge their responsibilities. The action plan should be costed out, and an annual operating plans and budgets should be prepared. The program-based budgeting (PBB) being implemented by the MoF represents a first important step for coordinating the financing and accountability of social protection programs. However, the PBB should explicitly develop functional budget codes for core social protection programs. Reduced dependence on external financing and clearing of government payment arrears would help increase the sustainability of the social protection system. LIPW is currently fully funded and LEAP is partially funded by external sources. To ensure their sustainability over the longer term, external resources should be replaced gradually by domestic funding. In addition, both NHIS and SSNIT require important reforms to ensure their future financial sustainability. The creation of a legal basis for key social protection programs may also contribute to their sustainability. viii Executive Summary Context and Purpose For most of the last two decades, Ghana’s economy has been on a high growth path , which has been accompanied by significant reduction in poverty. Ghana became a lower-middle- income country (LMIC) in mid-2011 and began commercial oil production the same year; as a consequence, it will have to rely more on domestic resources to finance its development in the future. Per capita growth rates averaged more than 2.5 percent between 1983 and 2006, and increased to around 6 percent from 2006 to 2011. The only breaks in the rising growth record since 2000 took place in 2009, as the global financial and economic crisis hit most countries hard, and in 2012, when growth reverted to its long-term trend after the start of commercial oil production. Poverty has declined from 52 percent in 1991 to 21 percent in 2012.1 Despite this progress, large macroeconomic imbalances threaten growth and continued poverty reduction. The major drivers of the large fiscal deficit have been the high public wage bill, interest payments on public debt, and energy subsidies. The government of Ghana (GoG) has embarked on a strong fiscal consolidation and reform process to reduce existing imbalances and improve public expenditure management. In early April 2015, the Board of the International Monetary Fund (IMF) approved a three-year arrangement under the Extended Credit Facility to support the GoG program. The World Bank recently approved a Macroeconomic Stability for Competitiveness and Growth Credit, which is complementing the IMF program by tackling the fiscal imbalance, and other development partners (DP) are also supporting the program. Rapid economic growth and poverty reduction has been accompanied by growing inequality in terms of income and geographical areas. The poverty rate in the northern part of the country has declined much less than in the rest of the country, largely reflecting the region’s much higher rate of subsistence farming and much lower rate of urbanization. In 2012, more than one-third of the nation’s poverty was concentrated in the three northern regions, which have only 17 percent of the population. The disparities in access to basic health and education services between north and south and between income quintiles remain large. A significant proportion of the population remains vulnerable to falling back into poverty in the event of major shocks. The GoG has a comprehensive National Social Protection Strategy (NSPS) in place and is finalizing the National Social Protection Policy. The draft NSPS of 2012 represents the government’s vision for creating an inclusive society by providing sustainable mechanisms to protect people living in extreme poverty or facing vulnerability and exclusion. The NSPS establishes a framework that draws on the 1992 constitution, on international conventions, and on Ghana’s Shared Growth and Development Agenda (GSGDA). It seeks to reduce the population’s exposure and increase resilience to risks, shocks, and exclusion, and provide a social protection floor across the life cycle, particularly to the extreme poor and most vulnerable. The Ministry of Gender, Children and Social Protection (MoGCSP) coordinates and oversees social protection 1 For the sake of comparability, the poverty headcount in 2012 was calculated using the 1999 poverty line. Using the official new poverty line the poverty rate at national level in 2012 would be 24 percent. See World Bank 2015b. ix interventions implemented by multiple ministries and government agencies and is currently seeking to develop an overarching social protection policy and vision for the sector. Given existing macroeconomic challenges, resources allocated to social protection programs should be used in the most cost-effective and pro-poor manner. There is a need to target programs better and to expand coverage of well-targeted programs to mitigate the potentially negative impact of fiscal consolidation on the poor. This report aims to contribute to the development of the national social protection policy under preparation and the rationalization of social protection expenditures, building on recent work by the Government of Ghana (2014). The focus is on social assistance programs (or social safety nets), although social security and active labor market programs are also discussed. This assessment aims to contribute to the rationalization of social protection public expenditures, building on recent work by the Government of Ghana (GoG 2014b). The GoG report (GoG 2014b) includes an analysis of the social expenditure structure and coverage, based on data up to 2012 and on the previous round of the Ghana Living Standards Survey (GLSS-5, 2005). It offers recommendations to rationalize social protection expenditures, including through enhanced effectiveness and coordination of the social protection system. This study complements and further develops the GoG analysis by adding: (i) a more in-depth profile of the population risks and vulnerabilities and social protection needs, based on a life cycle approach; (ii) an update the “state of social protection” (part 1 of the GoG report), based on most recent program administrative data on spending and number of beneficiaries (for 2013 and 2014); (iii) a program effectiveness and incidence analysis using Round 6 of the Ghana Living Standards Survey (GLSS- 6, 2012); (iv) an analysis of the administration costs and operational efficiency of programs with available data; and (v) a deeper analysis of public expenditure management issues and the sustainability of existing programs. This assessment reviews the core social protection programs as identified for the GoG report, based on a consultation process with different ministries and relevant social protection stakeholders. This portfolio of social protection programs includes social assistance programs, social insurance, and active labor market programs. The scope of social protection–the sample of programs analyzed in this report–is the same as in the GoG report with minor adjustments. This report updates the analysis on core social assistance programs, active labor market programs, social insurance schemes including the National Health Insurance Scheme,2 and general fuel and electricity subsidies. The review does not include programs that have been discontinued (see section 1 and 3). On the other hand, the review includes scholarships programs that were covered only partially in the GoG 2014b report, and the National Apprenticeship Program, which was recently introduced. Although scholarships have not been considered as part of social protection in previous studies for Ghana and are not part of the 5 social protection flagship programs defined in the Ghana National Social Protection Policy3, they are used as social protection instruments around the world to help poor households 2 While there are other interventions that support the health delivery system, the report will focus on the NHIS. 3 To achieve its policy objectives, the National Social Protection Policy proposes to improve the design and implementation of five flagship programs in the short term, identifies as: the Livelihood Empowerment Against Poverty, the Labor Intensive Public Works, the School Feeding Program, the National Health Insurance Exemptions and the Basic Education Capitation Grants. Other programs and plans of action are recognized to be relevant to social protection delivery in Ghana and to contribute to promoting sustainable livelihoods and human dignity. x Particularly, social protection related interventions implemented by the Ministry of Health (MoH), Ministry of access education facilities. They can be cash to families and/or in the form of waivers of educational fees or vouchers that reimburse households or schools/facilities for actual expenditures. This study is expected to contribute to the development of the social protection policy under preparation4. The analysis of risks and groups at risk and of the existing portfolio of programs to address these risks, as well as of the analysis of the allocation and management of public resources to social protection and energy subsidies, aims to provide some guidance to establish policy priorities. Vulnerability and Poverty The vulnerable and poor in Ghana face a series of risks resulting from economic shocks, man-made or natural disasters (mostly droughts and floods), and poor household conditions. The role of the social protection system is to minimize the impact of these risks using a “portfolio of coherent programs” that work together to deliver resilience, equity, and opportunity.5 Household vulnerabilities are related to the conditions in which poor families live that expose them to a series of adverse situations. Ghana met Millennium Development Goal (MDG) 1 by halving poverty. However, more than 2.2 million Ghanaians (8.4 percent of the population) still live in extreme poverty: they cannot afford to feed themselves, even if they were to spend all their expenditures on food (GSS 2014a, 2014c). Extreme poverty is higher for children, and affects more than 1.2 million children (based on GLSS-2012). Despite steady progress in reducing poverty and improving social conditions, major challenges remain for all age groups. While Ghana has made considerable progress in reducing maternal, infant, and under-five mortality rates, these rates remain higher than would be expected at the nation’s income level. Regional disparities are high. The under-five mortality rate (per 1,000 live births) is 128 in the Northern Region, compared to a national average of 82. Chronic malnutrition (stunting) affects 22 percent of children under 5 nationwide, but 37 percent of children in the Northern Region. Ghana is unlikely to achieve key Millennium Development Goals to reduce the proportion of children under five who are underweight, and to cut infant, under-five, and maternal mortality rates. While access to health care has increased following the introduction of the National Health Insurance Scheme in 2003, out-of-pocket health payments, after declining initially, have increased, and remain at around 28 percent of total health costs. Likewise, despite rapid increase in enrollment in basic education in recent years, more than 623,000 primary school- age children do not attend primary school, and one in five children is involved in child labor. Many poor parents cannot afford to send and keep children in basic school, while upper secondary school and tertiary education are out of reach for most poor children. Youth employment is a challenge in Ghana, as in many other countries. Many people in Ghana are working poor because of poor quality of jobs, particularly in the agricultural and informal sectors. Meanwhile, few elderly people (above 60 years old) in Ghana retire with a pension, while one in four old people is poor. Social Protection Assessment Education (MoE) and Ministry of Food and Agriculture (MoFA) are relevant for integration in the short and medium term implementation of the policy (MoGCSP 2015). 4 The Ghana National Social Protection Policy (MoGCSP 2015) was under preparation at the time the analysis for this report was conducted. 5 The World Bank’s Social Protection and Labor Strategy 2012 –2022 (World Bank 2012b, xiv). xi To respond to these challenges, Ghana has built an impressive social protection system. Ghana has established social assistance programs that address the principal risks facing Ghanaians. It is developing the tools to implement and monitor social assistance programs in an efficient way. It is developing a governance framework for the sector at the national, regional, and local levels, and the Ministry of Gender, Children and Social Protection (MoGCSP) is formulating an overarching social protection policy and vision for the sector. But challenges remain to consolidate a social protection system. The Ghana social protection system includes social assistance, scholarships, social insurance and labor market programs. It has in place a portfolio of social protection programs, including LEAP (Livelihood Empowerment Against Poverty program), the country’s flagship cash transfer program for the poor (orphans and vulnerable children, OVC; the elderly; persons with severe disabilities and more recently pregnant women through the LEAP 1000); in-kind assistance programs (school feeding and free uniforms and exercise books); a labor-intensive public works program (LIPW) for the poor unemployed; education and health insurance fee waivers; scholarship programs; a three-tier social security system, which covers both the formal and informal sectors; and a number of youth employment programs. Ghana is developing a series of management tools and putting in place a governance framework. The tools include the single registry, a revised targeting mechanism (including a poverty map and proxy means test), an e-payment system for benefits, complaint and redress mechanisms, monitoring and evaluation systems, and management information systems (MIS) for the specific programs. There are also efforts under way to strengthen social auditing of programs at different levels and the existing social case management system. The Cabinet recently approved a multi-sectoral governance framework for social protection at the national, regional, district, and community levels, which is being put in place. Challenges remain in terms of coverage of the poor and generosity of benefits of key social assistance programs. Coverage of the extreme poor by LEAP is still low (less than one-third of extreme poor families), as is the coverage of LIPW, health insurance, and social security. A micro- simulation of the impact of LEAP with 2014 coverage and generosity parameters using GLSS-6 shows only a modest reduction in the extreme poverty gap (8 percent). In contrast, there is substantial program fragmentation of scholarships and youth programs. LEAP transfers, despite doubling in 2012, are relatively small compared to the basic needs of the beneficiaries. On the other hand, current social security pensions are reasonable, but future replacement rates seem too generous and unsustainable. There is room to improve the accuracy of targeting of the key social assistance programs, particularly education (uniforms and exercise books) and scholarships programs. In addition, to put in place the governance framework for the social protection system, planning and coordinating tools must be developed: a social protection policy, a medium-term program, action plan, and budget; and monitoring and evaluation and management information systems. And the social protection system must overcome important sustainability challenges: LIPW is fully financed by development partners, and both the National Health Insurance Scheme (NHIS) and Social Security National Insurance Trust (SSNIT) require important reforms to ensure their future financial sustainability. Social Protection Public Expenditure Review xii Access and availability of basic program monitoring data such as number of beneficiaries and actual spending should be strengthened. Information about social assistance programs is not easily accessible, often program level data are fragmented, dispersed, and incomplete. With a few exceptions, most program management units could not provide information about their target group, how many people they were reaching, how much they were spending, and their sources of financing. Ghana’s spending on social protection is limited compared to international standards, which is consistent with the findings of relatively low coverage and modest generosity of major social assistance programs. Ghana spends 1.4 percent of GDP on social protection, 0.5 percent of GDP on social assistance including scholarships (compared to a global average of 1.6 percent of GDP among developing countries),6 0.9 percent of GDP on pensions and 0.1 on ALMPs. Social assistance (excluding scholarships) represented only 0.9 percent of total public expenditures and 0.3 percent of GDP. Spending on poorly targeted scholarships and subsidies on SHS is almost commensurate to spending on social assistance representing 0.8 percent of budget and 0.2 percent of GDP in 2014. Within social assistance school feeding represent 54 percent of social assistance spending, followed by fee waiver with 16 percent; the Livelihood Empowerment Against Poverty program (LEAP) and labor-intensive public works (LIPW), with 13 percent each; and other programs, with 4 percent. LEAP and LIPW, which are well targeted to the poor, are responsible for only 26 percent of total social assistance expenditures. Most programs are financed with local resources with the exception of LEAP, LIPW, and the Girls Participatory Approaches for Student Success (Girls-PASS) scholarships (through the United Kingdom’s Department for International Development, DFID). Ghana spends a fraction of the amount spent by its Sub-Saharan African lower-middle- income (LMI) peers on social assistance. Table ES.1 presents a comparison of social assistance spending in Ghana (including scholarships for international comparisons) with spending in five other lower-middle-income Sub-Saharan African countries for which data are available, as well with other 35 LMI countries worldwide,.7 Ghana spends 0.5 percent of GDP on social assistance (including scholarships), compared to 2.1 percent of GDP spend by its Sub-Saharan African peers, and to 1.6 percent of GDP by the larger group of LMI countries. Table ES.1: International Comparison of Social Assistance Expenditures Public expenditures (% of GDP) Ghana (2014) 0.50 Average of lower-middle-income countries (35 countries) 1.60 Bolivia (2013) 1.36 India (2014) 0.72 Lesotho (2010) 6.60 Mauritania (2012) 1.30 Philippines (2013) 0.57 6 World Bank 2014e, 2015a. Scholarships are considered as part of social assistance according to international definitions and accounted in global averages. In Ghana, scholarships have not been considered as part of the 5 flagship programs in the recently approved National Social Protection Policy, as well as other programs considered in this study. Without scholarship, spending on social assistance was about 0.3 percent of GDP in 2014. 7 These data were compiled by the World Bank and presented in the State of Social Safety Nets 2015 (World Bank 2015a). xiii Swaziland (2010) 2.10 Vietnam (2010) 0.52 Zambia (2011) 0.50 Public expenditures (% of GDP) Ghana (2014) 0.50 Average of lower-middle-income countries (35 countries) 1.60 Average of Sub-Saharan African lower-middle-income 2.10 countries (5 countries) Lesotho (2010) 6.60 Swaziland (2010) 2.10 Bolivia (2013) 1.36 Mauritania (2012) 1.30 India (2014) 0.72 Philippines (2013) 0.57 Vietnam (2010) 0.52 Zambia (2011) 0.50 Source: World Bank (2015a, Annex D). The table refers to social assistance only (noncontributory benefits), not to spending on pension and labor market programs. Note: Excludes general subsidies. Data are for the most recent year available between 2009 and 2014. The process of public expenditure management in Ghana has improved in recent years but still presents some challenges. Public expenditure management is highly fragmented in Ghana, which makes it difficult to analyze the allocation of public resources. The central government budget covers the Consolidated Fund and the transfers from this fund to the earmarked funds. The five major earmarked funds (District Assemblies Common Fund, Ghana Education Trust Fund, National Health Insurance Fund, Petroleum Funds, and Road Fund) accounted for 21 percent of total central government noninterest current expenditures in 2014. The financial statements prepared by the Accountant General that show the executed budget includes the accounts of the Consolidated Fund and other funding sources. Similarly, the Ghana Auditor General (GAG) report on the Public Accounts of Ghana covers only the Consolidated Fund. The GAG prepares separate audit reports for the statutory funds and MDAs (Ministries, Departments, and Agencies) and MMDAs (Municipal, Metropolitan, and District Assemblies). There is no consolidated account of the public sector or of the general government. And there is no readily available information on budget outturns by sector, so it is very difficult to assess the impact of public expenditures. Fiscal decentralization has been under way for several years. Many responsibilities have been devolved to local governments, including welfare services; however, the degree of discretion granted to MMDAs by the central government differs. In some sectors, such as health and education, functions are deconcentrated and control remains in the center. In others areas, such as planning and construction of roads, parks, and cemeteries, functions have been fully devolved. Most MMDA resources continue to be under the control of the central government, such as the recruitment and compensation of local governments’ human resources. While the process of public expenditure management in Ghana has improved in recent years, major challenges remain in the areas of budget formulation, execution, and auditing. Actual revenues have been off target in recent years when compared with projections, which complicates budget monitoring and execution. Mid-year budget revisions lead to unanticipated cuts in some programs, with negative consequences for service delivery. Transfers to statutory funds— that is, funds established by a statute of the legislature—suffer significant delays, which in turn affects xiv budget execution by MDAs and MMDAs and payment to contractors and suppliers. Fiscal discipline in budget execution has been weak. This results from the unpredictability of funds and delays in the release of funds, as well as ineffective expenditure controls, reflected in overspending and in the accumulation of expenditure arrears. The implementation of social protection programs suffers from these same problems. Three major issues were reported by program managers interviewed: late disbursement of funds; incomplete disbursement of funds from the Ministry of Finance (MoF); and unpredictability of disbursement from statutory funds8. While the accounts of MDAs and MMDAs are audited every year, there is no follow-up with respect to the recommendations of Ghana Auditor General (GAG). Several initiatives are under way to improve budget formulation, management, and execution. In 2009, MoF launched the Ghana Integrated Financial Management Information System (GIFMIS) with the objective of improving budget comprehensiveness and transparency, consolidated cash management, monitoring and control of outstanding commitments, and payroll management. As part of the initiative, a new chart of accounts has been introduced and has been adopted by all MDAs and is progressively being adopted by MMDAs. In 2011, MoF created a Public Investment Division, which is seeking to establish a policy framework to guide public investment decisions. In 2014, MoF introduced program-based budgeting (PBB) to replace the existing practices of activity-based and line-item budgeting. In 2016, the MMDAs will also begin preparing their budgets on a PBB basis. The government is also implementing a fiscal decentralization action plan, which covers revenues and internally generated funds; intergovernmental transfers; borrowing; and budget formulation, execution, reporting, and external auditing. Conclusions and Recommendations To establish a more effective and sustainable social protection system, the report recommends addressing the following four challenges: consolidating the substantial gains achieved to date and building a stronger program portfolio; strengthening expenditure management; expanding well- targeted programs and rationalizing poorly targeted programs; and ensuring the sustainability of key programs. 1. Consolidating Gains and Building a Stronger Program Portfolio The social protection policy under development should set specific goals to be achieved and provide the elements for the development of a medium-term program and action plan to achieve the goals. Currently, there are no clear, rallying goal(s) for the social protection system that are also realistic and can be monitored. The goal(s) should be set with the participation of all key stakeholders. They could include, for example, MDG-type goals—such as eliminating extreme poverty or reducing the poverty rate by half—that could help mobilize Ghanaians and development partners to support the social protection programs, in a manner similar to the MDGs. The elimination of extreme poverty in Ghana would require a minimum investment of 0.5 percent of GDP on a well-targeted transfer program. This estimate is based on the GLSS-6 and the expenditures gap (which measures the distance between the average expenditure of the extreme poor to the extreme poverty line as a percentage of the extreme poverty line). The expenditure gap of the extreme poor is 27 percent, or GH¢215. Therefore, if the expenditure 8 The IMF program has defined strategies for clearance of arrears of statutory funds. xv capacity of the extreme poor was increased by the estimated expenditure gap, the incidence of extreme poverty would be eliminated. A transfer of GH¢215 to all the extreme poor would cost GH¢479 million, which is equivalent to 0.5 percent of GDP in 2013. It is important to note that this assumes that the transfer is perfectly targeted to the extreme poor households and the transfer is tailored to each household expenditure gap. A transfer of GH¢215 across the board to those in extreme poverty would still leave those with expenditure gaps below the average in extreme poverty. Only an across-the- board transfer of GH¢754 would ensure that all extreme poor would exit extreme poverty; this amount is equivalent to 1.8 percent of GDP. But in this case, this redistribution of resources would also reduce overall poverty by 20 percent and the poverty gap by 56 percent. While setting specific goal(s) should help galvanize support for social protection, the social protection policy should also seek to achieve broader objectives to consolidate an effective social protection system in Ghana. As mentioned, the broader objective of the social protection system is to minimize the impact of risks and vulnerabilities facing Ghanaians using a “portfolio of coherence programs that work together to provide for increase resilience, equity and opportunity.” The reduction in poverty is related to equity, but the social protection system should also focus on increasing resilience and opportunity. Table ES.2 summarizes the strengths of and issues with the existing portfolio of programs and presents recommendations to strengthen it. Table ES.2: Issues to Be Addressed in the Social Protection Programs Social Protection Function Strengths/Issues Recommendationsa Programs Main/Secondary Social Assistance Unconditional cash Equity Support to extreme poor elderly Increase coverage and keep benefits transfer (LEAP) poor, OVC, chronic ill, pregnant levels in real terms (short-term, ST). women. Target all families in extreme Issues: Low coverage, low poverty (medium-term, MT). generosity to ensure adequate minimum security. In-kind transfers Equity/opportunity Helps increase and maintain Keep expenditures under control and (school feeding) enrollments. payments up to date (ST). Issues: Payment delays to providers. Cash and in-kind Equity/opportunity Help keep students in class. Target uniform and textbooks to transfer (capitation, Issues: Leakages, payment arrears, poor and vulnerable children (ST). uniforms, and lack of targeting. textbooks) Waivers/NHIS Equity/opportunity Free access to health care for the Target exemptions only for the poor exemption poor. and vulnerable (MT). Establish a Issues: Low coverage of the poor, mechanism to ensure automatic low rate of renewals, lack of renewals of LEAP beneficiaries targeting. (ST). Public works/LIPW Equity/resilience/ Insure against drop in income, Continue to expand the program to opportunity involve community in selection of the most food vulnerable districts workers. (ST) Issues: Low coverage; Keep compensation at a level to oversubscription in work sites; ensure some degree of self-selection working conditions of women with (ST). children; fully financed from Ensure appropriate working external sources. conditions of women with children (ST). Link with skills/income- producing opportunities (pilot underway) (MT). xvi Start allocating local resources to the program (ST). Scholarships/Subsidies Opportunity/equity Promote human capital SS should publish and disseminate to SHS students accumulation. regulations of different programs Issues: poor selection mechanisms, (ST). lack of targeting, payment arrears, Target awards to the poor (ST). lack of monitoring. Potential Monitor beneficiaries of tertiary duplication and overlapping of scholarships, ensure relevance for existing programs. the country of tertiary scholarships, and revise bonding arrangements (MT). Rationalize all scholarships programs (MT). Social Insurance Resilience Insure against drop in income during Continue to increase social security old age coverage and adopt reforms to make Issues: Low coverage, future it sustainable (MT). benefits too generous, not sustainable. Active Labor Market Opportunity Enhance youth employability and Establish labor market information Programs workers’ skills. system, job search assistance, career Issues: Lack of labor market counseling, evaluate youth information services, little attention programs, and consolidate youth to basic skill formation and second programs (MT). chance education programs, overlapping and duplication of youth programs, no impact assessment of programs. Note: LIPW=Labor Intensive Public Works program; NHIS=National Health Insurance Scheme; OVC=orphans and vulnerable children; SHS=senior high school; SS=Scholarship Secretariat. a. Timeline of implementation: ST=short term (within one year); MT=medium term (within three to five years). The social protection medium-term program and action plan should be supported by a budget and monitoring and evaluation (M&E) framework. Once clear goals for the social protection have been established, policies and programs should be geared toward achieving those goals. The medium-term action plan should be costed out, and an annual operating plans and budgets should be prepared. For the M&E framework, a few indicators should be selected to ensure that program inputs are being transformed in the desired outputs, and the outputs transformed into the desirable outcomes. For a selected number of programs, process evaluation, targeting accuracy evaluation, public expenditure tracking survey (PETS), or beneficiary satisfaction surveys should be conducted, when considered cost-effective. Similarly, as already contemplated, program impact evaluation should be conducted for a selected number of programs, particularly those that can inform policy and program design. The social protection policy under preparation should clearly define the scope of social protection in Ghana and the core social protection programs it comprises. The MoGCSP should coordinate budget formulation, execution, and evaluation of these programs with the MDAs responsible for their implementation and the MoF. The program-based budgeting (PBB) being implemented by MoF should explicitly present these core social protection programs, and GIFMIS should include a functional code for social protection that will aid in appropriately tracking social protection spending. As MoGCSP implements the new governance framework, it is important to stress the need to substantially strengthen the District Social Development Departments (DSDD), as they are in the front line of social welfare services delivery. As the new framework is implemented, xvii there is a need to revalue the work of the district social workers (DSDOs) through training and improvements in working conditions. On the other hand, the DSDOs must become proactive in monitoring vulnerable groups in the community, identifying any family issues in a timely manner, and linking families to existing programs and services. Closer links between social services (referrals, integrated case management, and sensitization) and health, nutrition, and education programs, as well as with skill development programs, could help address some of the existing family challenges and provide opportunities for beneficiaries to “leap” from poverty. Referral protocols and cooperation agreements between institutions should be established. Training program for DSDOs should be developed and implemented. 2. Strengthening Management of Expenditures Good practice suggests to be conservative projecting revenues and realistic projecting the expenditures in budget preparation, while earmarking revenues should be avoided. Revenue projections have been quite off target in recent years, affecting budget preparation and budget execution. While a Medium-Term Expenditure Framework (MTEF) has been formally in place since early 2000, it has not contributed much to further fiscal discipline. Budget formulation is complicated by the existence of five major statutory funds, which receive earmarked funds. Earmarking treasury funds reduces the flexibility of fiscal management and makes program-based budgeting (PBB) less effective, and should be avoided, to the extent possible. All major social assistance programs should be shown explicitly in the budget. The program based budgeting being implemented by the Ministry of Finance represents a first important step for coordinating the financing and accountability of social protection programs. However, the introduction of the PBB has not made it possible to identify the major social assistance programs in the budget nor can GIFMIS currently provide program information. MoGCSP should work with the MoF to ensure that all social assistance programs are explicitly presented in the budget, including their recurrent expenditures, transfer or payments, and corresponding financing sources. This will make it possible to build a social protection budget to keep track of the amount of resources being allocated to the sector, their composition, and sources of financing. It will also facilitate the coordination and monitoring of the execution of sector expenditures. Ongoing reforms to improve public sector financial management should improve the execution of social protection programs. The implementation of social protection programs suffers from the same general weakness of overall budget implementation: late disbursement of funds; incomplete disbursement of funds from the MoF; and unpredictability of disbursement from statutory funds. The full implementation of GIFMIS, the adoption of the new harmonized chart of accounts, and the establishment of a unified treasury account should help with budget execution. Budget sector outturns should be prepared, published, and evaluated annually. The adoption of program-based budgeting should help the reporting and evaluation of public expenditure outcomes. This should include regular public reports on MDA budget outturns, using economic and functional classifications, and covering the different sources of financing (consolidated fund, statutory funds, internal generated funds, development partners). It should include reports on actual expenditure of MMDAs and source of finance. And it should include spending on social assistance programs and sources of financing. The functional responsibilities of different levels of government should be better defined and intergovernmental financial transfers should be geared toward providing MMDAs with the xviii resources to effectively discharge their responsibilities. Social welfare and community development functions, among others, have been devolved to MMDAs, but the MoGCSP operational departments (such as Social Welfare/Social Development) still discharge some related functions. At the same time, the “center” still controls “hiring and firing” of MMDA staff and payment of salaries. New legislation (L1971) has been reviewed to support the retooling of function of MMDAs. There is now a need to expedite efforts to provide MMDAs with the human and budget resources to discharge these functions in an effective way. Personnel decision should be transferred to MMDAs. District Assembly Consolidated Fund resources should be transferred directly to the District Assemblies (DAs). The Audit Report Implementation Committees (ARICs) in MDAs and MMDAs should be made responsible for ensuring that the recommendations of the Ghana Auditor General (GAG) are implemented by MMDAs. Recommendations to correct the problems identified in the audits are included in GAG’s report to the Parliament. There has been no follow-up on the report’s recommendations in the MDAs and MMDAs, , and no accountability. The same problems keep on reoccurring year after year: (i) misappropriation of funds; (ii) fraudulent payroll deals; (iii) lack of adherence to existing public procurement laws and process; (iv) irregularity in pension schemes; (v) pure embezzlement and theft; and (vi) inappropriate document and data storage procedures and poor record keeping. The Audit Service Act of 2000 establishes Audit Report Implementation Committees (ARICs) in MDAs and MMDAs to implement the GAG’s recommendations. However, these committees have not been functioning appropriately. The GAG is issuing new guidelines to address this problem. Support should continue to be provided to implement the financial decentralization framework with strong local participation. A Training program on financial management for leaders and staff at the MMDAs should be considered priority. 3. Expanding Well-Targeted Programs and Rationalizing Poorly Targeted Programs Poor parents continue to have difficulties in accessing basic services and sending their children to school. There is thus a need for a program of cash transfers to the poorest families, such as LEAP, with the adequate coverage and generosity (benefit amount with respect to household income or total consumption) to help parents overcome some of the constraints they face. The expansion of LEAP to cover all extreme poor families appears feasible within the confines of the ongoing fiscal consolidation. The GoG is committed to using part of the resulting fiscal space created by the ongoing fiscal consolidation to expand well-targeted programs such as LEAP. This study indicates that the gradual expansion of LEAP to cover all extreme poor families within the next five years would cost only 0.12 percent of GDP in 2019—assuming that the current level of transfer is kept in real terms. A micro-simulation exercise using the GLSS-6 indicates this increase in LEAP coverage would have a visible impact, with the incidence of extreme poverty declining by 22 percent and the extreme poverty gap falling by 28 percent from their 2013 levels. Consideration should also be given to expanding LIPW and to promote the inclusion of the poor and vulnerable in productive activities as a medium-term strategy to keep beneficiaries out of extreme poverty. As mentioned, LIPW has shown positive results and serves an important safety net function in areas where seasonal unemployment is a major problem. However, there is xix a need to scale it up, invest domestic resources in the program (now fully donor funded) and to create income-generating opportunities that go beyond the life of the public works and generally to promote sustainable livelihood of social assistance beneficiaries who have the capacity to work and the potential to graduate. The pilot under preparation —Complementary Income Generating Skills—will provide LIPW and LEAP beneficiaries with skills to pursue income-generating activities, and is a welcome initiative that could help them eventually graduate from the program. The District Social Development Departments (DSDDs) need to be strengthened, as they are in the front line of delivering social welfare services and linking these services with health, nutrition, and labor market programs. DSDOs must become proactive in monitoring vulnerable groups in the community, identifying any family issues in a timely manner, and linking families to existing programs and services. Closer links between social services (referrals, integrated case management, and sensitization) and health, nutrition, and education programs, as well as skill development programs, could help address some of the existing family challenges and provide opportunities for beneficiaries to “leap” from poverty. Referral protocols and cooperation agreements between institutions should be established. Training program for DSDOs should be developed and implemented. The Scholarship Secretariat (SS) program should be restructured and placed on an equitable, transparent, and efficient footing. The potential inequity of scholarships was already flagged in the Development Education Plan 2010–20, which recommended: (i) rationalization of some boarding facilities, except for those from disadvantaged areas and for girls; (ii) means testing and cost-recovery from parents and guardians (especially those who can afford to pay); (iii) grants and/or loans for poor students; (iv) provision, where feasible, of free transport for day-students living within 20 km of schools; and (v) student self-help schemes (cooking and cleaning).9 These recommendations are still valid and should be implemented. More generally, the GoG should address the overlaps and poor targeting of scholarships and student subsidies to senior high school. The recently released poverty maps10 should be used for further analysis to improve the targeting of secondary scholarships. Scholarships at the tertiary level should be based on the requirements of the economy to fill existing skill gaps. These scholarships should be awarded in support of a national human resources development policy. Indeed, the GSGDA II (2014–2017) recognizes the need to develop such a policy. All beneficiaries of the scholarships should be required to enter into effective bond agreements, which would ensure that the country is able to recover the investment made. Student loans with appropriate guarantees could replace scholarships for most students. Effective labor intermediation services in Ghana should be created. These services include: (i) labor market information; (ii) job search assistance; (iii) career counseling; and (iv) special programs for the unemployed, including second chance education programs. The World Bank and the ILO plan to help Ghana’s Department of Labor set up a modern labor market information system that can meet the needs of Ghanaian firms and job seekers. Job search assistance and career counseling services should be built up gradually. 9 Ministry of Education 2010, 30. 10 Ghana Statistical Services 2015. xx There is significant overlap and duplication among youth programs, and these program should be consolidated and rationalized. Despite the multitude of youth programs, it is not clear what works in Ghana because programs have not been evaluated for their impact. The only exception we are aware of is the National Apprenticeship Program (NAP), which is currently being evaluated. The rationalization of existing programs should take into consideration the existing large informal apprenticeship system, which is a specific feature of the Ghanaian economy. However, youth apprenticeship and entrepreneurship programs should not be geared toward providing youth with low-productivity jobs in the informal sector. In Ghana, most youth programs appear geared toward the creation of microentrepreneurs, without the proper evaluation of the capabilities of individual young people or the market needs. 4. Ensuring the Sustainability of Existing Programs Better targeting, increased cost-efficiency, and reduced dependence on external financing would help increase the sustainability of safety net programs. While during the fiscal consolidation period, the GoG will continue to require the support of development partners to finance the operations and expansion of key social assistance programs, provisions must be made for the future reduction in their dependence on external financial. The creation of a legal basis for key social protection programs may also contribute to their sustainability. The GoG should start allocating budget resources to finance the LIPW program. As discussed, LIPW is currently fully funded by external sources. To ensure its sustainability over the longer term, external resources should be replaced gradually by domestic funding. Continuation of efforts to reduce poorly targeted energy subsidies should help create fiscal space for strengthening social protection programs. Spending on electricity and fuel subsidies absorb more than twice the amount of public resources allocated to social assistance programs. Reducing these subsidies will make it possible to reallocate some of these resources to well- targeted social assistance programs. A strategy and action plan should be developed to implement the reforms required to ensure the sustainability of the NHIS. Adverse selection, lack of restrictions on demand, lack of copayments, and cost escalation are all characteristics of the current system that make it financially unsustainable. These problems and reform options have been well studied and discussed at length during the celebrations of the ten-year anniversary of the NHIS. Among these reform options are: cost containment, additional funding, intensify clinical audits, increased coverage of the poor, improved computerization of operations, and shortened claims processing and payment time. Recent studies11 have also clearly laid out additional reforms options, including redefining the benefit package (core services package and other services) and introducing copayments for the nonpoor and noncore services. Similarly, SSNIT requires additional reform to ensure its longer-term sustainability. While out of the scope of this study, a previous World Bank evaluation of the pension system suggests that following reforms should be considered: (i) establishing a linear accrual rate, along with other parametric reform measures; (ii) gradually increasing the age for eligibility for normal retirement 11 For example, Bitran 2014. xxi benefits; (iii) automatically indexing pensions to the consumer price index and wage inflation; and (iv) extending the reference period for benefits.12 12 World Bank 2013a, vii. ii I: Context and Purpose This introductory chapter briefly describes the context of this assessment and expenditure review, including the current challenges facing the government of Ghana, and the purpose and organization of the report. A. Context Ghana’s development success during the last 30 years has been accompanied by challenges recently. Ghana met the Millennium Development Goal of halving poverty and became a lower- middle-income country (LMIC) in mid-2011. It began commercial oil production the same year. As a consequence, it will have to rely more on domestic resources to finance its development in the future. This new environment requires Ghana to address four challenges. First, the government of Ghana (GoG) must reduce existing macroeconomic imbalances, which threaten continue growth and poverty reduction, while protecting the most vulnerable groups. Second, it must improve public expenditure management to avoid future fiscal crises. Third, it must consolidate development gains and deepen structural change and private sector development. Fourth, it must make further progress in reducing poverty and regional disparities and increasing social inclusion. These challenges are reflected in the challenges facing the social protection sector discussed in this report. For most of the last two decades, Ghana’s economy has been on a high growth path, which has been accompanied by significant poverty reduction. Per capita growth rates averaged more than 2.5 percent between 1983 and 2006, and increased to around 6 percent between 2006 and 2011.13 The only breaks in the rising growth record since 2000 occurred in 2009, as the global financial and economic crisis hit most countries hard, and in 2012, when growth reverted to its long-term trend after the start of commercial oil production. Poverty declined from 52 percent in 1991 to 21 percent in 2012.14 Ghana’s poverty reduction over the last two decades has been driven by rural development and increasing urbanization. Small-scale farmers, especially cash crop growers, benefitted from increasing agricultural productivity (notably in cocoa) and rising incomes and consumption, while in rapidly growing urban areas characterized by growing service sectors, large numbers of the labor force, including migrants from rural areas, were absorbed, mostly in informal economy. Ghana’s social indicators compare well with its LMIC African peers; indeed, Ghana has the best Human Development Index (Table 1.1). This rapid growth has been accompanied by growing inequality in terms of income and geographical areas, however. The Gini index deteriorated from 37.5 in 1991 to 40.8 in 2012, about a 9 percent increase. The big jump in the Gini occurred between 1998 and 2005; it was around two points.15 The poverty rate in the northern part of the country has declined much less than in the rest of the country, largely reflecting the region’s much higher rate of subsistence farming and much lower rate of urbanization. In 2012, more than one-third of national poverty was 13 World Development Indicators. 14 For the sake of comparability, the poverty headcount in 2012 was calculated using the 1999 poverty line. Using the official new poverty line the poverty rate at national level in 2012 would be 24 percent. See World Bank 2015b. 15 GSS 2014a, 2014c; World Bank 2015b. 1 concentrated in the three Northern Regions, which have only 17 percent of the population. The disparities in opportunities to access basic health and education services between north and south and between income quintiles remain large. A significant proportion of the population remains vulnerable to falling back into poverty in the event of major shocks. Table 1.1: Selected Social Indicators Ghana Cameron Congo, Côte Mauritania Nigeria Senegal Zambia Rep. d’Ivoire Population (million) 26 22 4 20 4 174 14 15 GDP per capita (US$, 1,850 1,316 3,172 1,521 1,070 3,006 2,212 1,540 2013) Life expectancy (years) 61 55 58 50 61 52 63 57 Maternal mortality (modeled per 100,000 380 590 426 720 320 – 320 280 live births) Under-five mortality rate (per 1,000 live births) 78 95 49 100 90 117 55 87 Adult literacy (15 years and above) 72 71 – 41 – 51 52 61 Human Development 138 152 140 171 161 152 163 141 Index (2013) a Sources: World Development Indicators (http://databank.worldbank.org/data/databases.aspx) and UNDP (http://hdr.undp.org/en/data). Note: Most recent available information, mostly for 2013. – = not available. a. Out of 187 countries. The macroeconomic situation presents important challenges. It is characterized by a large fiscal deficit, lower growth, and higher inflation. The GoG has been addressing these imbalances, but as of late 2014, significant imbalances remained, including a fiscal (cash) deficit of about 9.5 percent of GDP, driven in part by the high public wage bill, interest payments, and energy subsidies. Progress in addressing these vulnerabilities is crucial, as macroeconomic instability compromises the capacity to deliver basic services, affects the poor who are least able to hedge against its adverse impact, and undermines investment and growth. The 2015 budget adopted by the Parliament includes a strong set of measures to begin reducing existing imbalances. The fiscal consolidation effort includes the reduction of energy subsidies, a new tax on petroleum products, and stronger containment of the wage bill. In early April 2015, the Board of the International Monetary Fund (IMF) approved a three-year arrangement under the Extended Credit Facility to supports the GoG’s fiscal consolidation and public sector management reform program.16 The World Bank and other development partners (DP) are also supporting the program. B. Purpose The government of Ghana has a comprehensive National Social Protection Strategy (NSPS) in place. The 2012 draft NSPS represents the government’s vision of creating an inclusive society by providing sustainable mechanisms to protect persons living in conditions of extreme poverty, vulnerability, and exclusion. The NSPS establishes a framework that draws on the 1992 constitution, on international conventions, and on Ghana’s Shared Growth and Development 16 http://www.imf.org/external/np/sec/pr/2015/pr15159.htm. 2 Agenda (GSGDA).17 It seeks to reduce the population’s exposure and increase its resilience to risks, shocks, and exclusion, and provides a social protection floor across the life cycle, particularly for the extreme poor and most vulnerable. The NSPS has five strategic objectives: (i) promote sustainable mechanisms for social protection; (ii) provide a framework to strengthen stakeholder collaboration and coordination of social protection; (iii) harness and promote positive cultural values of social protection; (iv) develop sustainable funding instruments; and (v) institutionalize monitoring, evaluation, and reporting mechanisms across sectors.18 It considers several areas of intervention, including cash transfers, public works, pensions, social insurance, subsidies, public health programs, subsidies, welfare services, and skills and training programs. It further contemplates the development and implementation of efficient institutions arrangements and targeting, information management, and monitoring and evaluation (M&E) instruments. Given existing macroeconomic challenges and the need for fiscal consolidation, there is a need to ensure that resources allocated to social protection programs are used in the most cost-effective and pro-poor manner. There is a need to target social protection programs better and expand coverage of well-targeted programs to help mitigate the potentially negative impact of fiscal consolidation on the poor. A social protection policy is under preparation at the time of writing this report; this assessment is meant to serve as an input to that effort. The purpose of this assessment is to contribute to the rationalization of social protection public expenditures building on recent work by the International Labor Office (GOG 2014b).19 This assessment updates the analysis of the risks and vulnerabilities facing Ghanaians based on the recent (2012/13) household survey (GLSS-6); assesses to what extent the existing social protection programs and expenditure allocations contribute to address those risks and vulnerabilities; and suggests areas to strengthen the social protection system. More specifically, the assessment will complement and further develop the analysis undertaken in the recent GOG 2014b study on rationalizing social protection expenditures by: (i) Providing a more in-depth profiling of the population risks/vulnerabilities and social protection needs, based on a life cycle approach. (ii) Updating the State of Social Protection (part 1 of the GoG 2014b report) with the most recent program data on spending and number of beneficiaries. (iii) Identifying program gaps and duplications by mapping existing social protection programs to address risk and vulnerability for the different population groups. (iv) Assessing program effectiveness analysis using the GLSS-6 and administrative data. The following indicators have been used to measure program effectiveness for selected programs: program coverage and overlaps, adequacy of benefits with respect to consumption per capita, targeting efficiency (incidence), cost-effectiveness, and management tools. (v) Conducting a deeper analysis of public expenditure management issues and the sustainability of existing programs. This discussion includes sources of funding for social protection (such as the Consolidated Fund, Statutory Funds, and external funds). 17 NDPC 2010a, 2010b. 18 MESW 2012a, 32. 19 GoG 2014b. 3 (vi) Providing specific recommendations to strengthen the social protection system at the systemic and program level. The scope of social protection—the sample of programs analyzed in this report—is the same as in the GOG 2014b report, with minor adjustments. The assessment updates the analysis on core social assistance programs, active labor market programs, social insurance schemes including the National Health Insurance,20 and general fuel and electricity subsidies. The Ghana- Luxembourg Social Trust (GLST) is not covered here because it terminated in mid-2014.21 Likewise, the Social Inclusion Transfer (SIT), a cash transfer program supported by the African Development Bank, is not covered, as it ended in 2012. On the other hand, the review includes scholarships programs that were covered only partially in the GoG 2014b report, and the National Apprenticeship Program, which is a new program. Although scholarships have not been considered as part of social protection in previous studies for Ghana, they are used as social protection instruments around the world to help poor households access education facilities. They can be cash to families and/or in the form of waivers of educational fees or vouchers that reimburse households or schools/facilities for actual expenditure. The assessment is expected to contribute to the development of the social protection policy under preparation. The analysis of risks and groups at risk and of the existing portfolio of programs to address these risks, as well as of the analysis of the allocation and management of public resources to social protection and energy subsidies, aims to provide some guidance to establish policy priorities. This assessment is organized as follows. Chapter II briefly reviews risk and vulnerabilities facing Ghanaians, with particular attention to the most poor and vulnerable groups and regions. Chapter III reviews the basic features of core social protection programs now being executed to address those risks and vulnerabilities. Chapter IV draws on the previous two chapters to assess existing social protection programs for their coverage (including gaps and overlaps), generosity,22 targeting accuracy, cost-efficiency, sustainability, monitoring and evaluation, and coordinating mechanisms and institutional arrangements. Chapter V reviews social protection expenditures in the context of overall public expenditure management. Chapter VI offers recommendations on how to improve the effectiveness of the social protection system, which may be of help in preparing the social protection policy. 20 While there are other interventions that support the health delivery system, the report will focus on the NHIS. 21 It was a conditional cash transfer initiated in 2009 that targeted poor pregnant women and children. It was implemented by ILO and supported by the Luxembourg trade unions. 22 Generosity is defined as the level of a program benefit as a share of the poverty line or other type of indicator, such as the minimum wage, the average wage, or the total consumption of beneficiary households . 4 II: Analysis of Risks and Vulnerabilities This chapter identifies the major sources of vulnerability for Ghana households, the most vulnerable groups (the potential target groups of social protection programs), and the principal risks they face. It focuses on the major risks facing individuals and families during the life cycle that, if not addressed, can perpetuate the intergenerational transmission of poverty. A. Vulnerability and Poverty 1. Sources of Vulnerability The poor and vulnerable in Ghana face a series of important risks. These risks result from: (i) economic shocks and consequent variations in income and consumption; (ii) change in labor demand; (iii) man-made and natural disasters, mostly drought, floods, fire outbreaks, and accidents; and (iv) household conditions that expose the poorest families to a series of adverse situations and make them vulnerable. The role of the social protection system is to minimize the impact of these risks, using a “portfolios of coherent programs” that work together to deliver resilience, equity, and opportunity.23 Most working-age people in Ghana work—but in low-productivity sectors and in jobs that offer limited job and income security. Most jobs are in low-productivity self-employment activities that generate little earnings. In general, employment in Ghana takes place outside of the labor market, as most people work for themselves or their families. For their primary jobs, four out of five workers work in farm and off-farm household enterprises, and only one in five is a wage worker (and thus actually employed and receiving a wage from a firm or person).24 The majority of these informal workers lack social security coverage (paid holidays, pension contributions, and sick leave), making them vulnerable to such shocks if they occur. The recent global financial crisis severely affected the poor in Ghana. An evaluation of the impact of the crisis indicates that in its wake, it is very important to maintain agricultural productivity, provide for income-earning opportunities for the poor in rural Ghana, and protect real incomes of wage earners.25 It is also very important to promote pro-poor, broad-based agricultural growth and the expansion of social safety nets, so the country is prepared to meet similar crises in the future. To diminish the risks originating from external economic shocks and cyclical variations in economic activity, it is important to pursue prudent macroeconomic policies. Maintaining prudent fiscal and monetary policies will help contain inflationary pressures, as high inflation is one of the most damaging burdens that can be imposed on the poor population, particularly those living on a fixed income and/or earning subsistence wages. Maintaining a solid financial system is crucial to ensure that depositors’ funds are preserved and credit and other services are available to firms and families, particularly during difficult economic periods. 23 The World Bank’s Social Protection and Labor Strategy 2012 –2022 (World Bank 2012b, xiv). 24 World Bank 2016a. 25 World Bank 2010, 2. 5 Ghana has taken important steps to lessen the risks associated with natural disasters by putting in place institutional coordination and capacity through a responsive social protection system. Social protection programs can not only mitigate the effects of a crisis after it has hit, but are also capable of playing a preventive role in advance of a crisis. Ghana is using a multisectoral approach to risk management. In 1996, Ghana created the National Disaster Management Organization (NADMO) with the mission to “manage disasters by co-ordinating the resources of government institutions and nongovernmental agencies, and developing the capacity of communities to respond effectively to disasters and improve their livelihood through social mobilization, employment generation and poverty reduction projects.” 26 Ghana has also been taking steps to implement activities under the Hyogo Framework of Action to develop institutional capacity and identify at-risk areas and populations. NADMO functions under a national secretariat; ten regional secretariats; 243 municipal, metropolitan and district secretariats; and more than 900 zonal offices throughout the country. NADMO relies on GoG support, donor technical and financial assistance, and social mobilization of civil society to provide relief and to build the capacity of society at large to prevent and manage disasters. The organization also collaborates with its partners in reconstruction following a disaster.27 The northern regions of the country is prone to droughts and floods and food insecurity. The World Food Programme (WFP) Comprehensive Food Security and Vulnerability Assessment study conducted in 2009 revealed that about 5 percent of the Ghanaian population, or 1.2 million people, could be classified as chronically hungry. In addition, about 2 million people, or nearly 10 percent of the population, were at risk of hunger in the event of shocks. Contributory factors include prolonged dry seasons, which lead to recurrent serious seasonal food insecurity, and inadequate food storage facilities. The report, which focused on the northern part of the country, also noted serious regional disparities in food insecurity, where 34 percent of the population in the Upper West Region, 15 percent in the Upper East, and 10 percent in the Northern Region, were food insecure: totaling about 0.5 million people. The lowest prevalence of food insecurity was found in Accra (2 percent) and the rural areas in Greater Accra (1 percent) and Western region (1 percent).28 The vulnerability of northern Ghana can be traced to several factors. The region is landlocked and has a poorly developed transport infrastructure, which limits access to markets and social services. In comparison with the South, it has less rainfall, greater land and soil degradation, and a predisposition to droughts and floods. This forces agricultural households to adopt low-risk and low-input strategies, creating a virtual cycle of poverty. Many households engage in seasonal migration to the South. While labor migration as a whole has helped reduce poverty substantially in Ghana, the North has benefited to a lesser extent from remittances than other parts of the country. Northern migrants tend to have lower skills, which relegates them to the lowest wage segments of labor markets. To bridge the North-South divide, the Savannah Accelerated Development Strategy (SADS) and the Savannah Accelerated Development Authority (SADA) were established. The SADS specifies a two-track strategic framework to address regional disparities between the North and the South: (i) igniting and sustaining market and private sector growth—particularly in the agriculture, tourism, and mining sectors—and developing nonfarm 26 http://nadmo.gov.gh/. 27 MESW 2012a, 75. 28 WFP 2009, 14. 6 enterprises; and (ii) strengthening food security and social protection among the poor and vulnerable segments of the population.29 The Global Facility for Disaster Risk Reduction (GFDRR), a global partnership managed by the World Bank, is supporting the implementation of the Disaster Risk Management (DRM) Country Plan, which was jointly drafted by the GoG and the GFDRR. The program is being implemented in several phases. The White Volta Flood Hazard Assessment constituted the first phase. The second phase is a US$1.3 million grant to strengthen the institutional capacity of the agencies responsible for flood and disaster risk management.30 Through the Labor Intensive Public Works program (LIPW), the government, with World Bank support, is also investing on small climate change public works and afforestation projects on degraded lands. Household vulnerabilities are related to the conditions in which households live.31 These vulnerabilities are usually associated with a lack of skills, assets, employment, income-generating opportunities, sociocultural risks, income insecurity in old age, and a lack of access to basic services and nutritious food. They are also related to specific conditions that can affect groups such as the elderly; orphaned, abandoned, and abused children and child laborers; the physically disabled; and the chronically ill, including those living with HIV/AIDS. Existing vulnerabilities can magnify the effects of economic shocks or natural disasters. For example, if a youth has no skills, he or she may find it difficult to secure employment even in good economic times, but in bad economic times, even more better-qualified job seekers will be competing for the same few jobs. On the other hand, the coping strategies that individuals and families tend to use can make them less vulnerable to risks associated with changes in economic activity or natural disasters. For example, a member of a farming families may get a job in a nonagricultural sector (diversification of production), migrate to urban areas, or participate in a training course (investment in human capital). The family may open a savings account (diversification of assets) or contribute to a pension (insurance). These are self-insurance, self-protection, and market insurance strategies at the individual level that help households confront the kind of risks that can affect entire communities or even the entire country. These household vulnerabilities that affect poor families are the main focus of this report. 2. Poverty Ghana has achieved Millennium Development Goal 1 of halving the poverty rate; poverty dropped from 52 percent in 1991 to 21 percent in 2012.32 Since 2005, the incidence of poverty and extreme poverty declined by 7.7 and 8.1 percentage points, respectively (Table 2.1).33 The reduction in poverty can be traced to output growth and pro-poor interventions, as inequality—as measured by the Gini coefficient—has increased from 1991 to 2005 and then remained at similar levels until 2012. 29 World Bank 2010, 2. 30 World Bank Project Ghana-Disaster Risk Management Country Plan (P144828), 2014. 31 Risks can be classified as idiosyncratic, when they affect only one individual or household (such as the death of the breadwinner, old age, disability, or even sociocultural risks such as domestic violence) or covariate, when they affect several households or even the whole country (such as the recent oil price shock). 32 For the sake of comparability, the poverty headcount in 2012 was calculated using the 1999 poverty line. Using the official new poverty line the poverty rate at national level in 2012 would be 24.2 percent. See World Bank 2015b. 7 33 GSS 2014a, 2014c. Table 2.1: Incidence of Poverty and Inequality, 2005 and 2012 Poverty Extreme Poverty Gini Incidence (%) 2005 31.9 16.5 0.419 2012 24.2 8.4 0.423 Difference -7.7 -8.1 0.000 Poverty gap (%) 2005 11.0 5.0 2012 7.8 2.3 Difference -3.3 -2.7 People (million) 2005 7.0 3.6 2012 6.4 2.2 Difference -0.6 -1.4 Source: GSS 2014a, 2014 c. Note: The 2012 poverty and extreme poverty (adult equivalent) lines are GH¢1,314 and GH¢792 per year. The incidence of poverty in the Northern Region is two to three times the national average. While the incidence of poverty in Greater Accra is 5.6 percent, it is 71 percent in Upper West (Figure 2.1). In 2012, the rural population comprised 50 percent of the population of Ghana, yet it accounted for 78 percent of those in poverty. The poverty incidence in the rural savannah contributes more than 40 percent to the overall poverty in Ghana: more than in the rural coastal and forest combined.34 Figure 2.1: Poverty and Extreme Poverty by Region, 2012 80 Incidence of poverty and extreme poverty 70 60 50 40 30 (%) 20 10 0 Poverty Extreme poverty Source: GSS 2014a, 2014c. The poorest regions are also the more vulnerable to changes in economic conditions. The relatively heavy concentration of the population near the poverty line makes them vulnerable to 34 Ghana Statistical Service 2014c, 12. 8 changes in income, whether seasonal or annual. This can be seen in Figure 2.2, which shows the impact on the incidence of poverty of a 15 percent and 25 percent increase in the poverty line (equivalent to a reduction of the same magnitude in purchasing power of households). The impact is generally greater in the poorest regions. For example, an increase in the poverty line by 25 percent in rural areas would increase the incidence of poverty from 37.7 percent to 50.9 percent. In the Upper West Region, if the purchasing power dropped by 25 percent, poverty would increase from 70 percent to nearly 80 percent. Figure 2.2: Sensitivity of Incidence of Poverty to Changes in the Poverty Line, 2012 90 80 70 60 50 40 30 20 10 0 Poverty line, official Poverty line +15% Poverty line +25% Source: World Bank staff estimates based on GLSS-6. Note: Incidence of poverty with an increase in the poverty line of 15 percent and 25 percent. The main poverty covariates are employment status, education, ownership of assets, and access to basic services such as potable water, sanitation, and electricity.35 The poverty incidence among uneducated household heads is 37.6 percent, while it is 3 percent among those with tertiary education. The poverty incidence is highest among households where the head is self- employed in the agricultural sector (39.2 percent) or is unemployed (28.1 percent). Poverty incidence among male-headed households is higher (25.9 percent) than female-headed households (19.1 percent).36 This is similar to the pattern found in 2005 (GLSS-5 survey). More than 2.2 million Ghanaians live in extreme poverty (8.4 percent of the population): that is, they cannot afford to feed themselves, even if they were to spend all their expenditures on food.37 Extreme poverty is mostly a rural phenomenon, with as many as 1.8 million persons living in extreme poverty in rural areas. The incidence of extreme poverty is less prevalent in urban localities, with Greater Accra Metropolitan Area (GAMA) contributing only 0.9 percent to the incidence of extreme poverty. Urban localities contribute 11.2 percent to the national incidence of extreme poverty.38 35 GSS 2014a. 36 GSS 2014c, 21 and 22. 37 They are unable to purchase the minimum basket of food with 2,900 calories per adult equivalent per day. 38 GSS 2014c, 14. 9 Extreme poverty and poverty incidence rates are higher for children; more than 1.2 million children live in extreme poverty, and 3.4 million live in poverty. Children under 18 years of age have much higher rates of poverty and extreme poverty than the country average (Table 2.2). The elderly follow, in terms of higher poverty and extreme poverty incidence rates. Somewhat unexpected, youth aged 18 to 35 have lower incidence rates of poverty and extreme poverty than adults. Table 2.2: Incidence of Poverty by Age Group, 2012 Age (years) Extreme poor All poor (%) Number of Number of poor (%) extreme poor 0–17 (children) 9.9 28.3 1,211,013 3,462,715 18–35 (youth) 6.7 19.6 493,721 1,450,781 36–59 (adult) 7.2 20.8 370,167 1,079,348 60+ (elderly) 8.5 23.9 153,819 435,227 Total 8.4 24.2 2,228,720 6,428,071 Source: World Bank staff estimates based on GLSS-6. Access to basic services is much better in urban areas than in rural areas. Access to basic services is determined by both their availability and affordability. Table 2.3 indicates that in 2013, about 90 percent of all households used potable water, 45 percent had improved sanitation, and 71 percent had electricity. Access to these services is much lower in rural areas and for the poorest households. Table 2.3 Access to Basic Services, by Quintile, 2012 Ghana Q1 Q2 Q3 Q4 Q5 poorest richest Watera 90.0 79.3 83.9 89.3 93.0 96.0 Urban 98.2 92.7 96.6 97.7 98.8 99.3 Rural 74.2 76.2 70.6 72.2 73.8 79.1 Sanitationb 45.3 25.6 34.5 40.4 49.8 59.2 Urban 53.8 30.2 40.3 45.0 53.5 63.8 Rural 34.8 24.5 30.9 35.6 43.4 44.1 Electricityc 70.7 36.9 55.3 68.9 79.5 88.2 Urban 88.5 69.9 74.7 84.1 91.1 94.7 Rural 48.6 28.0 43.3 52.9 60.2 67.1 Source: GSS 2014c, based on GLSS-6. a. Water includes inside pipe, water vendor, neighbor/private; public standpipe; borehole, well. b. Improved sanitation includes flush, pit latrine, pan/bucket, and Kumasi Ventilated Improved Pit (KVIP). c. Electricity includes main grid and generator. B. Main Vulnerable Groups and Risks This section characterizes the main vulnerable groups in Ghana and the risks they face during the life cycle. The risks were identified on the basis of the analysis of the data from the latest Ghana living standards survey (GLSS-6), available studies, and the input of key stakeholders. The focus is on the existing constraints these groups face to access basic services — 10 particularly from the demand side, which can be relaxed through social protection interventions. The general age groups considered are: pregnant women and very young children (less than 3 years); school-age children (3–17 years); youth (18–35 years); working poor (36–59 years); the elderly poor (60 years and over); and poor people and families with disabilities and/or HIV/AIDS. Different age groups may be used depending on specific context and data availability. 1. Poor Pregnant Women and Young Children In the last several years, the poor have increased their use of health services, and health outcomes have improved in Ghana. These improvements have been attributed to increased incomes and the increased supply of services through the National Health Insurance Scheme (NHIS) and community-based health and nutrition programs.39 Despite significant progress, poor children—who are Ghana’s most vulnerable group—face important health risks. Their weight is low at birth and they experience debilitating and life-threatening diseases. Their diet is inadequate, and they lack early stimulation, which can impair their development and may perpetuate their poverty. The health of an infant depends on the health of the mother. To promote the birth of healthy babies, the health system should ensure, first, that reproductive health services are available so that only desired pregnancies take place and that pregnancies are sufficiently spaced. In Ghana, 26.4 percent of women who are married or in union have an unsatisfied need for contraception services (Table 2.4). Percentages vary widely by region, with unmet needs greatest in Volta (38 percent) and the Northern Region (26 to 30 percent) and lowest in Greater Accra (20 percent). Unmet needs are much more marked in rural areas and among the poorest. Second, pregnant women should receive early and regular prenatal care. The World Health Organization (WHO) recommends that all pregnant women have at least four prenatal care visits and be immunized against tetanus. In Ghana, 86.6 percent of pregnant women received in 2011 four prenatal care visits, ranging from a low of 75.1 percent in the Northern Region to a high of 93.2 percent in the Eastern Region. Again, rural women and the poorest women fared worse than those in urban areas and in the richest quintile. Third, all pregnant women should be protected against maternal and neonatal tetanus. In Ghana, only about two-thirds of women are protected, with significant differences between rural and urban areas and rich and poor groups. Fourth, the delivery of babies should be attended by skilled medical personnel. The average for Ghana is 68.4 percent, with a low of 38 percent in the Northern Region, where more than half the babies are delivered by a traditional birth attendants or a relative or a friend. In rural areas, only 53.9 percent of deliveries are by skilled personnel, compared to 88.2 percent in urban areas. By income group, the disparity is even greater: 38.6 percent for the poorest quintile (Q1), compared to 97.6 percent for richest quintile (Q5).40 One in four poor women did not use prenatal care in public clinics because they could not afford it. For another 31 percent, such care was too far away or was not available. The remaining 46 percent thought “it was not necessary” or had other reasons, data from GLSS-6 indicate. As discussed in Chapter III, pregnant women have access to free health care under the NHIS; therefore, these data suggest that additional efforts are required to reach poor pregnant women. 39 World Bank 2013b, 38. 40 MICS 2011. 11 Table 2.4: Reproductive Care Indicators (%) Unmet needs for Antenatal care Protected Delivery by any contraception 4 or more visits against tetanus skill personnel (women age 15–49 (women age (women age 15– (women age 15– years married or in 15–49 years 49 years who 49 years who union) who had a live had a live birth had a live birth birth in the last in the last 2 in the last 2 2 years) years) years) Ghana 26.4 86.6 70.3 68.4 Urban 24.4 94.1 77.5 88.2 Rural 28.4 81.2 65.0 53.9 Q1 (poorest) 32.5 74.3 59.2 38.6 Q5 (richest) 16.2 98.5 88.2 97.6 Western 25.1 78.5 65.6 64.5 Central 22.6 88.0 74.6 63.4 Greater Accra 19.7 91.9 82.1 89.7 Volta 37.9 82.6 67.3 64.4 Eastern 22.4 93.2 72.1 77.9 Ashanti 29.8 91.0 76.2 73.7 Brong Ahafo 27.3 83.9 59.9 63.7 Northern 30.4 75.1 61.3 37.7 Upper East 26.3 88.8 56.5 67.0 Upper West 28.0 90.3 60.3 60.4 Source: MICS 2011. After birth, breastfeeding should be initiated immediately, mother and baby should have a postnatal care visit, and other WHO recommendations should be followed for the healthy development of the child. In Ghana, 58.5 percent of the newborns did not have the recommended postnatal care visit; the frequency for women in the poorest and richest quintiles was similar, and the rate was slightly higher for rural women than urban women (Table 2.5). WHO recommends that breastfeeding should start within the first hour of birth and should continue exclusively during the first six months. In Ghana, only 45.7 percent of children are breastfed exclusively during the first six months. The prevalence is higher in rural areas and for the poorest mothers. Beyond six months, breastfeeding should be accompanied by consumption of nutritionally adequate, safe, and appropriate complementary foods. In Ghana, only 31 percent of children 6–23 months are fed following the three recommended practices (breast milk or milk products, four or more food groups,41 and the minimum number of times or more). The low adoption of the recommend practices is widespread in rural and urban areas and for the poor and rich, which indicates a generalized lack of nutrition education. Finally, to avoid diseases that typically affect children, a child should receive a Bacillus de Calmette y Guérin (BCG) vaccination to protect against tuberculosis; a dose of polio administered at birth and three doses later (OPV3); three doses of diphtheria, pertussis, and tetanus (DPT); and a measles vaccination by the age of 12 months. In Ghana, 77 percent of children aged 12–23 months are fully immunized before their first birthday. The coverage rate for children aged 12–23 months who received all immunisations is 84 percent; less than 1 percent of children did not receive any vaccination. 41 Grains, roots, and tubers; legumes and nuts; dairy products (milk, yogurt, cheese); fresh foods (meat, fish, poultry and liver/organ meats); eggs; vitamin-A rich fruits and vegetables; other fruits and vegetables. 12 Malaria is endemic and perennial in all parts of the country and mostly affects children and pregnant women. Ghana’s entire population is at risk of malaria infection, but children under five years of age and pregnant women are at higher risk of severe illness due to lower immunity. Malaria is the principal cause of deaths in children under five (34 percent of all deaths). There is a significant negative link between malaria and nutritional status. Malnutrition in children is highly prevalent in areas where there is a high incidence of malaria. Conversely, malaria negatively affects the nutritional status of children, as well as pregnant and lactating women. Malaria infection has been associated with anaemia, being underweight, and stunting, especially among children under two. The National Malaria Control Strategy (2008–15) recognizes the use of insecticide treated nets (ITN) as one of the most effective measures for preventing malaria.42 In Ghana, only about half of households have at least one long-lasting insecticidal net (LLIN); the frequency is higher in rural areas and poor income groups (Table 2.5). The U.K. Department for International Development (DFID) and UNICEF have been implementing an important LLIN program in Ghana. Table 2.5: Child Health Indicators (%) No postnatal Exclusive Child feeding with all Complete Households care visit breastfeeding recommended vaccine regime with at least (newborns) (first 6 practices (6–23 (12–24 months) one insect net months) months) (LLIN) Ghana 58.5 45.7 31.0 84.4 48.3 Urban 55.2 48.9 34.8 81.7 38.4 Rural 60.9 43.3 28.2 86.5 59.5 Q1 (poorest) 56.3 53.7 28.6 83.8 65.4 Q5 (richest) 58.5 47.2 38.4 84.5 34.1 Western 59.5 46.8a 15.3 91.4 42.2 Central 75.5 39.5 35.7 78.4 31.6 Greater Accra 57.1 21.1a 34.7 78.3 24.7 Volta 51.6 49.1a 37.9 71.8 85.3 Eastern 76.1 42.6 23.1 88.0 77.8 Ashanti 59.6 39.0 38.5 86.8 39.0 Brong Ahafo 28.8 69.8 40.7 94.9 52.4 Northern 58.4 63.6 21.3 80.2 67.0 Upper East 48.7 58.4a 31.6 94.6 51.4 Upper West 52.3 67.0 38.3 93.4 59.5 Source: MICS 2011. Note: LLIN = long-lasting insecticidal net; HH = households; LLIN = long-lasting insecticidal net. a. based on 25-49 weighted cases. These gaps in care and poor practices translate into high maternal mortality rates, into high neonatal and infant mortality rates, and into underweight newborns who are not able to resist diseases. While Ghana has made steady progress in reducing maternal, infant, and under- five mortality rates (Figure 2.3), these rates remain higher than would be expected for a nation of its income level.43 Neonatal mortality (during the first 28 days after live birth), in particular, is responsible for 60 percent and 40 percent of the infant mortality rate (before the first year of life) and under-five mortality rates, respectively (Table 2.6). Child mortality rates are higher in rural areas than in urban areas and vary significantly among regions, with rates generally higher in the 42 http://www.who.int/countryfocus/cooperation_strategy/ccsbrief_gha_en.pdf?ua=1; World Bank (2014d, 1). 43 Saleh 2013, 10. 13 northern regions than the rest of the country. The under-five mortality rate (per 1,000 live births) is 92 in rural areas, compared to 72 in urban areas; 106 for the poorest quintile, compared to 52 for the richest; and 128 in the Northern Region, compared to a national average of 82. Figure 2.3: Trends in Mortality Rates 180 700 160 600 140 per 100,000 live births per 1000 live births 500 120 100 400 80 300 60 200 40 100 20 0 0 1988 1990 1993 1995 1998 2000 2003 2005 2008 2010 Linear (Infant mortality (per 1000 live births) Linear (Under-5 mortality (per 1000 live births) Linear (M Mortality Ratio (per 100,000 live births), right ) Source: WHO and others (2012, 40); EDS (http://www.statcompiler.com/). Table 2.6: Low Birth Weight and Mortality Indicators, by Region Low birth weight Neonatal mortality Infant mortality Under five (percent) (per 1000 live (per 1000 live mortality (per births) births) 1000 live births) Ghana 10.7 32 53 82 Urban 10.9 30 46 72 Rural 10.5 33 56 94 Q1 (poorest) 11.6 35 61 106 Q5 richest 9.7 24 38 52 Western 10.7 27 50 67 Central 10.5 36 55 88 Greater Accra 9.9 20 37 56 Volta 10.5 47 68 89 Eastern 10.6 25 38 61 Ashanti 11.4 27 43 86 Brong Ahafo 8.2 44 66 104 Northern 11.9 39 66 124 Upper East 14.5 34 58 98 Upper West 9.5 41 67 108 Source: MICS 2011. Child malnutrition remains a major problem in Ghana. Chronic malnutrition (stunting) and anemia affect 22 percent and anemia affects 57 percent of children under 5, respectively.44 Anemia results mostly from a lack of iron-rich foods and can cause poor mental, socioemotional, motor, 44 Malnutrition is an issue among pregnant women as well. 14 and neurophysiological functioning. There is conclusive evidence that children with iron- deficiency anemia are at developmental risk.45 Ghana’s stunting rate compares well with the other lower-middle-income countries in Sub-Saharan Africa (Table 2.7), but it is still unacceptable high because the health and development of nearly one in four surviving children is compromised. The national stunting rate masks wide income and regional disparities. Chronic malnutrition for children in the poorest quintile (33.2 percent) is three times higher than for children in richest quintile (11.6 percent). Similarly, chronic malnutrition is much higher for children in the Northern, Upper East, and Upper West Regions (Figure 2.4). Table 2.7: Comparisons of Under-Five Nutrition Indicators, Selected Sub-Saharan African Countries Ghana Cameron Congo, Côte d’Ivoire Mauritania Nigeria Senegal Zambia Rep. Height for age 22.7 32.6 30.0 29.6 22.0 36.0 19.2 45.8 Weight for height 6.2 5.8 8.0 7.6 11.6 10.2 8.9 5.6 Weight for age 13.4 15.1 11.0 15.7 19.5 24.4 16.8 14.9 Source: http://www.data.unicef.org/nutrition/malnutrition; UNICEF 2013. Data for Nigeria from World Bank World Development Indicators 2011. Note: Latest available data. Height-for-age (chronic malnutrition or stunting) reflects children’s nutritional history. Weight-for-height (acute malnutrition or wasting) reflects their current nutrition status. Weight-for-age (underweight) reflects both the past and current nutritional status of children (and takes into account both chronic and acute malnutrition). Figure 2.4: Under-Five Nutrition Indicators by Region 40 35 30 % of age group 25 20 15 10 5 0 Height for age (chronic) Weight for height (acute) Weight for age (underweight) Source: MICS 2011. Demand and supply constraints underline these results. Data from GLSS-6 indicate that demand for health care is relatively high. About two-thirds (66.2 percent) of those who got ill or were injured in the previous two weeks of the survey consulted a health practitioner.46 Of those 45 Walker and others 2007, 147 and 148. 46 GSS 2014d, 25. 15 who consulted a health practitioner, about half paid nothing. Of those who paid, the average bill per patient for consultation and treatment was GH¢157, and for hospitalization was GH¢151.47 For the extreme poor family, the average payment per patient was GH¢94 for consultation and treatment and GH¢29 for hospitalization, which corresponds to two-thirds and one-fifth of the extreme poor family’s monthly food budget, respectively. If more than one member was sick during the period, the burden would increase proportionally. Out-of-pocket payments continue to be high. Of the total payments (consultation and hospitalization), in about 46 percent of the cases, the major part of the health bill was paid by households, while in 52 percent of the cases, it was paid by insurance. GLSS-6 did not ask the question why one-third of those who got ill or injured did not visit a health facility, but one can infer that affordability may have been one of the major reasons. Indeed, World Health Organization data indicate that out-of-pocket health payments in Ghana, after declining with the introduction of the National Health Insurance Scheme in 2003, have since increased, and have hovered around 28 percent of total health costs in recent years (Figure 2.5). This analysis suggests that a greater effort is required to cover all extreme poor families with health insurance. As discussed in Chapter III, the recent MoGCSP initiative to extend LEAP to poor pregnant or lactating women (LEAP 1000) seeks to relax some of the demand constraints facing women in accessing health care and contributes to other existing Ministry of Health interventions such as the supplementary feeding program, aimed at improving nutrition and access to health services. Table 2.8: Health Payments per Patient for Those Who Visited Health Facilities Extreme Ghana poor All poor Non-poor 1.Consultation (GH¢) 156.8 93.6 95.1 165.0 Registration 5.7 2.9 4.3 6.0 Consultation 15.5 6.3 21.6 14.8 Diagnosis (lab, etc.) 34.4 32.3 11.1 36.3 Drugs and treatment 28.7 17.1 17.3 30.5 Treatment or services received 46.5 29.5 30.4 49.7 Other payment 25.9 5.6 10.4 27.7 Total % of monthly expenditures 43.3 68.2 48.9 41.7 2.Hospitalization (GH¢) 150.9 28.9 75.7 165.9 Payment for staying in a hospital/health 129.8 14.9 61.6 143.5 Payment for medicines and medical supplies 21.2 14.0 14.1 22.4 Total % of monthly expenditures 41.7 21.1 39.0 42.0 3. Who pays for the largest portions of health expenses, incl. consultations and hospital stays (%) 100.0 100.0 100.0 100.0 Out-of-pocket 42.1 46.3 43.6 41.8 Health insurance 53.7 52.2 53.2 53.8 Government 0.6 0.3 0.1 0.8 Other 3.5 1.2 3.1 3.6 4. Memo: Monthly household expenditures (GH¢) 362 137 194 395 Source: World Bank staff estimates based on the GLSS-6. Note: Average for only those who pay something. There was not much difference between payments for consultation made in public or private facilities. 47 The Ghana Statistical Service (2014d, 28) estimates an average payment of GH¢347 for overall treatment fees. 16 Figure 2.5 Total Health Expenditures and Out-of-Pocket Expenditures 35 8.0 7.0 30 % of total expenditure 6.0 25 % of GDP 5.0 20 4.0 15 3.0 10 2.0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Out of pocket (% of total exp.) Total health expenditure/GDP (%) (right) Source: WHO (http://apps.who.int/nha/database/ViewData/Indicators/en). Several initiatives area under way to strengthen health service delivery. To improve service delivery at community level, the Ghana Health Service implemented the Nutrition and Malaria Control for the Child Survival Project from 2007 to 2013 with World Bank support (a US$25 million IDA credit). The project helped strengthen the Community-based Health Planning and Services program. It established and rolled out a strong community-based program to promote children’s growth across 77 districts in 5 of the country’s 10 regions.48 A follow-up Maternal and Child Health and Nutrition Improvement project (a US$68 million IDA credit and a US$5 million grant from the Multi-Donor Health Results Innovation Trust Fund) was recently approved, to be implemented from 2014 to 2020. Likewise, in the Upper West Region, where seasonal hunger is quite acute, UNICEF and Catholic Relief Services (CRS) launched the Integrated Nutrition Action against Malnutrition to fight severe acute malnutrition (SAM). Ghana’s private sector and the GoG, with the support of Micronutrient Initiative, have also embarked on the iodization of salt and fortification of wheat flour and vegetable oil to ensure that essential nutrients are present in the diets of children and women. The program aims to fortify vegetable oil with vitamin A, and wheat flour with vitamins A and B, iron, and folic acid.49 2. School-Age Children One in ten children lives in extreme poverty in Ghana. Virtually all these poor children (99.9 percent) live with families (Table 2.9). About 7.2 percent of the extreme poor children do not live with their parents or grandparents; 5.5 percent live with other relatives; and 1.7 percent are adopted, live with nonrelatives, or live alone. These 7.2 percent (about 87,190 extreme poor children) who live in poor households without their parents or grandparents may be the most 48 World Bank 2014a, 5. IDA is the International Development Association, the arm of the World Bank that helps the lowest income countries. 49 MESW 2012a, 54. 17 vulnerable group;50 they include many of the orphaned and vulnerable children (OVC) and child laborers who require special support. Table 2.9: Living Status of Children (0–17 years) (% distribution) Ghana Extreme All poor Nonpoor poor Live alone 0.05 0.04 0.02 0.06 Live with families 99.95 99.96 99.98 99.94 Of those who live with families: 100.0 100.0 100.0 100.0 Live with parents 78.8 80.9 80.2 78.2 Live with grandparents 12.2 11.9 12.6 12.0 Adopted /foster/step child 1.8 0.8 1.2 2.0 Live with nonrelatives 0.8 0.5 0.4 0.9 Live with other relatives 5.9 5.5 5.2 6.2 Head of household 0.1 0.2 0.1 0.1 Other 0.6 0.2 0.3 0.6 Memo: No children 12,228,295 1,211,013 3,462,715 8,765,580 Percent 100.0 9.9 28.3 71.7 Source: World Bank staff estimates based on GLSS-6. Eight percent of children aged 0–17 years (980,000 children) in Ghana are orphans of one or both parents, and 1 percent are double orphans. About 44 percent of the orphans live with their mother; 35 percent live with neither parent; and 10 percent live with the father.51 UNAIDS reports that 180,000 children are orphans due to AIDS.52 The Livelihood Empowerment Against Poverty (LEAP), Ghana’s social protection flagship program, provides cash transfers for caregivers of orphans and vulnerable children (OVC) in extreme poverty (see Chapter III). One in five children is engaged in child labor. Some 1.9 million children aged 5–17 years (21.8 percent) are child laborers, according to the GLSS-6. The proportion of boys in child labor was slightly higher (22.7 percent) than girls (20.8 percent). The prevalence of child labor is much higher in rural areas (30.2 percent) than in urban areas (12.4 percent). About 1.2 million (14.2 percent) of children aged 5–17 years are engaged in hazardous child labor. The proportion of boys engaged in hazardous forms of child labor (15.4 percent) was slightly higher than girls (12.9 percent) and in rural areas (20 percent) than in urban areas (7.7 percent). Child labor in hazardous work mostly occurs in artisanal fishing, agriculture, artisanal mining, and quarrying industries. These children are driven to work by poverty. Children under age 15 also work in domestic work, services, and the commercial sex trade. Often they drop out of school or never attended school. Children engaged in child labor and other forms of hazardous work are often exposed to various forms of abuse at the workplace. About 91 percent of children who suffered abuse were involved in child labor and 87.4 percent of them were engaged in hazardous work.53 There is solid international evidence that early childhood development (ECD) programs benefit poor children development. Nutritional supplementation and earlier stimulation 50 For possible inequalities between biological and nonbiological children, see Roelen and Chetttri (2014). 51 MICS 2011, Table CP.10. 52 http://www.unaids.org/en/regionscountries/countries/ghana/. 53 GSS 2014b; Ghana National Commission on Children. 18 interventions administered together, especially to poor children under 2 years old, can help malnourished children catch up with other children.54 Early stimulation, enrollment in early childhood education and in preprimary school, and participation in other learning and school readiness activities help children develop and perform well when they enter primary school. In Ghana, only about 10 percent of children under 3 years old are enrolled in crèches or nurseries; the enrollment ratio is slightly higher for girls (Table 2.10). One the other hand, about two-thirds of children aged 4–5 years are enrolled in preschool or kindergarten, with slightly more girls.55 Data from MICS 2011 indicate that 68 percent of children 3–5 years old participate in some form of ECD, but there are wide regional and income disparities: 85 percent in Greater Accra compared to 51 percent in Northern Region; and 42 percent in the poorest quintile (Q1) compared to 97 percent in the richest quintile (Q5). Table 2.10: Enrollment in Early Childhood Development Programs, 2013 Total Boys Girls Gender parity ratio Crèche/Nursery (0–3 years) , not enrolled 2,629,376 1,339,812 1,289,564 Population 2,863,595 1,457,458 1,406,137 Enrolled 234,219 117,646 116,573 Percent enrolled 8.2 8.1 8.3 1.0 Kindergarten (4–5 years), not enrolled 354,560 188,108 166,452 Population 1,409,527 718,494 691,033 Enrolled 1,054,967 530,386 524,581 Percent enrolled 74.8 73.8 75.9 1.0 Source: Ministry of Education- EMIS (http://moe.gov.gh/site/resources/emis.php). The major risk children of primary and secondary school age face is not attending school. This can be either because they did not enroll in the first place or because they were forced to drop out of school for financial, early pregnancy, or other reasons. Another risk is that children do attend school but learn very little because of the poor quality of education. A low level of schooling almost invariably leads to poor job market prospects and low salaries, and often to a life in poverty. In recent years, Ghana made significant progress in ensuring primary education for all children. Its primary education net enrollment at 84 percent is now in the middle range of its African middle-income peers. It is below Zambia, Cameroon, and the Republic of Congo, but above Senegal, Mauritania, and Nigeria (Table 2.11). Table 2.11: Country Comparison of Net Primary School Enrollment (%) Ghana Cameron Congo, Côte Mauritania Nigeria Senegal Zambia Rep. d’Ivoire Total 84.1 91.5 90.2 61.9 69.6 63.9 73.3 93.7 Boys 84.2 97.1 86.4 67.4 67.1 69.4 70.6 93.0 Girls 83.9 85.9 94.0 56.3 72.1 58.1 76.1 94.4 Gender ratio 0.996 0.885 1.088 0.835 1.075 0.837 1.077 1.015 54 World Bank 2005, 136. See also The Lancet (2013, 2); Black and others (2008); Grantham-McGregor and others (2007, Figure 1); Horton (2008, 179); and Lake (2011, 1277). 55 In Ghana, the official entrance age for preprimary education is 3 years, with a duration of 3 years; the entrance age for primary is 6 years with a duration of 6 years; the entrance age for junior high school (JHS) is 12 years with a duration of 3 years; and the entrance for senior high school (SHS) is 15 years with a duration of 3 years. 19 Source: Ghana -Ministry of Education, EMIS; other countries-World Bank Development Indicators (latest available data). Note: Net primary enrollment is the ratio of children of official primary school age who are enrolled in school to the population of the corresponding official primary school age. It excludes overage and underage children. Nonetheless, there are still over 623,000 primary school-age children who do not attend primary school (314,000 boys and 309,000 girls) (Table 2.12) and large regional disparity remain. Net primary enrollments are below the national average in Greater Accra, Volta, and the Eastern and Upper West Regions. On the other hand, the gender parity ratio is close to unity in most regions, with the exception of the Northern Region, where girls lag behind boys, and in the Upper West, where girls do better than boys (Figure 2.6). Table 2.12: Net Enrollment Rates and Children Who Do Not Attend School, 2013 Total Boys Girls Gender parity ratio Primary (age 6–11) (percent) 84.1 84.2 83.9 0.996 Junior high school (age 12–14) (percent) 47.8 51.0 44.5 0.873 Memo Primary age children who do not attend 623,385 313,846 309,539 primary school Population (age 6–11) 3,909,857 1,987,431 1,922,426 Enrolled (age 6–11) 3,286,472 1,673,585 1,612,887 Source: Ministry of Education- EMIS. http://moe.gov.gh/site/resources/emis.php. Figure 2.6: Primary and Junior High School Net Enrollment and Gender Ratios by Region, 2013 1.20 120.0 1.00 100.0 0.80 80.0 ratio 0.60 60.0 % 0.40 40.0 0.20 20.0 0.00 0.0 NER Primary (rigth) NER JHS (right) Primary Gender ratio JHS Gender ratio Source: Ministry of Education, EMIS (http://moe.gov.gh/site/resources/emis.php). 20 Enrollments are much lower in junior high school (JHS). At this level, net enrollment is only 47.8 percent, with girls lagging behind boys by a significant margin (44.5 percent compared to 51 percent). Net enrollments are below the national average in the Northern and Eastern Regions, Volta, and Brong Ahafo. The gender parity ratio works against girls in all regions, but is near parity in the Upper East and West regions (Figure 2.6). As expected, net enrollments in senior high school (SHS) (25.2 percent)56 are even lower. A World Bank study estimates that 1 million children aged 6–16 years do not attend school in Ghana.57 Primary and junior high school enrollments grow with income. Net enrollments at primary and junior high school are lower for the poorest quintiles than the rich quintiles, with the difference more pronounced in junior high school (Figure 2.7). Figure 2.7: Primary and Junior High School Net Enrollment Rates, by Quintile, 2012 90 80 70 60 50 % 40 30 20 10 0 Q1 poorest Q2 Q3 Q4 Q5 richest Ghana Primary JHS Source: GSS 2014c. To reduce the number of children who do not attend school, the government is implementing several programs. The DFID-supported Complementary Basic Education (CBE) Program enrolled 24,117 children in the Northern, Upper East, Upper West, and Brong Ahafo regions in 2014. For 2015, the CBE is expected to enroll 100,000 additional students in six regions.58 The government is also providing 155 schools with some school infrastructure under the “Elimination of Schools under Tree Program” and supporting the training of more teachers by expanding educational facilities, with the view of meeting the target of 95 percent of trained teachers at the basic level in the medium term.59 At the secondary level, the government began building 50 senior high schools in selected districts across the country. In 2015, an additional 50 schools were constructed and 125 existing schools will be supported with improved educational facilities. This 56 UNESCO Institute of Statistics. 57 Darvas and Palmer 2014, 2. 58 http://www.presidency.gov.gh/. The goal of the CBE program is to support 120,000 out-of-school children from disadvantaged backgrounds in attending a high-quality, nine-month flexible schooling program. Under this program, children who have not attended school are taught in their mother tongue, enabling them to rapidly develop numeracy and literacy skills. On completing this nine-month program, these children will be able to enter primary school at the appropriate class level—usually class three or class four. (Complementary Basic Education. Business Case, March 21, 2012; DFID Ghana). 21 59 MoF 2014a, 14. construction is supported by the World Bank Secondary Education Improvement Project (US$156 million, 2014–19). Demand constraints have also been relaxed. In addition to the capitation payment to schools (payments to school based on the number of students enrolled) to compensate for the elimination of the registration fee in preprimary, primary, and junior secondary public high schools, the government is also providing other assistance and incentives to parents to send and retain children in school, including the free uniform and exercise books programs, and the school lunch program. There is also a large program of scholarships for senior high school students. These programs are discussed in Chapter III. Nonetheless, the direct cost of education to parents remains a major constraint to school attendance for poor families. Data from GLSS-6 indicate that half of children (4–18 years) who never attended school give demand-related reasons for not attending. Fourteen percent of the poor children say they cannot afford school (Table 2.13). On the other hand, one-fifth of poor children give supply-related reasons, such as “No school” or “School too far.” Table 2.13: Reason for Never Attending School (Children 4-18 Years Old) (%) Ghana Extreme Poor all Non- Poor poor Supply 17.9 24.1 20.1 14.3 No school /school too far 17.5 22.8 19.4 14.3 School not safe 0.4 1.3 0.7 0.0 Demand 48.9 51.3 50.5 46.1 Cannot afford schooling 12.9 10.1 14.0 11.2 Family did not allow schooling 9.8 7.3 8.4 12.2 Not interested in school 9.9 13.7 11.1 7.8 Education not considered valuable 1.1 1.1 1.3 0.9 To learn a job 0.6 0.4 0.5 0.8 To work for pay 0.4 0.0 0.0 1.1 To work as unpaid worker in family business/farm 7.4 11.6 8.7 5.3 Help at home with household chores 6.7 7.0 6.7 6.8 Other 33.2 24.7 29.4 39.6 Disabled/ illness 3.9 1.1 2.8 5.7 Too young 26.7 20.9 23.4 32.2 Other 2.6 2.7 3.2 1.7 Source: World Bank staff estimates based on the GLSS-6. The direct cost of schooling still places a heavy burden on the budgets of poor households. While public basic school is expected to be free for parents, data from the GLSS-6 indicate that only 4 percent and less than 1 percent of extreme poor children did not pay anything to attend primary and junior high school public schools, respectively. Of those who paid something, the data indicate that an extreme poor family spent GH¢52 and GH¢85 per year to send one child to public primary school and public junior high school (JHS), respectively, which represents 3.2 percent and 5.2 percent of the family annual food consumption, respectively (Table 2.14).60 This means that an extreme poor family will have to go without food for 12 or 19 days to send one child to public primary school or junior high school. An extreme poor family with a child in public 60 The Ghana Statistics Service estimates the cost of attending school at GH¢459 per year, but it does not differentiate by level of education or private/public school. See Ghana Statistics Service 2014d, 16. 22 primary school and another child in public junior high school would need to go without food for one month. In order to send one child to public senior high school (SHS), the sacrifice in terms of food would be 88 days, or over two months. Most of the costs for poor families with children in primary school and junior high school are for food, boarding, and lodging at school, but the expenditures on uniforms and sport clothes, as well as books and school supplies, is also significant. For senior high school, most of the costs are for fees and registration, followed by food and boarding. The President of the Republic announced that public senior high school is to become free progressively, starting in 2015/16.61 However, the continued burden of education costs on poor families calls into question whether the existing school programs (capitation, school feeding, uniforms, exercise books) are having the desirable impact of facilitating the access of the poorest to education, or are spread too thin and should be targeted on those who need them the most. Table 2.14: Annual Cost of Primary, Junior High School, and Senior High School Public Education per Student (GH¢) Public primary school Public junior high school Public senior high school Ghana Extreme Poor Non- Ghana Extreme Poor Non- Ghana Extreme Poor Non- Poor all poor Poor all poor Poor all poor School fees and 12 3 3 16 21 6 8 27 528 194 233 594 registration fees Contributions to 6 3 4 6 11 6 7 12 23 8 18 24 PTA Uniforms and 15 8 10 18 19 12 13 22 27 17 19 28 sports clothes Books and 18 8 10 21 36 18 20 42 80 38 49 87 school supplies Transportation to 6 0 1 9 18 1 1 25 47 5 9 55 and from school Food, board, and 82 26 40 104 128 33 56 155 286 116 103 327 lodging at school Expenses on 15 3 5 21 34 8 13 43 49 14 17 56 extra classes In-kind 3 1 2 4 3 1 2 3 16 6 4 19 expenses Total 156 52 75 199 270 85 120 328 1055 399 451 1190 Total % 3.6 3.2 3.2 4.2 6.2 5.2 5.1 6.9 24.3 24.2 19.3 25.1 household food consumption Household food 4348 1646 2333 4743 4348 1646 2333 4743 4348 1646 2333 4743 consumption Source: World Bank staff estimates based on the GLSS. Note: PTA = Parent Teachers Association. Improving the quality of schooling is a major challenge in Ghana, as in many other countries where enrollments have grown very rapidly. It is estimated that nearly three-quarters of the students leave basic education without a proficient level of literacy and numerical skills. Learning results are lower for students from the poorest families and from rural areas, making it impossible for them to progress academically. Very few of these poor youth have opportunities to improve their basic skills through second chance programs because they are generally obliged to engage in unskilled, often unpaid household jobs, in agriculture, or in street peddling. Some of these youth 61 http://www.presidency.gov.gh/. 23 get into informal apprenticeships. However, informal apprenticeships in Ghana generally do not provide some basic literacy or numeracy skills (see further discussion in Chapter VI).62 3. Youth The principal risk facing young people is never having attended school or leaving school without the skills that can get them a good and stable job. Most youths without adequate skills remain jobless. The overall rate of unemployment for all age groups reported by the GLSS-6 is 5.2 percent of the labor force: 4.8 percent for men and 5.5 percent for women; and 6.5 percent in urban areas and 3.9 percent in rural areas.63 Of those employed, GLSS-6 establishes that one-third are underemployed. Unemployment among youth (age 15–25) is much higher, at 10.9 percent; 16.3 in urban areas and 7.1 percent in rural areas (Figure 2.8). The youth unemployment rate is slightly higher for girls than boys (11.7 percent versus 10.2 percent). Alternative estimates show youth unemployment at much higher rates. For instance, the OECD 2012 Africa Economic Outlook estimates youth unemployment (15–24 years) in Ghana at 26 percent.64 Figure 2.8: Unemployment Rate per Age Group, 2012 18 16 14 12 10 % 8 6 4 2 0 15 - 25 25 - 44 45 - 64 65+ All Urban Rural Ghana Source: GSS 2014d, 55. Youth employment (unemployment) is a major development issues facing the country, according to a survey conducted by the World Bank Group in September 2011.65 Ghana’s population is young; 40 percent is between the ages of 15 and 34 years. Rapid growth of the youth population, together with migration of young people from rural areas to cities, has outstripped the demand for urban labor.66 About 250,000 young Ghanaians enter the labor market every year, but on average only 2 percent of them are employed in the formal sector. Thus many young Ghanaians join the informal sector.67 62 Darvas and Palmer 2014, 2. 63 Ghana Statistical Service (http://www.statsghana.gov.gh/glss6_presentation.html). 64 http://www.africaneconomicoutlook.org/fileadmin/uploads/aeo/PDF/Ghana%20Full%20PDF%20Country%20Note. pdf. 65 World Bank 2013b, 18. 66 World Bank 2013b, 31. 24 67 Schwegler-Rohmeis and others 2014, 20. Only 45 percent of graduates from junior high school continue to senior high school, on average, according to the Ministry of Education. Of the remaining 55 percent, 5 percent go to formal vocational schools, 7 percent to formal technical institutes, and 10 percent to traditional informal apprenticeship. The remaining 33 percent of junior high school graduates may fall within the category Not in Education, Employment or Training (NEET) bracket or may start working informally.68 Youth unemployment and graduate joblessness are serious concerns in Ghana. As mentioned, about 250,000 youth enter the labor market every year in Ghana.69 Many have limited basic skills and are obliged to engage in unskilled, often unpaid household jobs. Some get informal apprenticeships, which usually do not provide the basic literacy and numeracy skills required to move to higher-skill, higher-productivity jobs.70 Of those joining the labor force every year, about 70,000 are tertiary-level graduates, but only 5,000 are employed by the formal sector. The rest survive in the informal sector or remain unemployed.71 A large number of youth are in skill training and apprenticeship programs at any time, but these programs need to be better targeted. About 500,000 youth are in apprenticeships at any one time, according to a recent World Bank study; of these, 80 percent are in the informal sector and only 20 percent are in the formal public and private training centres.72 Only 5 percent of extreme poor youth (15–25 years) have participated in apprenticeship, compared to 13 percent of the nonpoor, data from GLSS-6 indicate (Table 2.15). In addition, the availability of regular labor market information is needed to facilitate access of jobseekers— in particular, youth and students— to labor markets. Table 2.15: Apprenticeship Participation by Youth (15–25 years) Ghana Extreme poor All poor Nonpoor Currently in program 5.6 2.5 4.6 5.9 Participated in the past 6.0 2.3 3.3 6.9 Never in a program 88.4 95.2 92.1 87.2 Paid a fee 8.9 3.5 5.7 9.9 Did not pay a fee 3.9 1.7 3.0 4.2 Source: World Bank staff estimates based on the GLSS-6. 4. Working Poor and Underemployed Poverty is related to where people work, rather than whether they work. Of those people who work, 22 percent are poor and 7.5 percent are extreme poor, according to data from GLSS-6. The survey indicates there were about 11 million people working; therefore 2.5 million were working poor and 840,000 were working extreme poor. The highest incidence of poverty is among those self-employed in agriculture (Figure 2.9). This persistent poverty results from the very low productivity of subsistence farming, particularly in the northern regions of the country. As 68 Schwegler-Rohmeis and others 2014, 48. 69 Schwegler-Rohmeis and others 2014, 20. 70 Darvas and Palmer 2014, 2. 71 http://gebssghana.org/. 72 Darvas and Palmer 2014, 6. 25 expected, poverty is also high among the unemployed and inactive.73 The incidence of poverty for all type of workers has declined since 2005. Figure 2.9: Poverty Incidence by Type of Employment of Head of the Household (Percent) 50 45 40 35 30 25 20 15 10 5 0 Poverty 2005 Poverty 2012 Source: GSS 2014c. The rural extreme poor family is an especially vulnerable group. This group, which numbers about 313,000 families, has a higher incidence of extreme poverty (10.6 percent) than the national average (Table 2.16). Many of these households make their living in low-productivity, seasonal agricultural activities. About 92,000 working-age people in extreme poverty in rural areas are dedicated to agricultural activities (15 years and over, without employees). The incidence of extreme poverty among this group is more than three times the national average. Labor-intensive public works is a good instrument to help this group gain some income during the agricultural slack period (see Chapter III). Table 2.16: Rural Extreme Poor Households Incidence of Extreme Poor all Nonpoor Ghana extreme poor poverty (%) Rural HH 10.6 312,846 831,523 2,108,880 2,940,403 Heads of rural HH male 11.4 247,415 645,969 1,526,601 2,172,570 Heads of rural HH female 8.5 65,431 185,554 582,279 767,833 Rural people over 15 years in 25.4 92,049 172,186 190,337 362,523 agriculture Ghana, total 8.4 2,228,720 6,428,071 20,191,316 26,619,387 Source: World Bank staff estimates based on GLSS-6. HH = household. Migration has been a strategy by poor households, many from the rural North, to diversify their sources of income and provide for a steady flow of income. Migratory movement within 73 GSS 2014c; World Bank 2016a. 26 Ghana has usually been from the North to the South and from the less developed rural areas to the relatively developed urban areas. Accra (GAMA) has the highest proportion of migrants (60.3 percent), followed by the rural forest (51.6 percent). The other urban areas have 46.7 percent of migrants, while rural coastal has 44.6 percent. Rural savannah (37.5 percent) has the lowest proportion of the migrant population. With the exception of Accra (GAMA), where the proportion of male migrants (60.9 percent) is slightly higher than females (59.8 percent), the proportion of females is higher than males in all other localities. The 25–29 age group constitutes the highest proportion of migrant population for both sexes. In Accra (GAMA), the main reason for migrating is to seek employment (26.2 percent), followed by other family reasons (22.3 percent). 74 5. Elderly Poor Most of the elderly poor live with families. Of the poor elderly, only 6 percent live alone and 2 percent live only with their spouse or partner (Table 2.17). Ninety-two percent of the elderly poor live with family members. Among the better-off, more live alone (16 percent) or only with a partner/spouse (7 percent). Therefore, targeting poor (or extreme poor) families would reach most of the poor (or extreme poor) elderly. Table 2.17: Elderly Living Arrangements Ghana Extreme All poor all Nonpoor Poor Elderly (60+) 100 100 100 100 Live alone 13.6 7.4 5.9 15.9 Live with spouse/partner 5.9 2.6 2.2 7.1 Live with families 80.5 90.0 91.9 77.0 Memo: Number elderly 1,820,313 153,819 435,227 1,385,086 Percent distribution 100.0 8.5 23.9 76.1 Source: World Banks staff estimates based on GLSS-6. Few elderly people in Ghana retire with a pension, and one in four is poor. Affiliation in the Social Security and National Insurance Trust (SSNIT) is mandatory for all private and public employees. There are two requirements to receive the SSNIT pension: (i) to contribute at least 180 monthly contributions; and (ii) to be 60 years old (or 55 years old, if the worker opts for early retirement with a reduced pension). An estimated 1.8 million people were aged 60 years and over in 2014.75 Of these, only about 142,000 had a pension, or just 7.7 percent (Table 2.18). The incidence of poverty and extreme poverty among the population 60 years and over is 24 percent and 8 percent, respectively, according to GLSS-6. Table 2.18: SSNIT Pensioners 2010 2011 2012 2013 2014 Number of pensioners 107,312 112,522 119,323 129,691 142,076 Percent of population 60+ 6.5 6.7 6.8 7.2 7.7 Memoranda: Population 60+ 1,643,381 1,690,763 1,743,182 1,797,088 1,838,930 Source: SSNIT and Ghana 2010 Population and Housing Census. 74 GSS 2014d, 64. 75 Projection based on the Ghana 2010 Population and Housing Census (GSS 2012). 27 The government is implementing an Ageing Policy and Action Plan. With support from the United Nations Population Fund (UNFPA-Ghana), implementation of the Action Plan began in December 2011. The policy seeks to mainstream older persons into the national development process. To ensure the implementation of the policy, the Ministry of Gender, Children, and Social Protection (MoGCSP) intends to propose legislation on the policy to Parliament and establish a National Council on Ageing. 6. Poor Persons and Families Living with Disabilities and/or HIV/AIDS About 358,000 families in Ghana are extreme poor. These families cannot afford to provide the minimum amount of food required by their members, even if they spend their entire budget on food (Table 2.19). Many of these families need also to provide for members with disabilities, members living with HIV/AIDS or other chronic diseases, and their elderly. Table 2.19: Composition of Vulnerable Families (%) Ghana Extreme poor All poor Nonpoor Families 100.0 100.0 100.0 100.0 Single 20.1 6.1 5.5 22.9 No children 13.8 11.5 10.7 14.5 1 child 17.0 8.8 10.8 18.2 2 children 16.4 12.9 14.4 16.8 3 children 13.5 13.7 15.4 13.1 4+ children 19.2 47.0 43.1 14.6 Number of families 7,525,543 357,822 1,081,189 5,511,912 Percent distribution 100.0 5.4 16.4 83.6 Source: Word Bank staff estimates based on GLSS-6. The poorest families tend to be larger. Poorest families generally have a larger number of children. Nearly half of the extreme poor families have four or more children (Table 2.19). Only 6 percent of those in extreme poverty families are single, and less than 12 percent have no children. Better-off families tend to be smaller; more than one-third are single or have no children. Therefore, programs that place limits on the number of beneficiaries per family will greatly disadvantage most of the extreme poor families and fail to reach their objectives. More than 2 million persons with disabilities (PWD) are estimated to be living in Ghana. The MoGCSP estimates that 20 percent of the population in Ghana has some form of disability.76 On the other hand, the Ghana Federation of the Disabled places the incidence of disability at between 7 and 10 percent of the population, with the three most prevalent types of disability related to visual impairment, hearing impairment, and physical disabilities.77 By way of comparison, the disability prevalence rate in the adult population in 56 lower-income and higher-income countries averaged 18 percent and 11.8 percent, respectively, according to the 2011 World Health Organization and the World Bank study on disability worldwide.78 For Ghana, the report indicates 76 MoGCSP 2014b, 51. 77 http://www.gfdgh.org/disability%20situation%20in%20ghana.html. 78 WHO and World Bank 2011, 28, Table 2.1. The difference in estimates derives in part from different definitions of disability. 28 a disability rate of 12.8 percent for 2002–04.79 Therefore, it can reasonably be assumed that there are at least 2 million people with disabilities in Ghana. Data from GLSS-6 based on self-reporting of disability indicates that the incidence of disability, as expected, is greater for the elderly and among the poor. The National Council on Persons with Disabilities was established in 2008 to implement the provisions of the Persons with Disabilities Act 715 of 2006. The Council is mandated to establish and maintain a register of PWDs and nongovernmental organizations (NGOs) working on disability in Ghana. The Council, which is attached to the MoGCSP, is setting up resource centers in all districts as business incubators to provide business skills training and start-up kits for all PWDs. It is also promoting the enactment of legislation to require easy access of PWDs to all public buildings. The Council receives 2 percent of the District Assembly Common Fund (DACF) annual resources: about GH¢10 million in 2014.80 In Ghana about 220,000 people were living with HIV/AIDS (PLWHA) as of 2013. The prevalence rate among adults (15–49 years) is 1.3 percent, and varies from 3.6 percent in the Eastern Region to 0.9 percent in the Northern Region. HIV afflicts 35,000 children aged 0 to 14. As mentioned, about 180,000 children (0–17 years) have been orphaned because of AIDS.81 Prevalence among young people 15–24 years—which is used as a proxy for new infections—was 1.3 percent in 2012, down from 2.1 percent in 2009.82 Because of insufficient funds and difficulties in accessing test kits and antiretroviral drugs, only 60 percent of eligible people living with HIV were receiving antiretroviral therapy (ART) in 2013. Similarly, only 76 percent of HIV-positive women who were pregnant were receiving ART to reduce mother-to-child transmission. An important program that benefits AIDS-affected households is the World Food Programme food assistance to food-insecure households, which provides food rations to patients on ART in the Northern, Upper East, Upper West, and Eastern Regions. Recently, the GoG removed the GH¢5 fee charged to patients when they collect their monthly ARVs. In sum, the Ghanaian population is vulnerable to a series of shocks, including idiosyncratic shocks that affect particular families. The poor are more vulnerable to these shocks because they have fewer means to deal with them. The social protection system should help minimize the impact of these shocks using a “portfolios of coherent programs” that work together to deliver resilience, equity, and opportunity.83 Despite steady progress in improving health conditions, major challenges remain, as Ghana is unlikely to achieve key MDG targets, particularly reducing the proportion of children under five who are underweight, and infant, under-five, and maternal mortality rates. Continued support to ease demand and supply constraints to access quality health care is necessary. Likewise, while most children are attending school, many poor families are unable to send their children to school because of demand and supply constraints; these constraints need to be further relaxed. Youth employment is a challenge in Ghana, as in many other countries. Many adults in Ghana are working poor, as the quality of jobs, particularly in the agricultural and informal sectors, often does not provide a minimum income for families to evade poverty. Orphans and vulnerable children, child workers, persons living with disabilities, and persons living with HIV/AIDS are vulnerable groups that should continue to receive special attention. 79 WHO and World Bank 2011, 273. 80 MoGCSP 2014b, 51. 81 http://www.unaids.org/en/regionscountries/countries/ghana/. 82 Ghana Aids Commission 2014. 83 World Bank 2012b, xiv. 29 III: Core Social Protection Programs to Address the Risks Ghana’s social protection programs include social assistance programs (or social safety nets), scholarships, social insurance, and active labor market programs that target youth. This chapter briefly reviews the core social protection programs as well as electricity and fuel subsidies. This chapter reviews core social protection programs that are currently being implemented in Ghana. The discussion closely follows the definition of social protection and the list of programs covered by the GoG study (2014b) that was arrived at in consultation with the Government of Ghana (GoG).84 Social protection programs include social assistance programs, scholarships, social insurance, and active labor market programs. The chapter covers basically the same programs as the GoG report, with a few minor adjustments. The Ghana-Luxembourg Social Trust (GLST) is not covered here because it terminated in mid-2014.85 Likewise, the Social Inclusion Transfer (SIT), a cash transfer program supported by the African Development Bank, is also not covered, as it ended in 2012. On the other hand, the review includes scholarships programs that were covered only partially in the GoG 2014b report, and the National Apprenticeship Program, which is a new program. Although scholarships have not been considered as part of social protection in previous studies for Ghana and are not part of the 5 social protection flagship programs defined in the Ghana National Social Protection Policy86, they are used as social protection instruments around the world to help poor households access education facilities. They can be cash to families and/or in the form of waivers of educational fees or voucher that reimburse households or schools/facilities for actual expenditures. This chapter describes the main design and implementation characteristics of the various programs, based on information available, before turning to their assessment in Chapter IV. A. Social Protection Programs 1. Social Assistance The core social assistance programs being implemented in Ghana are described below and summarized in Table 3.1. Whenever possible, the program’s objectives, target group, eligibility criteria, benefits, cost, and impact are described. Social assistance programs are noncontributory transfer programs targeted in some manner to the poor and those vulnerable to poverty and shocks. 84 GoG 2014b. 85 It was a conditional cash transfer initiated in 2009 that targeted poor pregnant women and children. It was implemented by ILO and supported by the Luxembourg trade unions. 86 To achieve its policy objectives, the National Social Protection Policy proposes to improve the design and implementation of five flagship programs in the short term, identifies as: the Livelihood Empowerment Against Poverty, the Labor Intensive Public Works, the School Feeding Program, the National Health Insurance Exemptions and the Basic Education Capitation Grants. Other programs and plans of action are recognized to be relevant to social protection delivery in Ghana and to contribute to promoting sustainable livelihoods and human dignity. Particularly, social protection related interventions implemented by the Ministry of Health (MoH), Ministry of 30 Education (MoE) and Ministry of Food and Agriculture (MoFA) are relevant for integration in the short and medium term implementation of the policy (MoGCSP 2015). Table 3.1: Selected Social Protection Programs, 2014 Program Selection method Number of beneficiaries Benefits Social assistance LEAP Targeted 90,754 households (2014) Cash transfers: GH¢24 per month for 169,790 persons 1 eligible member; GH¢30 for 2; 83,238 persons under 18 GH¢36 for 3 ; GH¢45 for 4+ 23,173 adults Free NHIS membershipa 63,379 persons over 65 Education capitation grants Universal 5.6 million children Free school registration in public school (preprimary, primary and JHS). Schools receive GH¢4.5 per student per year. Since 2013, schools in 57 deprived districts have received additional grants School feeding Universal 1.7 million children in public pre- Every school day one hot nutritious primary and primary schools meal Take-home rations for girls Targeted 32,073 girls in 508 schools in Each girl receives 8 kilograms of northern regions and Volta maize, 2 litters of oil, and 1 kilogram of iodized salt at the end of the month School uniforms Targeted 400,000 children (2013) One free school uniform a year Free exercise books Targeted 4.8 million books (2013) Free exercise books Elimination of child labor Targeted 21.8 percent of children aged 5–17 Provides assistance to pursue (NPECLC) years in child labor alternative livelihoods NHIS exemptions Targeted 6.9 million exempt Exempt from payment of premium (children under 18, persons 70+ years, the indigent, LEAP beneficiaries, pregnant women and SSNIT pensioners) LIPW Targeted 30,042 people Public works-GH¢7 per a six-hour working day, for an average of 101 days a year Social welfare services Targeted 2,234 orphaned and poor children Shelter and care, counselling and 3,000 children in conflict with the assistance, social welfare services, law vocational and skill training 3,672 aged persons 1,165 persons with disabilities Scholarships Scholarships (SS) Targeted 113,689 (senior high school) Boarding 23,080 (tertiary) Tuition, books, living expenses Social insurance SSNIT 142,000 pensioners Pensions, medical and disability benefits Workmen’s Compensation Work injury benefits Active Labor Market Programs National Apprentice Targeted 6,000 people Skills upgrading, apprenticeship Program 31 LESDEP Targeted 15,005 people Local enterprise and skills development Youth Employment Agency Targeted 100,000 youth Oversee the development, supervision, and coordinating the creation of jobs for the youth Source: Program managers and staff estimates. LEAP = Livelihood Empowerment against Poverty program; LESDEP = Local Enterprises and Skills Development Program; LIPW = Labor Intensive Public Works program; NHIS = National Health Insurance Scheme; NPECLC = National Program for the Elimination of Child Labor in Cocoa; SS=Scholarships Secretariat. a. Benefit amounts have been updated. While new benefits amount have been applied since September 2015, the analysis in this report relies on the benefit amounts specified in this table. Livelihood Empowerment Against Poverty (LEAP). This cash transfer program is GoG’s flagship social protection program. It was launched in 2008 to supplement the income of “dangerously poor households,” reach the “bottom 20 percent poorest households in Ghana (about 1 million),” and help them “leap out of poverty.”87 It is managed by the Department of Social Development (DSD) in the Ministry of Gender, Children, and Social Protection (MoGCSP). LEAP has a decentralized implementation structure. Primary implementation responsibility lies with the District Social Welfare Officers, with oversight from the Regional Social Welfare Officers and the National LEAP Management Unit in DSD. The program is supported by the World Bank (US$15.7 million from 2014 to 2017),88 the U.K. Department of International Development (DFID) (£36.4 million from 2012 to 2016), and the U.S. Agency for International Development (USAID) through UNICEF (the United Nations Children’s Fund) (US$2.5 million). The main objective of the LEAP Program is to reduce poverty by increasing consumption and promoting access to services and opportunities among the extremely poor and vulnerable. Its specific objectives are to:89  Improve basic household consumption and nutrition among children younger than 2 years, the aged (65 years or older who lack productive capacity), and people with severe disabilities  Increase access to health care services among children younger than 5 years, the aged (65 years or older who lack productive capacity), and people with severe disabilities  Increase basic school enrollment, attendance, and retention of beneficiary children between 5 and 15 years of age  Facilitate access to complementary services (such as welfare, livelihoods support, and improvement of productive capacity) among beneficiary households. To be eligible for a cash transfer from LEAP, a household must be extreme poor as determined by the targeting process. It must also have at least one member who is elderly (65 years old or older) and lacks financial/material support; or an orphan and vulnerable child (OVC); or a person with severe disability (who lacks productive capacity). The selection of beneficiary households is done through a four-step process. First, there is geographic selection of the poorest districts and communities. Second, the Community LEAP 87 MESW 2012b, xii and 105. 88 This is part of the US$50 million Ghana Social Opportunity Project additional financing that continues to support the strengthening of the social protection system. The initial support was part of the Ghana Social Opportunity Project (US$89.1 million) approved in 2010. 89 MoGCSP 2014a, 1. 32 Implementation Committee (CLIC) selects the eligible households in the community. Third, a proxy means test (PMT) is administered and the households are ranked according to their poverty "score" and checked against the eligibility criteria of the relevant category (vulnerable children, older people, severely disabled persons). Finally, there is a community validation mechanism to determine whether or not the selected households are among the most impoverished. Eligibility entitlements should be recertified every four years. Selected households receive a photo-ID beneficiary card that they must present to collect the benefits. In 2014, there were 90,754 LEAP beneficiary households and 169,790 direct beneficiaries, distributed as follows: younger than 18 years, 83,238;90 adults, 23173; and elderly 63,379. LEAP is operating in all 10 regions and in 144 of the 219 districts in the country. LEAP beneficiaries are eligible for free membership in the National Health Insurance Scheme (NHIS). In 2014, the program invested GH¢38.7 million: GH¢20 million financed by the GoG, and the remainder by DFID. The Additional Financing (AF) from the World Bank started disbursing in 2015. LEAP transfers cash to the eligible households. The amount transferred depends on the number of eligible members. In 2012, benefits were increased threefold, from an average of GH¢12 to GH¢36 per month per household. Since 2012, for one eligible member, the monthly benefit for the household has been GH¢24; for two eligible members, GH¢30; for three eligible members, GH¢36; and for four or more eligible members, GH¢45. The average household benefit stands at GH¢36 per month.91 LEAP grants are paid every two months through the Ghana Post Office. DFID has financed an e-payment pilot and preparation is under way to take it to scale. Electronic payment is being introduced through the telecommunication company MTN cell phone service, Ghana Interbank Payment and Settlement Systems, e-swich smart card, and Aya debit card. LEAP is essentially a conditional cash transfer program with an unconditional cash (UCT) component for the elderly and disabled in extreme poverty. The (soft) conditions or coresponsibilities are described in the Operational Manual and apply only to households with children under the age of 15 years.92 These include education coresponsibilities (enrollment and attendance in school for children between 5 and 15 years old); and health coresponsibilities (members of beneficiary households who have children under 5 years old must visit health facilities for vaccinations and participate in growth monitoring sessions every three months). Unconditional cash transfers are for households that are eligible and do not have children aged below 15 years.93 While the primary activity of the District Social Welfare Officers has been delivering transfer payments, they are also responsible for home visits to verify the status of caregivers, awareness of the program’s health and education coresponsibilities, and management of program complaints and grievances. In early March 2015, MoGCSP announced the expansion of LEAP to pregnant women and infants with support from UNICEF and the U.S. Agency for International Development (USAID). The new component (LEAP 1000) is initially covering 6,000 households with pregnant women and children under one year in ten districts in the Northern and Upper East Regions. From 90 Includes 23,337 beneficiaries whose age is currently unknown due to a data problem but who probably belong to this group. 91 At the time of writing this report, discussions were on-going to increase this to an average of GH¢48 (one member); GH¢64 (two members); GH¢72 (three members); GH¢90 (four members); and GH¢ 108 (four or more members). 92 MsGCSP 2014a. 93 MoGCSP 2014a, 3. 33 January 2016, the new category under LEAP 1000 will be integrated in the mainstream LEAP, in addition to its existing three categories. Like the exiting LEAP, LEAP 1000 provides the 6,000 households with a cash grant every two months, depending on the number of beneficiaries in the households during a three-year period. Beneficiaries are also provided with material explaining the key responsibility to abide by, including attendance at preventive health care services and health and nutrition education sessions. Data collection on potential beneficiaries has been done electronically using smart phones.94 The program’s impact has been found to be positive, but some operational problems remain. Findings from an evaluation undertaken by the Institute for Statistical, Social and Economic Research (University of Ghana-Legon) and the University of North Carolina in 2012 have shown positive impacts, especially on education, food security, health, and productivity of beneficiaries.95 According to the study, compared to the control group, food insecurity has been reduced for LEAP families by 25 percentage points, especially for households led by women (by 32 percentage points).96 In addition, grade repetition was reduced by 11 percentage points, absenteeism was reduced by 10 percentage points for children in primary school, and by 11 percentage points for girls in secondary school, and school enrollment among secondary school-aged children increased by 7 percentage points. Ninety percent of all LEAP households was enrolled in the National Health Insurance Scheme (NHIS) (a 7 percentage point increase over the comparison group),97 and morbidity was reduced for children aged 6–17. LEAP households were 7 percentage points more likely to participate in paid work, with a significant increase in female labor on-farm activities ranging from 9 to 13 days depending on the size of the household. LEAP households—particularly female-headed households—were also 11 percentage points more likely to hold savings, receive gifts, and reduce loan holdings. Self-reported happiness increased by 16 percentage points over the control group.98 On the other hand, a World Bank supervision mission in 2014 identified the following operational challenges; (i) delayed payments to beneficiaries; (ii) an outdated proxy means test; (iii) a management information system (MIS) that needed to be updated; (iv) lack of social accountability and grievance redress mechanisms; and (v) inconsistent project communication, particularly with beneficiary communities.99 The Community LEAP Implementation Committee (CLIC) has also been found to be less participative than hoped. While as of early 2015 about 95 percent LEAP beneficiaries had been enrolled in the NHIS, many may have not renewed their memberships and therefore cannot take advantage of NHIS benefits. 100 Following these assessments, a number of reforms, including a revision of the proxy means test, adjustment of the benefit level, and revision of coresponsibilities, are being implemented. LEAP is currently piloting a new electronic data collection mechanism. LEAP is also developing a management information system for the program that will support data entry and analysis at the national and decentralized levels, is introducing a new electronic payments mechanism, and is training staff and building their capacity to implement LEAP more effectively. 94 http://www.citifmonline.com/2015/02/26/gender-ministry-expands-leap-program/#sthash.0IjTJlk5.dpbs. 95 The baseline data for the study was collected in 2010. See http://www.cpc.unc.edu/projects/transfer/countries/ghana. 96 Handa and others 2014. 97 Handa and others 2014, 8. 98 World Bank 2014d, 3. See also FAO 2013 on the economic impact of LEAP. 99 World Bank 2014d, 14. 34 100 NHIS officials indicate that 50 percent of the entire membership does not register again every year, as required. In 2014, a new monitoring and evaluation (M&E) system was introduced to improve LEAP operational processes, including performance indicators and new reporting tools. Social accountability and grievance redress systems for the program are also being strengthened.101 Recently, the members of CLIC started receiving GH¢10 during each payment cycle (every two months) to help cover some of the costs they incur on LEAP activities. The National Targeting Unit under MoGCSP is using the new targeting mechanism to build up a single registry database. The Ghana National Household Registry could be used by other social protection programs to manage membership. The World Bank is providing additional financing (AF) to the Social Opportunities Project program.102 The AF will seek to better structure the implementation of the program’s “soft conditionality” through regular workshops. The AF will continue to provide support to capacity building through the development of the M&E system for the program that will support data entry and analysis at the national and decentralized levels. The M&E will be coordinated with the Ghana National Household Registry, which is being developed. The AF will also support development communication activities, as well as monitoring and evaluation (regular spot checks of operation activities and impact evaluation studies), and social accountability and grievance redress systems for the program. A new impact evaluation study is being prepared. Education Capitation Grant to Schools. The government provides a transfer to public schools for each child enrolled in basic education in preschool, primary, and junior high school, in the form of a capitation (per student) payment. The payments are in lieu of registration fees, which were abolished in 2005. The capitation grant program aims to facilitate universal access to basic school by abolishing the school registration fee for parents and financing the implementation of the School Performance Improvement Plan (SPIP), while providing an element of self- management for the schools. Schools receive GH¢4.5 per year (increased from GH¢3, which was equivalent to the former registration fee) per student directly from the Ministry of Education. The capitation grant uses the school enrollment registries to allocated funds to the schools. The program has had an important impact on enrollment. In 2004, when it was initiated in the 40 most deprived districts, enrollments in preschool and in primary school increased by 36 percent and 14.5 percent, respectively. In 2005, the program was extended nationwide to all public schools. In 2014, 5.6 million pupils benefited from the program, at a cost of GH¢16.7 million, as the transfers corresponding to only two terms were disbursed to schools. As expected, smaller schools have found it more difficult to cover their costs with the capitation grants than larger schools. In order to compensate smaller schools—most of which are located in the most remote and poorest areas—for their higher fixed costs, an additional grant has been given to public schools in 57 of the poorest districts since 2013 from the Ghana Partnership for Education, which is managed by the World Bank (US$70.5 million, from 2013 to 2016). In 2014, additional transfers of GH¢800, GH¢ 1,200, and GH¢1,000 was made for preprimary, primary school, and junior high school, respectively, and represented an investment of GH¢18.7 million. In 2015, the amounts were increased to GH¢1,000, GH¢1,500 and GH¢1,200 for preprimary, primary school, and junior high school, respectively. The program also supports increased teacher training, supervision, and 101 World Bank 2014b. 102 World Bank 2014 AF Ghana Social Opportunity Project for US$50 million. 35 implementation of school report cards. Given the lack of rigorous evaluation, the program plans to finance an impact evaluation of the capitation grant.103 A public expenditure tracking survey (PETS) on the education capitation grants conducted in 2010 found inadequate recordkeeping and significant leakages (missing funds) in the course of the transfer from the Ministry of Finance/Ghana Education Service (GES) to the District Education Departments (DED), and from the DED to the schools.104 Leakages tended to be more pervasive with the transfer between districts and schools. The PETS indicated that the releases of the grant had been quite irregular from the MoF or from the DED, and most of the schools still charged levies/fees as a result of inadequacy and delays in release of capitation grant. Two-thirds of District Directors said the amount of the capitation grant received by schools in their districts was inadequate. This opinion was shared by the majority of head teachers, teachers, Parent Teacher Associations (PTAs), executives of school management committee, and parents. Only a small proportion of respondents concluded that the capitation grant had reduced significantly the financial burden of parents.105 The Ghana School Feeding Programme (GSFP). This program began in 2005 as a pilot in 10 schools (one from each region) under the Ministry of Local Government and Rural Development (MLGRD) and support from the Dutch government. The program has recently been located under the MoGCSP. The World Food Programme (WFP) is providing both cash transfers and technical assistance and to 50,000 beneficiary pupils for two days in every school- going week in the three regions in the north of Ghana. All the other development partners, such as Partnership for Child Development, SNV (a Dutch NGO), the World Bank, and UNICEF, are providing technical assistance to the program. In 2014, the program served meals to over 1.7 million children in 4,881 preprimary and primary public schools (about 17 percent of all public schools) in 216 districts all over Ghana, reaching approximately 39 percent of all children registered in public preprimary and primary education. The program seeks not only to increase school enrolment, attendance, and retention, but also to increase domestic food production to improve household incomes and food security in poor communities in Ghana, as part of the Home Grown School Feeding Initiative.106 The program provides children in public kindergartens and primary schools one hot nutritious meal prepared from locally grown foodstuffs every school day (195 days a year). In most cases, the meals are prepared in the schools by local caterers, who prefinance the preparation of the meals and are later reimbursed by the GoG. There are 5,000 caterers across the country. At the school level, the program is managed by the School Implementation Committee which is chaired by the PTA Chairman. The school feeding policy was prepared in early 2014 and has been approved by the Cabinet. A consultant has been procured with the support of WFP Centre of Excellence in Brazil to draft the policy into a bill for the legislature. 103 World Bank 2012a, 13. 104 The survey covered 30 public primary schools randomly selected across six districts in three regions as follows: ten in the Northern Region, twelve in Ashanti, and eight in the Western Region. 105 Ampratwum and Armah-Attoh 2010, 2. 106 http://hgsf-global.org/en/bank/news/498-improving-incomes-in-ghanas-school-feeding-program. 36 The payment of the school meal to the caterers was GH¢0.5 per student in 2014 and was increased to GH¢0.8 in 2015 to improve the quality of the meal. In 2014, the cost of the program was GH¢165 million, of which GH¢158 million was financed by the District Assembly Common Fund (DACF) and GH¢6.7 million from external sources (GH¢6.6 million from WFP and GH¢0.1 million from UNICEF). While there is no impact evaluation of the program, a series of program reviews credit it as having a positive impact on enrollments and retention.107 The School Feeding Program has attracted more students to the classroom, but it has suffered in the past from poor targeting. A World Bank study found that only 21 percent of the investment of the School Feeding Program was going to the poor.108 More affluent regions of the country were getting a larger share of the benefits than poorer communities. Based on this evidence, the GoG requested support from the World Bank and the WFP to retarget the program. Information about which regions and which schools should be receiving the program (and from which ones support should be phased out) was compiled by the World Bank using national statistics, data from WFP’s Comprehensive Food Security and Vulnerability Assessment 2008/2009, and spatial data variables. With this information, the government launched a retargeting effort in 2011. As a result, WFP estimates that 70 to 80 percent of the investment in school meals currently goes to the poorest communities.109 The program has incurred significant delays in paying the local caterers, which has affected quantity and quality of food provided. While in recent years, payment arrears of more than six months were common, arrears averaged about 125 days in March 2015; by June 2015, all arrears to caterers had been cleared. On the other hand, the expected connection with local agriculture has been less than anticipated. It was expected to provide at least of 80 percent of the food inputs for the program, but the estimated average is 54 percent purchases from farm-based organizations, varying from 85 percent in the Upper West to 30 percent in Ashanti.110 Other estimates place local content at less than 20 percent.111 The GoG has drawn on the Buffer Stock Company to supply basic grain (rice and maize) to the program, which are delivered to the caterers by the company and deducted from their payment. This has helped increase the local content of the program to about 65 percent.112 The program management has difficulties in monitoring the program implementation at local level and the involvement of communities and parents in the program varies widely across communities. To this end, UNICEF is supporting the Ghana School Feeding Program in building a robust monitoring system for the program. An important initiative is the provision of school feeding to special education schools, which are receiving increasing attention from the authorities. According to the Ghana Shared Growth and Development Agenda (GSGDA) II, policy intervention planned includes “integrating persons with special needs in mainstream education; ensuring that rehabilitated/new infrastructure are disability-friendly to students; strengthening the capacity of institutions responsible for persons with disabilities, including specialist teachers, resource assessment and rehabilitation centers; providing the incentives for special educators; enhancing the pedagogical skills of teachers of 107 See, for instance, studies referenced in Kedze (2013, 24). 108 Wodon 2012. 109 WFP 2013, 80. 110 SNV 2011, 65. 111 Kedze 2013, 13. 112 Information received from the responsible for the program in the MLGRD in an interview in March 4, 2015. 37 special education; and ensuring that children with special needs benefit from the advantages of modern technology.”113 Take Home Rations for Girls. This program began in 1999 and is administered by the Ghana Education Services (GES) with support from the World Food Programme. In 2014, it provided take-home rations for 32,073 girls in 508 schools in 12 districts (10 in the Northern region and 2 in the Volta Region). Each girl receives 8 kilograms of maize, 2 liters of oil, and 1 kilogram of iodized salt at the end of each month if they are in school for at least three-quarters of the month. The program has been credited as a major reason why two of the regions covered under the program, the Upper East and the Upper West, were the first in Ghana to attain gender parity in their schools, but no formal evaluation of the program is available. The cost of this program in 2014 was GH¢ 6 million, financed by WFP. Girls in the schools assisted by the program who excel in studies (as measured by their performance on the Basic Education Certificate Examination, BECE) are awarded a scholarship to ensure that they are able to complete senior high school, consisting of GH¢ 800 (US$ 76), as well as groceries and textbooks. This latter program which benefited 14 girls in 2014 (338 girls to date), is supported by WFP’s Gender Innovation Fund. 114 Provision of Free School Uniforms and Exercise Books. In 2009, the GoG started providing free school uniforms and exercise books to pupils in basic education as part of the effort to improve education enrollments. Uniforms are targeted to poor and vulnerable students. Under the program, 400,000 school uniforms were distributed annually in the last several years to students in poor communities across the country, and have reached 75 percent of the targeted schools, according to the Ghana Education Service (GES).115 About 4.8 million books have been distributed each year in recent years, covering nearly 90 percent of pupils enrolled in public basic schools. The procurement of the uniforms and exercise books are centralized in Accra and are awarded through competitive bidding to local producers. The program has also faced problems with payment delays to the suppliers. For example, the exercise books procured in 2012 were not paid until 2014. In 2014, the program invested GH¢7.6 million in exercise books. No information on spending for school uniforms in 2014 was available. For 2015, the President of the Republic announced that the program will be expanded: about 10,000 locally produced school sandals, 500,000 uniforms, 6 million textbooks, and 15 million exercise books will be distributed to schools across the country.116 A public expenditure tracking survey (PETS) on the distribution of books by the Ministry of Education found improper recordkeeping and leakages (missing books) from the Ministry of Education/publishers to the District Education Department (DED). The Ministry of Education’s Textbook Development and Distribution Policy is that each pupil in basic school has access on an individual basis to three government-designated core textbooks: English, mathematics, and science. The PETS published in 2012 and covering 6 districts (3 poor and 3 endowed) and 30 schools, found that pupil-core textbook ratios in two-thirds or more of the schools did not meet the Ministry of Education textbook policy.117 Despite serious problems with poor 113 NDPC 2014, 115. 114 http://www.wfp.org/node/3467/3711/642726 (accessed March 23, 2015). 115 NDPC 2013, 150; GoG 2014b, 43. 116 http://www.presidency.gov.gh/. 117 Ampratwum, Armah-Attoh, and Ashon 2012. The study was based on answers to a survey by 2 officials at the national level (GES financial controller, and logistics and supply officer) and 12 at the district level (6 District 38 recordkeeping of books delivered and received at all levels, the study concluded that the poor districts experienced more leakage of books than the endowed districts and that in 2008/09 the leakages in English and science text books from Ministry of Education/publisher to DED reached 10 percent and 8 percent, respectively. No mathematics books were distributed that year. Leakages from DED to schools could not be estimated in 90 percent of the cases.118 Many parents in Ghana still need to spend a significant amount of money to send and keep their children in public schools, in spite of the GoG’s efforts to make public basic education free. Elimination of Child Labor. As discussed, 21.8 percent (1.9 million) children aged between 5 and 17 years are actively engaged in child labor, and 14.2 percent (1.2 million) are engaged in hazardous work. The Ministry of Employment and Labor Relations (MELR), in collaboration with ILO/International Program on the Elimination of Child Labor (IPEC), has implemented a program to eliminate the worst forms of child labor (WFCL) in Ghana, including in the cocoa industry. The National Plan of Action (NPA) for the Elimination of the Worst Forms of Child Labor of 2009 provides a comprehensive framework to significantly reduce the worst forms of child labor. The Ghana Shared Growth and Development Agenda (2010–13) incorporates the NPA and includes child labor as an issue in its Child Protection and Development focus area. Likewise, the Education Strategic Plan (2003–15) includes alternative education for children who are out of school, hard to reach, or in urban slums. The Labor Department of the Ministry of Employment and Labor Relations coordinates all interventions relative to the eradication of all worst forms of child labour. It operates through Community Child Protection Committees (CCPCs) that are active in more than 600 communities nationwide. As part of its annual work plan, the Department of Labor supported 7,825 child workers in 2014 and plans to support 8,523 child workers in 2015.119 Labor-Intensive Public Works (LIPW). Implementation of this program began in 2011 with support from the World Bank’ Ghana Social Opportunities Project.120 This program aims to provide targeted rural poor households with access to employment and income-earning opportunities, particularly during the agricultural off-season from November to March/April and in response to external shocks, through rehabilitation and maintenance of public or community infrastructure. LIPW provides scalable quick-response mechanisms against shocks, such as floods or droughts or during a crisis. The scope of works eligible for LIPW is defined based on their labor content and scope to generate significant local employment. The public works include rehabilitation and maintenance of feeder roads, small dams and dugouts, and soil and land conservation works. The LIPW program started in 49 districts in all regions of Ghana, with a concentration in the three northern regions.121In 2014, 147 subprojects were completed with an investment of GH¢38.5 million; 445 subprojects haven been completed since inception. The program is managed by the Ministry of Local Government and Rural Development (MLGRD). Directors and 6 logistics and supply officers), as well as 339 teachers, parents, and other school officials, and 6 focus groups at the district level. 118 The study could not determine if there was leakage of excesses from the Ministry of Education /publisher to districts in English and science books in 81 percent and 69 percent of the cases, respectively; or from districts to schools in English and science books in 90 percent of the cases, each. 119 MELR 2014, 43. 120 Ghana Social Opportunity Project (US$89.1 million from 2010 to 2017) and Additional Financing (US$50 million from 2014 to 2017). 39 121 World Bank 2014b, 5. LIPW activities are selected from district development plans and District Assemblies are responsible for implementation. The selection of beneficiaries uses a combination of geographical and community-based targeting mechanisms and self-selection. First, poor rural districts are selected on the basis of the incidence of food vulnerability and poverty; districts also need to have medium-term development plans. Second, interested persons self-select to participate in the program. In some project areas, particularly the poorest areas, there has been oversubscription of workers (as many as 500 beneficiaries per subproject instead of the expected 150). The wage paid of GH¢ 7 for a six-hour working day is similar to the minimum wage of GH¢7 for an eight-hour working day effective January 1, 2015. Workers are paid every two weeks. Where there is excess demand for the program, the program is now relying on the community to help select the beneficiaries through community-based targeting method, with no more than 200 people selected by the community to participate in the works per site.122 In 2014, the program employed 30,042 unskilled persons (of which 65 percent were women), providing on average 89 days of work and GH¢461 in payment, with a total investment of GH¢38.5 million. While the participation of women in the program at over 60 percent has been quite high, some observers have indicated the need to improve the conditions for childcare in working sites according to the standards contemplated in the program’s operational manual.123 The impact of the program has been positive. The impact evaluation by the Institute of Statistical Social and Economic Research (ISSER) undertaken from July to October 2012 in 11 districts confirmed the positive impact of LIPW. In particular, LIPW households reported an increase in paid employment of 3.1 percent between January and March 2012 over nonbeneficiaries. This impact was even more significant (11.3 percent) for beneficiaries 24 to 34 years old. Expenditures on food increased in beneficiary households. Expenditures on clothing, housing, and other expenses were reported to increase among female-headed households. The level of indebtedness for households with youth aged 25–34 years declined by 4 percent, and the likelihood of having savings (either at home or at financial institutions) increased by 6 percent. The use of curative care was also higher among LIPW adults (13.7 percent). LIPW reduced the hours of schools missed by children by 3.6 hours and reduced the likelihood of grade repetition by 0.5 percent. 124 As mentioned, the World Bank is providing Additional Financing (AF) to the Social Opportunities Project.125 The basic design of the LIPW component and menu of subprojects will be maintained. The program is being expanded to 60 districts. The AF program is also addressing key implementation challenges, including: (i) oversubscription at many sites; (ii) delayed payments as a result of the use of manual systems to record time sheets and prepare payrolls; (iii) weak communication of project results; and (iv) inadequate social accountability and grievance redress mechanisms.126 The program has a strong capacity-building component. The objective is to create capacity at the national and local level to implement LIPW projects in the selected 122 Interview with program management on March 3, 2015. 123 Interview with ILO staff on March 9, 2015. 124 World Bank 2014d, 2. 125 World Bank AF Ghana Social Opportunity Project (P146923) of 2014 (US$50 million) and Japanese Social Development Grant (US$3 million) for a graduation pilot. 126 World Bank 2014d, 11. 40 districts and thereby to strengthen the government's decentralization program and enable the related program strategy to be introduced nationwide. Other activities are being developed, including a social accountability and grievance redress mechanism, development communication activities, capacity building for the introduction of ICT-based operational enhancements (such as electronic registration and electronic timesheets linked to e-payments), and impact evaluation. The Ghana Social Opportunity Project (GSOP) is initiating a pilot project to upgrade the skills of selected LEAP and LIPW beneficiaries, to facilitate their graduation from the programs. The World Bank, through the Japanese Social Development Fund (JSDF), is supporting a two-year Complementary Income Generating Skills pilot project. The goal of the project is to improve incomes of poor households by supporting them to manage their farm and nonfarm activities more productively and sustainably. Through this project, 6,400 LIPW and LEAP beneficiaries are being trained to enhance their entrepreneurial skills, as well as provided with tools in order to earn a sustainable livelihood and eventually graduate from the programs. The pilot will be conducted in eight districts in the Upper East Region with Japanese financing (US$ 2.7 million). The GoG is also conducting a similar pilot with 1,000 beneficiaries in four districts in four regions— Volta, Brong Ahafo, Northern, and Upper West. The pilot will be evaluated for results upon completion. National Health Insurance. The National Health Insurance Scheme (NHIS) was established in 2003 to provide universal health coverage to Ghana’s population. Before 2003, Ghana had a proliferation of community-based health insurance (CBHI) schemes. In 2003, under Act 650, most CBHI schemes transitioned into district mutual health insurance schemes (DMHISs). By 2012, Ghana had about 145 DMHISs. These schemes were regulated and supervised by the National Health Insurance Authority (NHIA). In 2012, the National Health Insurance Act 852 integrated all DMHISs into a unified NHIS, creating a bigger risk pool and resolving some of the governance and administrative challenges of Act 650.127 Earmarked revenues are the main source of financing for the National Health Insurance Fund. These come from 2.5 percentage points of the value-added tax (VAT) and 2.5 percentage points of the payroll contributions from the SSNIT pension scheme. The NHIS manages the scheme and has the following functions, among others: (i) determining the contributions to the scheme, in consultation with the Minister of Health; (ii) registering members of the NHIS; and (iii) granting credentials to health care providers and facilities that provide health care services to members of the NHIS. The law requires all Ghanaians to enroll in the scheme, but only 39 percent of the population had done so as of 2014. The active membership of NHIS was 10.5 million at the end of 2014, up from 8.9 million in 2012. The Ministry of Health has identified consumers’ ability to pay the premiums as a deterrent for enrollment in NHIS, as well as the cost of medicine, the quality of care at health facilities, and long waiting time at those facilities.128 Of the active membership (10.5 million), 66 percent (6.9 million) were exempt from payment of premiums, while 34 percent were premium-paying members.129 Categories exempt from premium payments include children under 18 years (4.7 million), pregnant women (286,596), 127 Otoo and others 2014, 3. 128 Ministry of Health 2014, 44. 129 Information received from NHIS. 41 persons 70 years and over (280,157), and indigents (1.5 million).130 LEAP beneficiaries and SSNIT contributors and pensioners are also exempt from premiums. Only pregnant women, indigents, and LEAP beneficiaries are exempted from all fees (registration fees and premiums). Exempted persons under 18 represent 45 percent of the membership. The responsibility for identifying those who qualify for the indigent exemption lies with the MoGCSP’s Department of Social Development. Following the negotiation of a Memorandum of Understanding between the Ministry of Health and the LEAP program in 2013, the NHIS started to register LEAP beneficiaries automatically. NHIS estimates that 95 percent of LEAP households are registered in NHIS, though membership renewal is only about 50 percent.131 The annual premium ranges from GH¢7.20 (US$2.50) to GH¢48.00 (US$17), depending on income. All active members have a biometric registration.132 Assuming that all those exempted would have paid GH¢7.20 in premiums, the cost of the exemptions in 2014 are estimated at GH¢48 million. The NHIS has a single benefit package that it describes in its website as covering 95 percent of the disease conditions that affect Ghanaians. Benefits include outpatient and inpatient services and drugs on the NHIS medicines list. The NHIS package excludes certain forms of surgery, cancer treatment, organ transplants, dialysis, nonvital services such as cosmetic surgery, and some items such as HIV antiretroviral drugs, which are provided by the National AIDS Program at highly subsidized rates. In addition to the premium, NHIS does not require any cost sharing, annual or lifetime limits, or other limitation on demand for services. De facto limitations arise, however, from the lack of certain services and pharmaceuticals, particularly in rural areas. The program’s impact on the poor has been limited because many are not enrolled. GLSS- 6 indicates that the poor have a lower rate of enrollment than the nonpoor: 49 percent of individuals in the poorest quintile are not registered with NHIS, compared to 39 percent in the richest quintile (Figure 3.1). The main reason given by the poor for not enrolling is because they cannot afford the premium. According to GLSS-6 data, the indigent are included in all income groups, which appears to indicate no clear definition of who is indigent and/or discretion in the selection. In this context, the GoG 2014b study indicates that NHIS appears to have improved financial health protection and access to health services for NHIS members, but to a lesser extent for the poor because many are still not enrolled in the scheme.133 With the increase in enrollment in NHIS by LEAP beneficiaries, this problem is being minimized. Indeed, according to NHIS data, the number of indigent covered has more than tripled since 2012. Renewals rates may continue to pose a challenge, however. 130 An indigent is defined as someone who is unemployed and has no visible source of income, no fixed residence, does not live with a person who is employed and has a fixed residence, and does not have any identifiable consistent support from any other person. Data received from NHIS. 131 Interview with NHIS in March 2015. The renewal rate from 2013 to 2014 was 70 percent. The renewal rate is for the overall membership, not only for LEAP beneficiaries. 132 MoF 2014a, 116. 133 GoG 2014b, 41. 42 Figure 3.1: Registration in the National Health Insurance Scheme by Quintile 42% 37% 39% 49% 45% 23% 27% 26% 21% 19% 31% 32% 34% 34% 33% Q1 poorest Q2 Q3 Q4 Q5 richest Registered with card seen Registered, card not seen or not valid Not currently registered Source: World Bank staff estimates based on GLSS-6. Note: Registered with valid card. Similarly, a recent evaluation of the Free Maternal Health Service Initiative found a continuing trend of increasing use of health facilities for deliveries by pregnant women, but found some lag among the poor women. The Free Maternal Health Service Initiative of 2008 exempted all pregnant women from paying the NHIS premium and all other fees. Enrolment entitles women to six antenatal visits, childbirth care (including care for complications), and two postnatal visits within six weeks of childbirth, care of the newborn for up to three months, and other primary health care benefits. The initiative has helped increase the share of supervised deliveries from 24.3 percent in 2012 to 37.5 percent in mid-2013 and the rate of prenatal care to 46.3 percent.134 Nonetheless, the uptake from poor women has reportedly been “modest, highlighting other barriers to accessing health care such as lack of information, and difficulties in covering non-medical expenses (i.e. transportation).” 135 LEAP 1000 would help relax these constraints for an initial 6,000 beneficiaries. Social Welfare Services. The MoGCSP’s Department of Social Development (DSD) is responsible for implementing LEAP. It is also responsible for setting standards and managing nationwide social welfare programs. With the decentralization process, social welfare and community development services were devolved to the districts (see Chapter IV). The DSD is responsible for children shelters, correction institutions, and vocational training and rehabilitation centers. In 2014, excluding LEAP, DSD spent GH¢1.6 million.136 Its main programs include: 137 134 Otoo and others 2014, 3. 135 World Bank 2014a, 4. 136 This information was provided by the Department of Social Development. It should exclude the staff that works in the DA and are being paid by Ghana Local civil service. 137 MoGCSP 2014b, 49. 43  Child Rights Protection Program. DSD activities include care for orphaned and vulnerable children, which sometimes involves institutionalization. DSD provided care for 2,234 poor children in 2014 and planned support for 2,000 in 2015.  Counseling to children in conflict with the law. DSD processes juvenile court cases and adoptions, as well as supervision of adolescents placed on supervisory orders and liaison duties for the courts. In addition, DSD provides counselling, hospital welfare services, case work, and assistance to street children. About 3,000 children in conflict with the law were assisted in 2014, and 4,000 were expected to be assisted in 2015.  Welfare service to elderly persons in need. DSD coordinates the implementation of the Ageing Policy and Action Plan launched in 2011. The Ministry intends to propose legislation to Parliament to establish the National Council on Ageing. In 2014, the DSD supported 3,672 elderly persons and planned to support 3,700 in 2014, through payment of hospital bills, pauper burial, and psychosocial support. A number of destitute persons were cared for at the Bekwai Destitute Infirmary. In early January, 2015 the MoGCSP launched the EBAN Elderly Welfare Card to enable Ghanaians aged 65 and above to have priority access to social services such as transport, health (NHIS), and financial services.  Rehabilitation of persons with disabilities (PWD). In 2014 DSD provided skills and vocational training to 1,165 PWD and planned to support 900 in 2015. The District Assembly Common Fund provides 2 percent of its resources to the National Council on Persons with Disabilities (NCPD) to support its activities. The budget for 2014 was GH¢10 million. Among other things, the NCPD is expected to establish business incubators for PWD: 10 in 2015 and 40 in 2016.138 There is significant overlap of functions between MoGCSP’s DSD and the District Assemblies’ Departments of Social Welfare. The DA’s DSD are involved in most of the previous areas, and it is not clear where the function of MoGCSP’s DSD ends and DA’s DSD begin (see Chapter V). 2. Scholarships and Subsidies to Senior High Students Scholarships/Subsidies to Senior High Schools. Most GoG scholarships are awarded by the Scholarships Secretariat (SS), which depends directly on the Office of the President of the Republic. GoG also provides subsidies to students in senior high schools (SHS) (in general and technical education), known as GoG subsidies to SHS. SS scholarships used to be financed by the Ghana Education Trust Fund (GETFund), but since 2011 they have been financed by the Consolidated Fund (see below). GETFund is now running its own scholarship program. Total investment in scholarships by the SS in 2014 was GH¢117 million, financed by the Consolidated Fund. This includes arrear payments from adjustments to the amount of subsidies paid retroactively, new schools that qualify for subsidies, and past shortfalls in MoF disbursements. For the 2015, the planned investment is GH¢150 million. In 2014, GoG invested GH¢67 million in SHS subsidies, and GETFund invested GH¢37.4 million in scholarships, for a total GoG investment in scholarships and subsidies of GH¢222 million. 138 MoGCSP 2014b, 52. 44 The Scholarship Secretariat has an important scholarships program at the secondary level. These scholarships are designed to finance boarding (lodging and food) of students who must leave their homes and villages to attend SHS. In some cases, the scholarship also pays for the cost of final examination and offers other assistance. The Scholarships Secretariat provides four types of scholarships. (i) The Northern Scholarship is given to all SHS students in boarding schools with at least one parent born in northern Ghana. For those attending school in the Northern Region, the award is automatic on the basis of a list submitted by the school. Those living outside the region must apply. The Northern Scholarship was instituted to bridge the educational development gap between the South and North of the country, by increasing access by children of northern Ghana to secondary education. The award package includes feeding grants and final examination registration fees of beneficiaries. (ii) The Hardship Scholarship is given to brilliant but poor students in the regions other than the three northern regions, on the basis of economic and social circumstances of the parents: low income, incapacitation, and/or deprivation. Beneficiaries are proposed by the schools and District Assemblies. (iii) The Merit Scholarship is given each year to the three best students of each program in each public SHS. Awards cover the feeding grant only. (iv) The President’s Special Independence Award is given by the President of the Republic every year during Ghana’s Independence anniversary celebrations to deserving junior high school graduates who excel in BECE examinations (one male and one female from each of the ten regions of the country). The award package includes cash awards and feeding grants (for the entire duration of senior high school education). In 2014, about 113,689 students benefited from these scholarships. At the tertiary level, a variety of scholarships are awarded: (i) Tertiary Local awards consist of bursaries and thesis allowances for postgraduate students in public universities; grants for physically challenged students in public universities and polytechnics; and long stay awards for medical students in their clinical years of study. (ii) Tertiary Abroad awards are for undergraduate and postgraduate studies under existing bilateral cooperation agreements between the government of Ghana and country donors. In 2014, about 23,080 tertiary education students benefited from these scholarships: 21,935 in local institutions, and 1,145 abroad. The Ghana Auditor General’s performance audits of the scholarships programs highlighted several challenges facing the program. For the SHS program, the report indicates lack of accountability of all players involved and no clear established written rules and regulations for the program. The program is run on a paper basis with minimum use of the internet. Problems were found with the selection of beneficiaries (ineligible beneficiaries, ghost beneficiaries, beneficiaries receiving more than one award), long delays in processing the application and disbursing funds, and lack of monitoring of the program.139 For foreign scholarships, an early report also indicated several challenges: lack of information about scholarship offerings, inadequate funding, and ineffective monitoring of students abroad, ineffective bonding, and weak incentives for student to return to their place of work.140 While these reports are somewhat dated (2006 for the SHS, and 2003 for the foreign scholarships), no evidence was found that their findings are no longer valid. The Scholarship Secretariat conducted two monitoring activities in foreign and local institutions in 2014, but results are not yet available. 139 GAG 2006. 140 GAG 2003. 45 GoG subsidies to SHS have grown over the years, with enrollment and adjustments to the unit transfer. This is an amount of money per term per student that is “agreed” with Conference of the Heads of Assisted Secondary Schools. The subsidy is paid on the basis of enrollments reported by the MOE. The annual subsidy per student more than doubled from GH¢ 40.35 in 2008 to GH¢92.4 in 2014. The GoG investment increased during the same period, from about GH¢14 million to GH¢67 million. No information is available on whether these subsidies overlap with SS subsidies to SHS. As with other programs, there is been substantial delay in releasing the funds to the schools. The Ghana Education Trust Fund (GETFund) has been running its own scholarship programs since 2011. Until 2010, GETFund financed the Scholarship Secretariat programs, but in 2011 it decided to discontinue those and established its own program. In the last three years, GETFund has invested about GH¢40 million a year in scholarships. In 2013, GETFund announced that it was suspending its program of scholarships abroad “to prevent the recurring situation of students on GETFund scholarships outside the country getting distressed because of the government’s failure to pay both their tuition fees and other allowances” because of delays in transfers for GETFund.141 Most District Assemblies also provide scholarships and bursaries to their residents. This assistance is usually for senior high school and tertiary students. Most of the resources to finance these scholarships are from the District Assembly Common Fund and are budgeted under social services. In 2013, District Assemblies spent GH¢2.3 million on scholarships, equivalent to 11 percent of their social services budgets (or 1.7 percent of their overall budget).142 The Cocoa Board has an important scholarship program for the children of cocoa farmers. The Ghana Cocoa Board was established in 1947 with a mission to encourage and facilitate the production, processing, and marketing of premium quality cocoa, coffee, and Shea nuts. Under its corporate social responsibility initiatives, it has instituted a Cocoa Scholarship Award Scheme as part of its welfare services to cocoa farmers for the secondary education of their children. On average, it sponsors about 7,500 students per year for three years.143 Development partners also finance scholarships. Among these are the Girls-Participatory Approach to Student Success program (Girls-PASS) and Secondary Education Improvement Program (SEIP). Girls-PASS, financed by DFID (£46.3 million from 2012 to 2016) is helping 81,000 girls access and complete secondary education by providing scholarship packages to 31,000 girls through the Campaign for Female Education (CAMFED) and at least 50,000 girls through the Ghana Education Service (GES) in 75 poor districts.144 In 2013/14, CAMFED delivered 24,300 scholarships to girls in junior high school (11,700) and senior high school (12,600). The GES delivered 9,996 JHS scholarships in 21 pilot districts and is currently extending this to an additional 45,000 JHS girls in all 75 poor districts during the 2014/15 school year.145 Girls-PASS program disbursed GH¢36 million in 2014/15. The World Bank-supported Secondary Education Improvement Program (SEIP) (US$ 156 million from 2014 to 2019) will also provide 141 http://www.adomonline.com/news_details.php?article=2476. 142 Social services include basic education, teacher training, cultured, sports, water, health, disaster management, and community initiative projects. 143 https://cocobod.gh/social_esponsibility.php. 144 MoF 2014a, 14. 145 DFID 2014. 46 10,000 scholarships to poor students at the senior high school level. The loan became effective in October 2014. 3. Social Insurance Ghana has a three-tiered social security system. The National Pensions Act 766 of 2008 reformed the pension system and established a three-tiered scheme: (i) a mandatory basic first-tier social security scheme managed by SSNIT; (ii) a mandatory fully funded and privately managed second-tier occupational scheme; and (iii) and a voluntary fully funded and privately managed scheme (third tier). The National Pensions Regulatory Authority supervises and regulates the operations of the mandatory, occupational, and voluntary systems, including SSNIT, Licensed Trustees, Registered Pension Fund Managers, and Pension Fund Custodians, and ensures the effective administration of pensions in the country. The first tier is managed by the Social Security and National Insurance Trust (SSNIT). Affiliation in SSNIT is mandatory for all public and private employees. The only main exceptions are members of the armed forces and police.146 Employers are responsible for ensuring that everyone who works for them is registered. Self-employed workers can join on a voluntary basis. The 2008 reform increase the contribution from 17.5 to 18.5 percent of wages and reduced contributions to SSNIT by 4 percentage points, and earmarked this contribution plus the 1 percent increase to the privately managed Tier II scheme. The current contribution to the pension system is based on the employee’s salary, and is 13 percent paid by the employer and 5.5 percent paid by the employee. Contributions to the Tier I scheme are 11 percent of wages (5.5 percent paid by the employer and 5.5 percent paid by the employee). Contributions to the Tier II scheme are 5 percent of wages (all paid by the employer). Contributions to the National Health Insurance program are 2.5 percent of wages (all paid by the employer). Self-employed workers contribute 18.5 percent to the system. The second and third tiers are privately administered. They provide the worker with a lump sum on termination of service, death, or retirement, and if desired a pension through the purchase of an annuity, though the annuity market in Ghana is still underdeveloped. The Informal Sector Pension Scheme (ISPS), which is part of the voluntary third tier, was created by the SSNIT in 2005. It is open to workers who are not covered by the mandatory pension scheme under SSNIT. Half the contributions go to a savings account, and the other half goes to a retirement account. A member can withdraw funds from the retirement account only in the events of old age, disability, or death. A member can withdraw funds from the saving account after five months. A large number of members registered in SSNIT do not pay their contributions. Of those registered in the scheme, about 40 percent are classified as inactive members because they have not paid their contributions in a 12-month period. SSNIT active members in 2014 corresponded to 7.5 percent of the working-age population (Table 3.2). Contribution density is low, as only about 72 percent of active private sector members pay their contributions regularly. Table 3.2: SSNIT Membership and Pensioners 2010 2011 2012 2013 2014 146 Exemptions also include the military forces and those who joined the scheme before 1972 (article 71 of the Constitution). 47 A. Registered members 1,327,658 1,390,945 1,478,755 -- -- 1.Active members 900,332 963,619 1,051,429 1,120,512 1,189,168 Public 366,642 402,664 -- 447,400 457,548 Private 533,690 560,955 -- 673,112 731,620 Number who pay regularly 685,073 437,978 479,529 508,041 530,689 2. Number of establishments 34,360 36,534 40,664 42,946 46,595 Public 12 12 12 12 12 Private 34,348 36,522 40,652 42,934 46,583 Number that pay regularly 17,767 20,916 21,196 22,812 24,430 3. Active member/working age (%) 6.4 6.7 7.1 7.3 7.5 B. Number of pensioners 107312 112,522 119,323 129,691 142,076 Retirement (60+) 940 916 1,042 1,099 1,072 Retirement (60+) + 25% L/S residual 64,062 68,748 74,409 83,968 94,874 Early retirement (55–59) 1,015 1,537 2,700 4,260 6,565 Early retirement + 25% L/S residual 40,325 40,310 40,132 39,333 38,490 Invalidity pension 77 117 163 206 265 Invalidity pension + 25% L/S residual 893 894 877 825 810 Number of pensioners/pop 60+ (%) 6.5 6.7 6.8 7.2 7.7 Memorandum: Working age population (15–64) 14,040,893 14,434,573 14,875,154 15,328,101 15,793,166 15–54/ total population (%) 56.9 57.2 57.6 58.0 58.4 Population 60+ 1,643,381 1,690,763 1,743,182 1,797,088 1,838,930 Total population 24,658,823 25,235,268 25,824,920 26,427,760 27,043,093 Source: SSNIT; Ghana 2010 Population and Housing Census. L/S = lump sum. Few elderly people in Ghana retire with a pension, and one in four is poor. As discussed, only about 142,000 people 60 years and over (just 7.7 percent) had a pension. The incidence of poverty and extreme poverty among the population 60 years and over is 24 percent and 8 percent, respectively. 4. Active Labor Market Programs Active labor market programs (ALMP) should help address the challenge of youth unemployment. ALMP offer a series of services to firms and job seekers to facilitate their adjustment to changing labor market conditions and facilitate the transition from school or unemployment to stable quality jobs. These services include (i) labor market information; (ii) job search assistance; (iii) career counseling; and (iv) special programs for the unemployed, including second chance programs and youth employment programs. In Ghana, the Labour Department under the Ministry of Employment and Labor Relations (MELR) administers over 60 employment centers throughout the country. These centers reportedly have very little activity and impact on the employment situation. Labor market information is run manually and cannot help match job seekers and firms in a timely manner. The MELR suffers from severe financial and human resources constraints. The National Employment Policy focuses on 17 priority actions, including the development of a labor market information system (LMIS), as well as reforms in education, skills development and training, and promotion of youth employment.147 The ILO and the World Bank are supporting the MELR in setting up an effective labor market information system. Major youth employment program are briefly reviewed in the discussion that follows.148 Box 3.1 presents some of the many nongovernmental organizations (NGOs) that are also active in the field in Ghana. 147 Schwgler-Rohmeis and others 2014, 60. 148 Ghana has a large number of public and private programs that provide skills upgrading and technical and vocational and educational training (TVET) and over 200 public and 450 private TVET institutes through the country. Here the 48 Box 3.1: Selected Youth Skill and Entrepreneurship Development NGO Programs Several nongovernmental organizations (NGOs) and nonprofit organizations actively provide support to unemployed youth and promote skill development, entrepreneurship, and access to jobs or creation of microenterprises. These organizations try to operate in coordination with government programs. NGOs include, but are not limited to:  The Meltwater Entrepreneurial School of Technology (MEST) and the MEST incubator  Innovations for Poverty Action  Don Bosco Technical Institute  Ghana Skills Development Initiative (GIZ, German International Cooperation)  Technoserve (international nonprofit economic development organization)  The Junior Achievers Trust International  GRATIS Foundation. The National Apprenticeship Program (NAP). This program is managed by COTVET (Council for Technical and Vocational Education and Training), which is responsible for coordinating both formal and informal apprenticeships. The largest provider of skills training in Ghana remains the informal apprenticeship system, which trains more than 440,000 youth at any one time; there are about four informal apprentices for every trainee in formal public and private training centers combined. The NAP, which started in 2011, is the GoG’s latest attempt to improve informal apprenticeship by providing youth with on-the-job training with master-craftsmen. It targets junior high school graduates. In 2011/12, the program included 5,000 beneficiaries; in 2013/14, it covered 6,000 beneficiaries. The program runs in 70 districts in the country. For 2015, it plans to cover 7,000 youth in 100 districts. The program provides tools to the trainees and pays a stipend to the master-craftsman (GH¢150 per trainee per year). The NAP is a one-year training program. After the training, beneficiaries are encouraged to start their own businesses. COTVET has partnered with trade associations and district education offices to help administer the program. They help identify potential beneficiaries and master-craftsmen. In 2011/12, the program disbursed GH¢0.9 million to buy tools for the trainees. The program reports GH¢0.8 million are in arrears to the master craftsman for the two batches of beneficiaries. The GETFund is directly financing the purchase of equipment and the districts are also helping finance the program. USAID is financing a randomized control trial to evaluate the effectiveness of the program, which is being conducted (from 2012 to 2017) by the Jamel Poverty Action Lab. The National Vocational Training Institute (NVTI). NVTI was established by an act of Parliament (Act 351) of 12 January, 1970 to provide demand-driven employable skills and enhance the income-generating capacities of basic and secondary school leavers, and other persons through competency-based apprenticeship, testing, and certification, as well as career development. NVTI receives approximately 6,000 applications per year, while a maximum 5,000 participants per year are allowed in the program for formal training. The current employability rate after the program is 3 to 56 percent. NVTI is operating in 34 institutes throughout the ten regions of Ghana. The focus are on the publically financed programs that seek to reach poor youth, school drop-outs, or recent graduates who are jobless or unemployed and that provide basic skills as well as job placement opportunities. With the exception of the National Apprenticeship Program, which is a new program, all other programs are included in GoG 2014b. 49 institute provides a four-year training course for school-based apprenticeship and a one-year short program for informal sector workers. The latest annual budget was about GH¢3 million. Local Enterprises and Skills Development Program (LESDEP). LESDEP was created in 2011 as a result of partnership between the Ministry of Local Governments and Rural Development and the private sector to improve the skills of informal sector workers and create viable small enterprises. There are no age limits for the beneficiaries, though the program focuses on youth. The program provides entrepreneurial skills, start-up equipment, access to credit, and support following the start-up. Equipment is financed by loans, to be used as a revolving fund. The program operates nationwide through national, regional, and district steering committees. The program contracts out master-craftsman, who receive GH¢2,000–2500 for the six-month apprenticeship training period. Loans to graduates are administered by a private banking institution. Specialized training is conducted in collaboration with the Ministry of Employment and Labor Relations and other agencies. Largely funded by the government, LESDEP supported 15,005 youth in 2014 and planned to support 32,100 in 2015. Beneficiaries of the LESDEP program are given an average start-up capital of GH¢1,000, in a form of a loan, payable over a 6– 24 month period at a repayment rate of 50 percent. The businesses established through LESDEP include transport services, water services, catering services, mobile telephone and laptop assembly and repair, sales vans, farming equipment services, dressmaking and fashion design, beauty care, barber services, fruit juice processing, fishing gear, construction materials, and canopy and chair rental. Demand for the program is high, but there are no clear eligibility criteria for selecting those that are ultimately enrolled in the scheme. There have been problems in providing equipment in a timely manner to those who completed the training and are ready to set up their businesses. Repayment of loans also constitutes a challenge, which jeopardizes LESDEP's sustainability. Youth Employment Agency. In February 2015, the Parliament established the Youth Employment Agency to replace the Ghana Youth Employment and Entrepreneurial Development Agency (GYEEDA). The National Youth Employment Program was created in 2006 to address the growing youth unemployment problem and was renamed GYEEDA in 2012. The program was one of the major youth program in Ghana. It focused on disadvantaged youth and provided job placement, skill training, and seed capital to establish micro-enterprise to graduates from junior and senior high school and from technical and vocational schools as well as school drop-outs and illiterate youth. The program is being redesigned under the MELR and it plans to cover 600,000 youth in 2015. There are three new components and support services:149  Skills training and internship: This component focuses on identification of skills in demand in the labor market (in the short-, medium-, and long term) and the provision of requisite training using the formal and informal structures.  Entrepreneurial training: This component aims to provide sufficient entrepreneurial skills to all target beneficiaries so they will be ready for self-employment in both the formal and informal segments of the economy. 149 MELR 2014, 22. 50  Cooperatives system: Under this component, beneficiaries of skills training and internship programs will be assisted to form cooperatives/ trade associations to enable them to access support to use their skills and trades to generate and earn income.  Support services (accessing financial support): Organized and registered cooperatives and trade associations will be supported through enterprise funds in the form of tools, equipment and innovative ways of financing. B. Energy Subsidies The GoG has embarked on the difficult task of reducing existing energy subsidies. According to government authorities, fuel subsidies and electricity subsidies are being reduced, with the exception of the so-called lifeline, a subsidy to the households that consume less than 50 kWh per month. A brief analysis of electrify and fuel subsidies and implication for social protection follows. The analysis is limited, as it is based only on information readily available on the websites and from interviews with officials of the National Petroleum Authority (NPA) and authorities of the Public Utility Regulatory Commission (PURC). Additional data and information would make possible a more detailed analysis of the situation.150 Electricity Subsidies. The structure of electricity tariffs is revised every two to three years and is adjusted every quarter for changes in basic cost parameters (including related macroeconomic parameters), according to the Public Utility Regulatory Commission. The last structural adjustment was in 2013.151 At that time, the implemented structure contained an 11 percent discount over what would be required to recover costs. Since then, quarterly adjustments have taken place, but have not covered this initial implicit subsidy, which is costing the operators about GH¢200 million per year.152 The current tariff structure includes a subsidy to households that consume low levels of electricity and cross-subsidies among the different type of consumers. The lifeline tariff is GH¢0.20 per KWh for consumption levels of up to 50 kWh/month (Table 3.3). The reduced tariff does not apply to those households that consume higher levels. A World Bank study on targeting found that the share of the lifeline tariff subsidy going to the poor was only 8 percent, and most of the beneficiaries were located in urban areas.153 Indeed, according to GLSS-6, only 25 percent of households in the poorest quintile use electricity (see also Chapter IV, Section C). The World Bank study suggests that connection subsidies are better targeted to the poor, compared to subsidies on electricity consumption. The current connection cost of GH¢400 (US$143) is prohibitive for most poor households, as it is equivalent of 217 percent of the monthly food consumption of households in quintile 1. Therefore, many poor urban households often resort to 150 The information we requested from NPA and PURC was the annual cost of the subsidy and the amount settled by MoF in recent years. In addition, from PURC, we requested the current cost recovery tariff level and consumption and income by tariff blocks to analyze the cross subsidies. 151 The damage of the Nigerian section of the West African Gas Pipeline in 2012—which supplies Ghana (as well as Benin and Togo) with natural gas from Nigeria—cut off the gas supply to Ghana and forced the country to increase oil imports to generate electricity, causing the import bill to rise dramatically by approximately US$27 million per month. The cost, however, was not passed through the consumer, so the price became well below the actual cost (benchmark price) and the subsidy peaked. 51 152 Information provided during an interview on March 3, 2015 with PURC senior officials. 153 Wodon 2012. illegal connections from neighbors, paying higher rates and subjecting their households to various hazards. Table 3.3: Electricity Tariff Structure, Effective October 2014 to March 2015 Type of consumer Residential Nonresidential Blocks (KWh/month) 0–50 51–300 301–600 601+ 0–300 301–600 601+ GH¢ per KWh 0.205393 0.412072 0.534790 0.594211 0.592403 0.630380 0.994658 Source: Public Utility Regulatory Commission. According to MoF budget outturns, subsidies to utilities peaked at 1.2 percent of GDP in 2013 and have since declined sharply (Table 3.4).154 These subsidies were estimated to have cost the budget the equivalent to 0.3 percent of GDP in 2014. For 2015, the budget includes the lifeline subsidy and no other subsidies for the utilities (as it did in 2014). GoG payment arrears to the utility companies are reportedly significant, though an estimate of their amount is not available. Table 3.4: Energy Subsidies Reported in the MoF Budget 2012 2013 2014 2015 Budget Provision Budget Actuals Budget Revised Budget -al projection outturn GH¢ millions Social benefits (Lifeline) 31 0 39 2 49 49 60 Utility companies 0 186 795 1,079 0 291 0 Petroleum products 0 623 228 79 50 327 50 % of GDP Utilities 0.0 0.2 0.9 1.2 0.0 0.3 0.0 Petroleum 0.0 0.8 0.2 0.1 0.0 0.3 0.0 Memo: GDP (GH¢ million) 74,959 74,959 93,867 93,867 114,654 114,654 135,011 Source: MoF budget statements. Note: Actuals refers to outturns registered in MoF Budget Statements and are not accrual amounts. The sector is facing a major cash flow crisis, which is affecting the steady supply of energy, productive activities, and sector investments. The sector has been debundled. Three public enterprises are the principal actors: the Electricity Company of Ghana (ECG), for distribution; GRIDCo, for transmission; and the Volta River Authority (VRA) for generation. The sector regulator is the Public Utility Regulatory Commission (PURC). Many public entities are not paying their electricity bills to ECG, which in turn is in arrears to GRIDCo and the generation companies. There are also arrears in the purchase of gas from Nigeria. Generation problems and lack of investment because of the sector cash flow problems have contributed to a supply gap of between 300 and 600 MW. Recently, the President of the Republic announced the restructuring of ECG with U.S. support (a US$500 million grant from the Millenium Challenge Commission) and an emergency program to increase the supply by 1,000 MW in the short run.155 In addition, 154 An IMF study (IMF 2013, 54) estimated Ghana’s electricity sector quasi-fiscal deficit at 2.3 percent of GDP. Quasi-deficit is defined as the difference between the actual revenue charged and collected at regulated electricity prices and the revenue required to fully cover the operating costs of production and capital depreciation. 155 http://www.presidency.gov.gh/. 52 the Energy Fund Levy on Petroleum Products will be increased from GHp 0.05 to GHp 1.00 part of which will be used to establish a Renewable Energy Fund. Also, a services charge of GHp 1 on kWh of electricity transmitted will be established to finance solar energy. 156 From a social protection perspective, the challenge is to increase access to electricity for low- income families by subsidizing the connections, and ensuring the reliability and quality of supply for those who already have access. The disruption in electricity supply is adversely affecting basic services (for instance, for hospitals and the water company) and small-scale informal sector activities, which have to rely on expansive alternative sources of energy (such as generators) or cease activities. The public sector electricity payment arrears create contingent liabilities for the GoG, which may have to be settled eventually and then will reduce the fiscal space for other public expenditures, including social protection. Therefore addressing these issues is an urgent matter. Petroleum Products Subsidy. Since 2004, the GoG has been deregulating the oil industry. In 2005, it introduced a price formula for oil products and established the National Petroleum Authority (NPA) to review the prices twice a month. In 2012, there were no price adjustments. The gap between domestic and international prices increased substantially, and the subsidy reached 3.2 percent of GDP.157 Table 3.3 indicates that for 2012, MoF recorded only GH¢623 million as budget outlays with respect to fuel subsidies, equivalent to 0.8 percent of GDP, with the difference accruing to arrears or being absorbed by the different sector operators. Over the first half of 2013, the GoG adjusted the prices of petrol, kerosene, diesel, marine diesel, and liquefied petroleum gas (LPG) by between 15 and 50 percent. Prices reached market levels in mid-September 2013 (Figure 3.2). At the end of 2013, the price of petrol stood at its market rate, up almost 30 percent from the beginning of the year. Premix, used by fishermen, remains subsidised.158 As analyzed in Chapter IV, subsidies on petrol and diesel products mostly benefit the nonpoor, while kerosene subsides are more pro-poor. To compensate for the adverse impact of price hikes on the poor, the GoG’s 2013 budget included more than doubling of the budget for LEAP and a tripling of the School Feeding Program.159 156 http://www.presidency.gov.gh/. 157 IMF (2013, 54. Appendix Table 1 and 76). Estimated based on the price benchmark method. Fuel subsidy (tax) per liter is obtained by subtracting the relevant cost recovery benchmark price from the domestic retail price. Benchmark prices are computed by adding Cost, insurance and freight (CIF) fuel import prices, national margins, and costs (such as for transportation and distribution) and indirect taxes (VAT and excise taxes). 158 Cooke and others 2014, 7. 159 Cooke and others 2014, 5 and 6. 53 Figure 3.2: Fuel Price Adjustments (2008=100) 370.0 320.0 270.0 220.0 170.0 120.0 70.0 16-Jan-08 16-Jan-09 16-Jan-10 16-Jan-11 16-Jan-12 16-Jan-13 16-Jan-14 Gasoline Gasoil Kerosene Premix Source: National Petroleum Authority. Currently, the National Petroleum Authority sets indicative maximum prices to be charged to consumers; these prices take into account the cumulative cost structure of importing and distributing the products (Table 3.5). The ex-refinery price is calculated by taking into consideration the CIF price (Cost, Insurance and Freight), exchange rate, import tax, and refinery costs. In July 2014, the Energy Minister was reported as saying that the GoG was spending GH¢ 85 million biweekly on total fuel subsidies.160 Since international and domestic fuel prices have changed significantly since then, it is not sensible to extrapolate that amount to estimate current subsidies. The MoF budget fuel subsidies in 2014 are recorded at GH¢ 327 million (projected outturn), equivalent to 0.3 percent of GDP. According to NPA officials, in March 2015 the maximum prices fully reflected the cost recovery levels. Indeed, given the reduction in oil prices, there has been a lag in adjusting prices downward to compensate for the losses incurred in previous periods. Table 3.5: Indicative Maximum Price of Fuels, Effective January 2015 Premium Gasoil Kerosene LPG Premix (domestic) (fishing) Ex-refinery price 199.6427 201.5129 208.6808 211.5687 125.5008 Excise duty 2.7800 1.8000 1.0375 0.7246 Tor debt recovery levy 8.0000 8.0000 5.0000 Road fund 7.3231 7.3231 Energy fund 0.0500 0.0500 0.0500 Exploration 0.1000 0.1000 0.1000 0.1000 Cross-subsidy levya 5.0000 -2.6987 -4.8449 -18.4042 -0.3608 Primary distribution margin 6.5000 6.5000 6.5000 BOST marginb 3.0000 3.0000 3.0000 Fuel marking margin 1.0000 1.0000 1.0000 1.0000 Ex-depot 233.3958 226.5873 215.5234 198.8891 126.2400 Special petroleum tax 40.8443 39.6528 37.7166 34.8056 UPPF (unified petroleum price fund)c 9.0000 9.0000 10.0000 10.0000 9.0000 160 http://af.reuters.com/article/ghanaNews/idAFL6N0PO1HY20140713?pageNumber=2&virtualBrandChannel=0. 54 Marketers margin 12.8200 12.8200 12.8200 8.1500 12.5200 Dealers (retailers/operators) margin 8.9400 8.9400 8.9400 6.3800 8.9400 LPG filling plant/Premix/ local admin costs 4.7753 0.3000 Distribution compensation margin 1.0000 Promotion margin 7.0000 Indicative maximum price (ex-pump price) 305.00 297.00 292.00 264.00 157.00 Price in GH¢ per gallon and kg 13.725 13.365 13.140 2.640 Price in GH¢ for a 12.5 kg cylinder 33.0 March 2015 price at the pump 13.725 13.365 13.140 Memo: Prices in April 2013 9.50 9.50 25.0 Source: National Petroleum Authority. Note: All prices in Ghana pesewas per liter, except LPG in GHp/kg. Imperial gallon=4.5 liters. LPG= liquefied petroleum gas. a. Subsidy across fuels: This value is positive only for gasoline and negative for all other fuels. That is, the consumption of all other fuels is subsidized by the consumption of gasoline (the idea is to subsidize fuels mainly consumed by the poor). b. BOST is Ghana’s Bulk Oil Storage and Transport Co., Ltd. c. The Unified Petroleum Price Fund (UPPF) supports fuel transportation to rural regions while maintaining a countrywide uniform pump price; the goal is to ensure the supply in all regions. In sum, Ghana has an extensive safety net, including a unconditional cash transfer for the poor, OVC, the elderly, and people with severe disabilities; in-kind assistance programs (school feeding, free uniforms, and exercise books); a labor-intensive public works for the poor unemployed; and education and health insurance fee waivers. It has a three-tiered social security system that covers both the formal and informal sector. It has a number of youth programs, although active labor market services and second chance programs are limited. And it has large energy subsidies, which are gradually being reduced. 55 IV: Assessment of the Social Protection System On the basis of the analysis of groups at risks and existing social protection programs analyzed in the last two chapters, this chapter assesses Ghana’s social protection system, focusing on coverage, gaps, and overlaps; the generosity of benefits; the accuracy of targeting; cost-efficiency; sustainability; monitoring and evaluation; and institutional arrangements and coordinating mechanisms. A. Coverage, Gaps, and Overlaps The coverage of key social protection programs is low. Table 4.1 presents estimates of the size of various programs’ target groups and number of beneficiaries. Unfortunately, for some programs, the definition of the target group is not clear. Also, in a few programs that have an Operational Manual, the definition of target group may be clear, but it is not applied in actual practice. Therefore, what constitutes the target groups is, in some cases, open to interpretation. The characterization of the program in terms of coverage compares the actual number of beneficiaries with the target group. Thus this characterization in some cases may also be open to discussion. Moreover, some programs, such as LEAP and NHIS, have multiple target groups. This analysis takes the broader target group as the indicator for the characterization of program coverage: namely, all extreme poor families, in the case of LEAP; and the entire population, in the case of NHIS. Taking into account all these caveats, of the nine programs covered in the table for which a target group could be identified, four have low coverage and five have high coverage. Most importantly, coverage is low for three key programs—LEAP, LIPW, and NHIS. Programs with high coverage include capitation grants, school feeding, free uniforms and exercise books, and scholarships. 56 Table 4.1: SP Program Scale by Target Group Core social Characteri- Number of Targeting Age group/At-risk group protection and Target group zation of beneficiaries, 2014 mechanism ALMPs coverage Pregnant women a 950,000 women Low birthweight, 69,800 NHIS (pregnant Pregnant women Lactating/newborns, 286,596 Categorical n.a. women) 794,300b Children (0-17) 3,462,715 Extreme poor, 1,211,013 LEAP, children Stunted (<5 years) , Categorical/ Extreme poor children 83,238e n.a. <18 756,000 CBT/PMT Orphans (0-17) , 978,264 Preprimary, primary and JHS age Capitation grants 5.6 mc 5.6 m children Universal High school Extreme poor preprimary and School feeding 0.7 md 1.7 m children Universal High primary age school Take-home rations Geographical Poor girls students in JHS 145,000 (appr.) 30,000 girls n.a. for girls /categorical School uniforms 0.7 md 400,000 Categorical High Children in basic public education Free exercise d 0.7 m 4.8 m books Categorical High books Poor students in JHS Scholarships, JHS 362,370 21,696 Categorical n.a. Scholarships, SHS Geographic/ 115,779 12,000 girls n.a. (DFID) categorical Scholarships, SHS Geographic/c Poor students in SHS 115,779 113,689 High (SS) ategorical Girls, PASS 34,296 (2013/14) n.a. scholarships NHIS, children Children 12 m 4.7 million Categorical n.a. <18 years Youth (18-35) 1,450,781 Poor Poor youth attending tertiary f Scholarships (SS) 4,164 23,080 Categorical n.a. education LESDEP 275,950 15,005 Categorical n.a. Youth NAP 275,950 6,000 Categorical n.a. Adults (36-59 years) 1,079,348 Extreme poor rural HHs 313,000 Geographic/ self- selection/ Extreme poor in agriculture >15 30,042 Low LIPW 92, 49 community- years based targeting NHIS (indigents) 1.5 million Categorical n.a. Elderly (60 and over) 435,227 Elderly (60+) SSNIT 1.8 million 142,000 pensioners Categorical Low LEAP (65+ Categorical/ Extreme poor elderly (65+) 116,287 63,379 n.a. extreme poor) CBT/PMT 57 Elderly (70+) NHIS ( all, 70+) 911,176 380,157 Universal n.a. Poor families 1,081,189 Categorical/ Extreme poor families LEAP 357,822 90,754 Low CBT/PMT Entire population NHIS 26.6 million 10.5 million Universal Low Source: Program managers and staff estimates. Note: n.a. = not available; m=million; ALMP=active labor market program; CBT = community-based targeting; DFID = U.K. Department for International Development; JHS = junior high school; LEAP = Livelihood Empowerment Against Poverty Program; LIPW = Labor Intensive Public Works Program; NHIS = National Health Insurance Scheme; PMT = proxy means text; SHS = senior high school; SSNIT = Social Security National Insurance Trust; SS = Scholarship Secretariat. a. GLSS-6. b. UNICEF (http://www.unicef.org/infobycountry/ghana_statistics.html). c. All children in preprimary, primary, and junior high school public schools. d. Extreme poor children in public preprimary and primary public schools (data from GLSS-6). e. Includes 23,337 beneficiaries whose age is currently unknown (data from GLSS-6). f. The number refers to youth aged 18–35 who are below the poverty line and currently attending tertiary education, based on the GLSS-6. Old age pension coverage is also low. In Ghana, less than 8 percent of the elderly (60 years and over) receive a SSNIT pension. This compares to an average of over 20 percent for Sub-Saharan Africa and for the lower-middle-income (LMI) countries (Table 4.2). When comparing to Sub- Saharan Africa’s middle-income (MI) countries for which information is available, Ghana ranks lowest. If the number of pensioners is added to the elderly receiving LEAP grants, the coverage would be 11 percent. This means that approximately 90 per cent of older people are not receiving any pension benefits and thus are likely facing a high degree of income insecurity. Since most elderly live in families, this places an additional burden on poor families that are trying to make ends meet. Table 4.2: Old-Age Pension Beneficiaries in Sub-Saharan African Middle-Income Countries Contributory Non-contributory Total Year Ghana 7.7 3.5 7.7 2014 Botswana 100 100 2010 Cameroon 12.5 12.5 2011 Cabo Verde 18.2 37.5 55.7 2009 Congo, Rep. 22.1 22.1 2011 Côte d’Ivoire 7.7 7.7 2010 Lesotho 100 100 2010 Mauritania 9.3 9.3 2002 Mauritius 100 100 2010 Namibia 98.4 98.4 2011 Senegal 23.5 23.5 2010 South Africa 27.7 64.9 92.6 2010 Swaziland 86 86 2011 Sub-Saharan Africa 21.5 Low-income/middle- income 24.6 countries Source: GoG 2014b, Table B.6; SSNIT for Ghana. 58 SSNIT’s coverage of workers is also low. It is below the average for Sub-Saharan Africa and less than half the average for developing countries. If countries with universal or near-universal noncontributory pensions are excluded from the analysis (see Table 4.3), Ghana is in the middle of the range of its middle-income Sub-Saharan African peers. Even though the second tier is mandatory and everyone who contributes to SSNIT should contribute to this occupational system, as of February 2014, only 7,200 employers have, or about 17 percent of the 42,000 employers registered in SSNIT.161 Ghana’s National Pensions Regulatory Authority (NPRA) has no updated information on Tier II membership as of March 2015. SSNIT has taken important steps to correct this low social security coverage. As mentioned, in 2005 it piloted the third-tier voluntary pension scheme to cater to workers in the informal sector, who constitute the majority of workers in the country. The Informal Sector Pension Systems (ISPS) was formalized in 2008. By the end of 2011, ISPS had about 91,000 members and had accumulated GH¢23 million in contributions, according to the Ghana News Agency.162 According to NPRA, this fund has grown very rapidly in recent years; by April 2015, it had accumulated GH¢1,075 million in holdings. The NPRA has no updated information on Tier III membership as of March 2015. Table 4.3: Active Contributors to Pension Scheme in Sub- Saharan African Middle-Income Countries Active contributors/ Year working population, age 15–64 (%) Ghana 7.5 2014 Mauritius 39.7 2008 Cabo Verde 20.7 2010 Swaziland 15.2 2010 Botswana 12.5 2009 Mauritania 9.4 2005 Congo, Rep. 6.9 2012 Côte d’Ivoire 6.3 2010 Namibia 5.6 2008 Cameroon 5.2 2011 South Africa 3.5 2010 Lesotho 3.1 2005 Sub-Saharan Africa 10.5 2010 Developing economies 22.0 Source: GoG 2014b, Table B.6; SSNIT for Ghana. There is scope to reduce fragmentation in scholarship programs. The Scholarships Secretariat (SS) manages most scholarships for senior high school and tertiary education. At the same time, several donors finance scholarships for JHS and SHS students through the Girl Education Unit in the Ghana Education Service (GES) and nongovernment organizations (NGOs). Apparently there is no coordination between SS and GES programs. The Education Strategic Plan 2010–2020 does not mention SS. Similarly, the appraisal documents for DFID’s Girls-PASS and World Bank’s Secondary Education Improvement Project (SEIP), which finance scholarships, do not mention 161 http://www.spyghana.com/ghanaian-employees-ignores-2nd-tier-pension-scheme/. 162 http://www.ghananewsagency.org/economics/ssnit-informal-sector-mobilises-more-than-91-000-members- 38132. 59 SS. On the other hand, the Auditor General reviews found overlapping of SS’ senior high school scholarships and lack of oversight and monitoring of the program.163 In this context, it should also be noted there is no national human resource development plan to guide SS awards for tertiary education. Similarly, youth programs are characterized by significant overlap and duplication. Ghana has many different programs focused on apprenticeship and entrepreneurship. These programs have been developed independently in various Ministries, Departments, and Agencies (MDA) without a coherent policy framework.164 Many of the programs target the same population groups, which creates an “uncoordinated policy and institutional framework for youth development,” as the Ghana Shared Growth and Development Agenda (GSGDA) II puts it.165 Linkages with social protection programs are limited or nonexistent. The impact of most of these programs is not yet known, and none of these programs has been evaluated through rigorous impact evaluations. Accountability for results is limited.166 But there are also gaps in coverage. Many extreme poor families and children are still not covered by LEAP. Second chance programs, career counseling, and labor market information are lacking. As a recent report indicated, “there is a lack of a transition management and no systematic approach on extra preparation courses for drop outs and low skilled workers for their re- integration. There is a lack of a second chance to achieve the minimum standard of qualifications/competences.”167 In sum, Ghana has a comprehensive social protection system that covers most of the risks that have been identified, but coverage of the most vulnerable groups, such families and children in extreme poverty, by LEAP is limited, as is the coverage LIPW, health insurance, and social security. In contrast, there appears to be substantial program overlap and fragmentation of scholarship and youth programs. B. The Generosity of Benefits Are existing programs sufficiently generous to have the desired impact? To answer this question, the benefits paid by the major social assistance programs and social insurance are compared to international benchmarks. The focus is on the major programs for which there is comparative information: LEAP, LIPW, and SSNIT. Global estimates of the generosity of two types of social assistance programs, as reported in a World Bank study, are presented in Table 4.4. Program generosity in this study is defined as 163 GAG 2006. 164 A large number of institutions are involved in active labor market programs (ALMP) and related areas, including: the MoGCSP (vocational training programs, rehabilitation centers for skill upgrade of the disabled); MLGRD (LESDEP, Community Development Vocational/Technical Institutes); MELR (National Vocational Training Institute, Opportunities Industrialization Centers-Ghana, Integrated Community Centers for Employable Skills); Ministry of Youth and Sports (NYEP, Youth Leadership and Skills Training Institutes of the National Youth Authority); Ministry of Trade and Industry (Ghana Regional Appropriate Technology Industrial Service); Council for Technical and Vocational Education and Training (National Apprenticeship Program); and Ministry of Education/GES (36 Technical Training Institutes located in the urban areas of the 10 regions, some of which were taken over from the private sector). 60 165 NDPC 2014, 126. 166 Schwegler-Rohmeis and others 2014, 62. 167 Schwegler-Rohmeis and others 2014, 53. the program transfer divided by the average consumption expenditures (excluding durables, housing, and health care) of the poorest households (those in the first quintile, Q1). 168 LEAP is also considered a last resort program, as it targets the poorest families with children, among other categories (poor elderly and people with disabilities). Because it also targets elderly people in extreme poverty without any source of income, it could also be construed as a noncontributory social pension for this group. Table 4.4: Generosity of Programs in Developing Countries Relative to Average Consumption Programsa Median Range Denominator transfer as % of average transfer as % of average consumption consumptionb Last resort programs (20)d 23% 5%–45% Q1c Family allowance (15)e 18% 6%–28% Q1c Source: Grosh and others 2008, 136 and 137. Q1 = Quintile 1 (poorest) a. The number of countries with these programs is shown in parenthesis. b. The two highest and lowest values were excluded. c. Consumption expenditures exclude durables, housing, and health of the poorest quintile. d. Last resort programs are needs-based, usually means-tested, programs designed to help those who are not assisted, or not assisted enough to prevent poverty, by social insurance (pensions, unemployment insurance) or universal programs (child allowances, education, and the like). e. Family allowances are cash transfer for families with children. They can take various forms, such as means-tested child benefits, birth grants, or universal transfers for all children under a fixed age. Despite tripling in 2012, LEAP benefits are still fairly modest. The household benefit under LEAP ranges from GH¢24 to GH¢45 per month, depending on the number of eligible household members; the average is GH¢36. This corresponds to 19.5 percent of the monthly food expenditures of a Q1 family (GH¢ 184). This is only enough to buy food for 5.8 days, and compares with 23 percent for last resort program in Table 4.4. Microsimulations show that the impact of LEAP on extreme poverty could be greater. To evaluate the impact of LEAP on extreme poverty, we performed a microsimulation using Round 6 of Ghana Living Standards Survey (GLSS-6). For current parameters, we used 2014 household coverage and an average annual benefit of GH¢432 (or GH¢69.7 per capita). To perform the simulation, we added the average benefit to the expenditure level of 560,520 extreme poorest people (90,000 households) in the sample and evaluated the resulting poverty incidence and gap. As can be seen in Table 4.5, there is 8 percent reduction in the extreme poverty gap. The incidence of extreme poverty does not change. This simulation illustrates the potential of LEAP as a poverty reduction program, but with the current coverage and benefit the overall impact is relatively small. For the beneficiary’s families, the impact of the program is not inconsequential , as the program impact evaluations has shown (see Chapter III). Table 4.5: Simulated Impact of LEAP on Poverty Indicators Extreme poverty Poverty P0 P1 P0 P1 Baseline (2013) 8.373 2.274 24.148 7.726 168 We take food consumption to represent the adjusted consumption excluding durables, health, and housing. 61 LEAP transfers perfectly targeted 8.373 2.09 24.148 7.615 Percentage points change 0 0.184 0 0.111 % change from baseline 0.00% -8.09% 0.00% -1.44% Source: World Bank staff estimates based on GLSS-6. Note: P0= Poverty headcount rate; P1=poverty gap; The transfer to the elderly poor is also relatively low. An elderly LEAP beneficiary living on his or her own would receive GH¢24, equivalent to 36 percent of the extreme poverty line. If this transfer is equated to a noncontributory pension for the elderly, Ghana would also compare unfavorably with other lower middle-income countries in Africa. Table 4.6 presents data from HelpAge International’s Pension Watch for those middle-income countries with contributory pensions, in terms of pension benefitsas a percent of GDP per capita. The Ghana LEAP transfer would be equivalent to 8 percent of Ghana’s GDP per capita (2013), which is among the lowest in the middle-income Sub-Saharan African countries. Table 4.6: Noncontributory Pensions in Sub-Saharan Africa Middle-Income Countries Country Benefit per GDP per capita Percent of GDP Targeting month (US$) (US$), 2013 per capita Ghana 12 1858 8 Means-tested South Africa 125 6618 23 Means-tested Cabo Verde 63 3767 20 Means-tested Mauritius 118 9203 15 Universal Namibia 60 5693 13 Universal Nigeria 32 3006 13 Means-tested Swaziland 20 3034 8 Pensions-tested Zambia 12 1845 8 Universal Source: http://www.pension-watch.net/ and World Development Indicators. On the face of it, the compensation for LIPW workers appears to be on the high side. The compensation being paid in 2015 by LIPW for a six-hour working day is GH¢7 (GH¢1.16 per hour), equal to the minimum wage for an eight-hour working day (GH¢0.875 per hour). Thus on an hourly basis, LIPW compensation is 32 percent higher than the minimum wage. As can be observed in Table 4.7, setting the wage or compensation rate for public work programs below the minimum wage or the ongoing market wage for the unskilled workers is the norm in most labor- intensive public works programs in developing countries as a way to provide for self-selection and ensure that only those who really need the jobs take them. Table 4.7: Wage Setting in Labor-Intensive Public Works Programs Country Intensive public works programs Wage setting Bangladesh Food-for-Work Program Program wage < market wage Bolivia Emergency Employment Program Program wage < market wage (about 2/3) Ethiopia Productive Safety Net Program Program wage < market wage (cash or food) Malawi Public Works Program wage < minimum wage South Africa Expanded Public Works Program wage ≤ minimum wage Yemen Public Works Program wage < minimum wage Source: Grosh and others 2008, 487–92. In the case of Ghana, however, the minimum wage is reported to be well below the market wage for unskilled workers. International Labor Organization (ILO) and LIPW staff observed 62 that the wage paid by LIPW is below the ongoing wage for unskilled and agricultural workers in the project areas. The current minimum wage of GH¢7 is equivalent to 80 percent of the poor household average monthly food expenditures.169 SSNIT’s current pensions are fairly generous. In 2014, the SSNIT minimum pension was GH¢200, which is equivalent 1.8 times the poverty line (Table 4.8). About half the pensioners received this minimum pension. The large increase from 2013 to 2014 in the number of pensioners receiving the minimum pension is because the amount of the pension increased by 100 percent, which affected a large number of pensioners. For 2015, SSNIT announced that the minimum pension for “fresh” pensioners will be GH¢230, equivalent to 2.1 time the poverty line; for those receiving the minimum pension in 2014, their pension will be GH¢246.92.170 Table 4.8: SSNIT Benefits 2010 2011 2012 2013 2014 Total benefits paid (GH¢million) 311 355 443 692 944 Old-age pensions: Highest pension (GH¢ /month) 4,785 6,932 8,648 11,227 22,292 Average pension (GH¢ /month) 297 312 393 631 768 Lowest pension (GH¢ /month) 42 45 50 100 200 Average pension/average salary (%) 70 61 63 87 90 No. pensioners with minimum pension 2,479 850 706 7,661 67,621 Memo: Average salary (GH¢ /month) 425 508 621 726 849 Number of pensioners 107,312 112,522 119,323 129,691 142,076 Source: SSNIT. Pension for future retirees would be very generous. According to a World Bank study, the 2008 reform of Ghana’s public pension system will gradually result in substantial additional levels of income replacement for retirees, as they will receive unreduced Tier I benefits (despite the reduction in the contribution for Tier I), in addition to Tier II benefits.171 Indeed, the SSNIT Tier I scheme offers a replacement rate for a full-term worker after 30 years of 72.5 percent of the best three years of wages; while under the combined Tier I and Tier II schemes, the future replacement rate is projected at about 87.5 percent. The current replacement rate is already among the highest in Africa (Figure 4.1). The SSNIT benefit formula provides much higher benefits for the first 15 years of contributions (an accrual rate of 2.5 percent/year of contributions), while the accrual rate for additional years is less than half that (1.125 percent/year). It also provided for a minimum vested replacement rate of 50 percent after 15 years of contribution. This creates a disincentive for workers to continue to contribute after 15 years.172 169 Considering 22 days of work in a month. 170 http://ssnit.org.gh. 171 World Bank 2013a. 172 World Bank 2013a, vi, footnote. 63 Figure 4.1: Gross Replacement Rates in Selected Sub-Saharan African Countries 90 80 70 60 50 % 40 30 20 10 0 Source: World Bank 2013a, based on the World Bank pension database. Recently, the Parliament approved amendments to the pension legislation that introduce important reforms that affect the future replacement rates.173 These reforms include: (i) reducing the minimum vested replacement rate from 50 percent to 37.5 percent after 15 years of contributions; (ii) reducing the accrual rate for additional service from 1.5 percent to 1.125 percent per year; and (iii) reducing the maximum replacement rate from the SSNIT Tier I scheme to 60 percent. The reforms will reduce the projected replacement rates to some extent, although they will continue to be high by international standards. Summing up, LEAP pays relatively low benefits. The compensation paid by the LIPW program is higher than the minimum wage on an hourly basis, but is reportedly below the competitive wage for unskilled labor in the project areas. On the other hand, current social security pensions are fairly generous, while future replacement rates are too generous and most likely unsustainable. C. Targeting Accuracy An analysis of benefit incidence helps evaluate the targeting accuracy of the social assistance programs, or the extent to which the poor benefit from the programs. Benefit incidence can be estimated using the Coady-Grosh-Hoddinott (CGH) indicators or from concentration curves and concentration and progressivity indices, all based on the GLSS-6. LEAP, as Ghana’s flagship cash transfer program, is of particular interest for this analysis. LEAP and NHIS exemptions are the only social assistance program explicitly included in the GLSS-6. Unfortunately, only a small number of households indicated that they were beneficiaries of LEAP, given its limited coverage. In addition, there are problems with the way the questions concerning LEAP were formulated in the questionnaire and recorded. The unit of assistance for the LEAP program is the household, but in the GLSS-6 questionnaire the question concerning LEAP has “any member of household” as the unit; the amount of the LEAP transfer to a particular household was recorded as being received by each member of the household. Consequently, the amount of LEAP transfers per beneficiary household would be overestimated by a factor equal to 173 The National Pension Amendment Act, 2014 (Act 883), passed by the Parliament of Ghana and gazetted on December 31, 2014. 64 the household size. Also, there were apparently problems with quality control of the answers and/or data entry. LEAP makes bimonthly payments, or 6 payments during the year. However, in the GLSS-6, there are households that report 8 and 12 payments during the last 12 months. Despite these problems, the inclusion of LEAP in the GLSS-6 was an important step forward. It is highly desirable for other major social assistance programs such LIPW and school lunches to be included in future surveys. The questions in the questionnaires pertaining to these programs should be carefully revised with program managers. The Coady-Grosh-Hoddinott (CGH) indicator compares the portion of the benefits received by a population quintile (say, the poorest quintile) divided by the portion of population in that quintile (20 percent).174 A program with a CGH higher than 1 means that the poorest quintile gets more from the program than its share in the population. The greater than 1 that CGH is, the more highly targeted on the poor the program is; conversely, the lower than 1 CGH is, the less pro- poor is the program. The interpretation of the indicator may be different for programs whose primary objective is not targeting the poor; in that case, the indicator will show the extent to which the poor benefit from these programs. An example is primary education, which targets all children and not only poor children; in this case we may interested in knowing to what extent poor children benefit from government subsidies to primary education.175 The proportion of poor people benefiting from scholarships, apprenticeship programs is low, but the ratio is high for primary school and junior high school benefits, and in the middle for the NHIS exemption. In relation to scholarships, the results do not reflect the scholarships financed by DFID’s Girls Participatory Approaches for Student Success (Girls-PASS) and SEIP programs because Girls-PASS was about to initiate operations and SEIP was still not effective when the field work of GLSS-6 was done. The regressivity of energy subsidies in 2012 when the GLSS6 data were collected is shown by the lowest CGH indicators. Table 4.9: CGH indicator in the Poorest Quintile Programs Beneficiary or benefit incidence in poorest quintile Primary (number) 1.5 Junior high school (number) 1.2 NHIS-Under 18 (number) 1.1 Kerosene (GH¢) 1.0 NHIS-Elderly (number) 1.0 NHIS (number) 0.9 NHIS-Indigents (number) 0.9 NHIS-Maternal service (number) 0.9 Senior high school (number) 0.8 Apprenticeship (number) 0.6 Petrol (GH¢) 0.3 Electricity (GH¢) 0.3 Social security (number) 0.2 174 Coady, Grosh, and Hoddinott 2004. The poorest quintile of population (bottom 20 percent) is used to define poverty, as it corresponds roughly to the incidence of poverty in Ghana (24 percent). 175 For some programs, only the information on the participation of different quintiles is available in the GPSS-6, not the information on the amount of benefits the beneficiaries received. This is the case for school attendance, for example. In some of these cases, it is assumed that benefits are equally distributed among the participants and therefore beneficiary incidence and benefit incidence are similar. 65 Scholarships/bursaries (GH¢) 0.2 University (number) 0.1 Diesel (GH¢) 0.1 LPG (GH¢) 0.0 Source: World Bank staff estimates on the basis of GLSS-6 and the programs/transfers that are captured in the household survey (the school feeding, text and uniform, and LIPW programs are not captured in the GLSS-6). There are not enough observations to capture the targeting accuracy of LEAP benefits. Note: Values in the right column refer to either the beneficiary incidence (number) or the benefit incidence (GH¢). Targeting accuracy can also be determined from the concentration curve, which makes use of the information from the full distribution of benefits (or beneficiaries). The concentration curve shows the share of total benefits (vertical axis) going to the percentage of the population ranked from the poorest to the richest income per capita quintiles (horizontal axis). A program in which an equal amount of benefits is received by each quintile has a straight 45-degree concentration line. A pro-poor program is represented by a concentration curve located above the 45-degree line. A program is progressive if its concentration curve is above the consumption concentration curve. If the program’s concentration curve lies between the diagonal and the consumption curve, the program is not pro-poor, although it is progressive. If it lies below the consumption curve, it is not pro-poor and it is regressive, in the sense that its distribution is worse than the existing distribution of consumption. According to GLSS-6, both apprenticeship and scholarships programs are not pro-poor. Figure 4.2 shows that the concentration curves for these programs sit below the 45-degree line. Indeed, 25 percent of those benefiting from apprenticeship programs are in the richest quintile, compared to 11 percent of those in the poorest quintile. As for scholarships, almost 60 percent of the benefits go the richest quintile, compared to 8 percent for the poorest. The potential inequity of scholarships awarded by the Scholarships Secretariat has been already flagged in the Ministry of Education’s Strategic Plan 2010 to 2020. Figure 4.2: Concentration Curves and Benefit Incidence for Apprenticeship and Scholarship Programs 1 70% cumulative % of benefits 60% 0.8 50% 0.6 % benefits 40% 0.4 30% 0.2 20% 0 0 0.2 0.4 0.6 0.8 1 10% cumulative % of population, ranked from the poorest to the richest 0% Apprenticehsip (#) Scholarships (GHc) Apprenticehsip (#) Q1 Q2 Q3 Q4 Q5 Scholarhsips/burs. (GHc) Source: Staff estimates on the basis of GLSS-6. 66 Note: Distribution based on benefits (GH¢) or beneficiaries (number). Energy subsidies are also highly regressive, except for kerosene. As can be observed in Figure 4.3, the concentration curve for kerosene is close to the 45-degree line, indicating that it consumed by the poor. The consumption of other fuels is well below the 45-degree line, and therefore clearly not pro-poor. In case of electricity, this may reflect the lack of access by many poor households. This analysis corroborates the findings of recent studies that noted the inequity built into most energy subsidies.176 Figure 4.3: Concentration Curves and Benefit Incidence for Sources of Energy 1.00 0.9 0.8 cumulative % of benefits 0.80 0.7 % of benefits 0.6 0.60 0.5 0.4 0.40 0.3 0.2 0.20 0.1 0.0 0.00 Petrol Diesel Kerosene Electricity 0 0.2 0.4 0.6 0.8 1 (GHc) (GHc) (GHc) (GHc) cumulative % of population, ranked from the poorest to the richest Q1 Q2 Q3 Q4 Q5 Kerosene Electricity Source: World Bank staff estimates on the basis of GLSS-6. When the concentration curve of a program crosses the 45-degree line or is too close to it, as in the case of kerosene, its concentration index helps determine if the program is pro- poor or not. This index is calculated from the concentration curve in a similar manner as the Gini coefficient is calculated from the distribution of per capita consumption. Negative values of the concentration index indicate a pro-poor program. The larger the index in absolute terms, the more accurately the program is reaching the poor. Figure 4.4 shows the concentration index for those transfer and subsidy programs for which there is information. As can be observed, the concentration indices are negative and pro-poor for several programs (primary education and junior high school education) and positive and not pro-poor for most other programs. 176 See, for example, Molini, Clementi, and Schettino (2014). 67 Figure 4.4: Concentration Indexes for Selected Social Protection Programs -0.400 -0.200 0.000 0.200 0.400 0.600 0.800 Primary (#) JHS (#) NHIS-Under 18 (#) NHIS-Maternal service (#) Pro-poor NHIS-Aged (#) NHIS (#) Kerosene (GHc) NHIS-Indigents (#) Apprenticeship (#) SHS (#) Social security (#) Electricity (GHc) Not pro-poor Petrol (GHc) Scholarships/burs. (GHc) LPG (GHc) University (#) Diesel (GHc) Source: World Bank staff estimates on the basis of GLSS-6. Note: Distribution based on benefits (GH¢) or beneficiaries (number). When the concentration curves cross the consumption line or is too close to it, as in the case of scholarships or electricity, a progressivity index can be calculated to evaluate whether the program helps to increase equity (that is, whether the program is progressive or regressive). The technique most commonly used to measure the progressivity of public expenditures is the method proposed by Kakwani, which is equal to the program’s concentration index minus the Gini coefficient of consumption. A program that increases equity (that is progressive) has negative values in its progressivity index. Table 4.10 presents the progressivity index for each program from the most accurate in reaching the poor (primary education) to the least accurate in reaching the poor (diesel). Most programs are progressive, implying that they help reduce the considerable inequity that still exists in Ghana as measured by the Gini coefficient; however, several programs are not pro-poor. Apprentice programs and the NHIS premium subsidy for indigent people are not pro-poor but they are progressive, as per 2012 when the data were collected. It has to be noted though that NHIA has since 2013 engaged to improve the targeting of the poor of NHIS, especially through the establishment of the National Registry, in close collaboration with LEAP and MoGCSP. On the contrary, scholarships and subsidies on petrol, diesel and LPG are not only not pro-poor but are also regressive. Table 4.10: Concentration and Progressivity Indexes of Selected Programs Programs Concentration Progressivity Index Characterization Primary (number) -0.24 -0.64 Pro-poor and progressive JHS (number) -0.12 -0.53 Pro-poor and progressive NHIS-Under 18 (number) -0.07 -0.47 Pro-poor and progressive NHIS-Maternal service Pro-poor and progressive (number) 0.00 -0.41 NHIS-Aged (number) 0.02 -0.39 Not pro-poor but progressive 68 NHIS (number) 0.04 -0.37 Not pro-poor but progressive Kerosene (GH¢) 0.04 -0.37 Not pro-poor but progressive NHIS-Indigents (number) 0.10 -0.31 Not pro-poor but progressive Apprenticeship (number) 0.12 -0.29 Not pro-poor but progressive SHS (number) 0.12 -0.29 Not pro-poor but progressive Social security (number) 0.28 -0.13 Not pro-poor but progressive Electricity (GH¢) 0.38 -0.03 Not pro-poor but progressive Scholarships/burs. (GH¢) 0.55 0.05 Not pro-poor and regressive Petrol (GH¢) 0.52 0.11 Not pro-poor and regressive LPG (GH¢) 0.57 0.16 Not pro-poor and regressive University (number) 0.59 0.19 Not pro-poor and regressive Diesel (GH¢) 0.67 0.26 Not pro-poor and regressive Source: World Bank staff estimates on the basis of GLSS-6. Note: Distribution based on benefits (GH¢) or beneficiaries (number). The GoG has been improving its targeting mechanism for social protection programs. With World Bank support, the GoG is establishing a National Targeting System (NTS). It consists of an updated targeting tool (poverty map and proxy means test, PMT) and a unified beneficiary identification registry, the Ghana National Household Registry (GNHR) to immediately serve the improve targeting of LEAP and NHIS. The poverty map and PMT are being revised and updated using the GLSS-6. GNHR is a single registry of the poorest, most vulnerable, and excluded households and also contains basic information on their socioeconomic status. The registry is to be accessed by all social protection programs in the country. The GNHR will therefore be the basis for a new National Targeting System (NTS), which will help social protection programs to identify, prioritize, and select households living in vulnerable conditions to ensure that different social protection interventions effectively reach their target populations. The GNHR will be based in MoGCSP. The GNHR is expected to eliminate the duplication of efforts and cost since MDAs involved in the targeting of the vulnerable groups will not have to build their own database. As mentioned, MoGCSP is also introducing e-payments to expedite and better control transfer payments. In sum, there is ample room to improve the accuracy of targeting of the key social assistance programs. The benefit incidence analysis indicates that apprenticeship programs are not pro-poor but progressive, while scholarships and subsidies on petrol, diesel, and scholarships are not only not pro-poor but also regressive. Ongoing efforts to improve the targeting mechanisms and create a single household registry will help make the social protection systems more equitable and effective. D. Cost-efficiency The share of operating costs to the total cost gives an indication of whether programs are cost-efficient or not.177 Operating costs in the context of this analysis for social assistance programs are estimated as the difference between total program costs and the value of the transfers or benefits delivered to program beneficiaries (or the cost of delivering the benefits to the beneficiaries). Table 4.11 presents the operating costs of different types of social assistance 177 Cost-efficiency compares cost of inputs needed to produce program outputs; for example, the cost of delivering the transfers. Cost-effectiveness compares cost of outputs needed to produce program results or outcomes; for example, the reduction in the poverty gap per US$1 in transfers. See Grosh and others 2008, 204; GoG 2014b. 69 programs from a variety of countries, as reported in a World Bank study.178 The range of the share of operating costs is quite wide and may be attributable to different factors, which makes comparisons difficult. Moreover, a program can have relatively high operating costs because of its small scale and its low benefits or because it is new and has high start-up costs. On the other hand, a program may have low operating costs because it does not invest in key activities and processes such as staff training, communications, auditing and controls, information management, and monitoring and evaluation systems. As a result, the World Bank study suggests developing an index of operating expenditures to take some of these factors into consideration. The proposed index is calculated by multiplying the generosity of program by the percentage of operating costs in the total program cost.179 Table 4.11: Typical Operating Costs as Share of Total Program Cost (Percent) Type of Program Minimum Maximum Average Median Number of programs Food-related programs a 6.0 55.5 25.4 22.0 19 Cash and near-cash b 2.0 16.6 8.2 8.9 16 Conditional cash transfer 4.0 13.0 8.2 6.7 9 Public works 1.6 24.0 10.3 4.9 6 Source: Grosh and others 2008, 411. a. Includes school lunches and nutrition programs. b. Includes old age pensions, child allowance, food stamps, guaranteed minimum income. LEAP cost-efficiency appears to be on the low side. Current estimates of LEAP recurrent costs were not available.180 At start-up (in 2010), they were estimated at 53 percent of total costs and projected to decline to 23 percent in 2012 as it gradually expanded, and more recently declined to 10.3 percent.181 The program project implementation manual sets a target of 20 percent or less for 2014 to 2016.182 Taking this latter figure, we can follow the procedure suggested by the World Bank study183 and calculate an index of operating costs for the LEAP. Recall that the program’s generosity was estimated at 19.5 percent. Assuming that operating costs as a percentage of total costs are 20 percent, it would yields an index of operating costs of 3.9 (=19.5*0.20). This compares with indices ranging from 0.5 to 2.25, with a median value of 1.1 for 10 social assistance programs in various parts of the world for which information was reported in the World Bank study.184 By these standards, the operating costs of LEAP would appear to be high. Nonetheless, as LEAP expands coverage and adjusts benefit levels, its overhead should decrease to levels similar to other more mature transfer programs. 178 Grosh and others 2008. 179 Grosh and others 2008, Box 9.3, 392. The formula for the proposed index is I = (B/C*Q1)*CA/TC; where B is the transfer; C*Q1 is the adjusted consumption for Q1 families; and CA and CT are the program’s operating costs and total costs, respectively. Thus a program with high generosity is expected to have relatively lower operating costs. 180 The information on LEAP expenditures received was fragmented. From MoGCSP, we received information on DFID financing for benefits and operating costs; from GSOP unit, we received information on World Bank financing. 181 White, Hodges, and Greenslade 2013, 34; and LEAP Management Unit. 70 182 MLGRD and MoGCSP 2014, 60. 183 Grosh and others 2008. 184 Grosh and others 2008, Box 9.3, 392. LIPW administrative costs appear still to be on the high side, mostly due to the capacity- building component of the program. The information provided by LIPW was timely and complete. LIPW’s share of total costs represented by administrative costs averaged 21 percent in the last two years, compared to a benchmark for this type of program of 10 percent (Table 4.12). An important part of the administrative costs have been spent on capacity building. Currently about 46 percent of the program benefits accrue to the workers in additional income; the target is to increase this share to 62 percent in the next three years.185 Table 4.12: LIPW Administrative Costs 2013 2014 Total cost (GH¢) 37,536,649 38,529,662 Benefits 29,505,531 30,418,981 Benefits to the poor 16,934,001 17,614,290 Infrastructure investment 12,571,530 12,804,691 Administrative costs 8,031,119 8,110,681 Administrative/total costs (%) 21.4 21.1 Benefits to the poor/total cost (%) 45.1 45.7 Memo: Workers (number) 52,254 30,042 Source: GSOP management. The Ghana School Feeding program has the lowest operating cost. The cost of the Ghana School Feeding Program (GSFP) reached GH¢165 million in 2014 and covered 1.7 million pupils. The reported administrative cost of the program is only about 2 percent of the total cost (Table 4.13). While GSFP does not transfer food to schools, with the purchasing and cooking foodstuffs being done at the local level, this estimate of administrative expenses appears to be low compared to the benchmark of 25 percent. Nonetheless, information received from program management in the Ministry of Local Government and Rural Development (MLGRD) indicates the annual cost of the program per pupil in 2014 was GH¢95 (equivalent to US$33), and was projected to increase to GH¢179 (about US$62) in 2015. A World Bank study reports that the cost of school feeding programs usually ranges from US$28 to US$63 per child per year, with an average of US$20 per child per year for the 19 countries in the sample.186 Therefore, the unit cost of the School Feeding Program in Ghana appears to be within international standards, but projected to increase. Table 4.13: Cost of the Ghana School Feeding Program 2013 2014 Total spending (GH¢ 000)) 96,363 164,687 Benefits 94,124 161,846 Administration cost 2,239 2,841 Administration costs/Total cost (%) 2.3 1.7 Memo: Cost per pupil 56 95 Number of pupils 1,728,681 1,740,000 Source: School Feeding Program management. 185 MLGRD and MoGCSP 2014, 59. 186 Bundy and others 2009, 60. 71 NHIS’ administrative costs are increasing rapidly. A study of Ghana’s health sector estimated the administrative costs of NHIS in 2008/09 at about 8 percent of the total cost, which it considered reasonable.187 A more recent estimate places administrative costs at 14.5 percent of total cost in 2013, more than doubling an average of about 6 percent in the previous three years (Table 4.14).188 This is a reflection of doubling of operating costs and five-fold increase in compensation expenditures. Table 4.14: Distribution of NHIS Expenses (Percent) 2010 2011 2012 2013 Operating expenses +compensation 5.8 4.8 7.6 14.5 Operating expenses 4.6 3.5 5.9 7.7 Compensation 1.2 1.3 1.7 6.8 Claims 73.5 71.9 75.1 75.0 Otherexpensesa 20.7 23.3 17.3 10.5 Overall expenditures 100.0 100.0 100.0 100.0 Source: NHIA Audited Financial Statements. a. Mostly includes support to the Ministry of Health. SSNIT administrative expenses are reasonable, though they have had a tendency to increase. SSNIT administrative costs averaged 8 percent of total contributions from 2010 to 2012 (Table 4.16). This compares well, for instance, with Botswana, where the administrative cost of the public pension plan (Botswana Public Officers Pension Plan) and private plans are 16 percent and 14 percent of contributions, respectively.189 By way of comparison, in Sierra Leone, the administrative cost of the National Social Security Investment Trust (NASSIT) relative to contributions is about 30 percent.190 In São Tomé and Principe, the cost of administration of Social Security is about 12 percent of the contributions, though by law they should not be higher than 7 percent.191 In 2013 and 2014, SSNIT administrative expenses increased to 11 percent and 15 percent of contributions, respectively, mostly because of a drop in contributions as a consequence of the build-up in payment arrears by the public sector. If arrears are taken into consideration, administrative expenses would average 7.7 percent of adjusted contributions (see bottom of Table 4.15), though there was clearly a tendency for these cost to increase in 2013 and 2014. Table 4.15: SSNIT Expenses 2010 2011 2012 2013 2014 Contributions (GH¢ million)a 512 749 1,192 1,048 887 Public 140 155 632 335 35 Private 372 594 559 714 852 (Payment arrears, net increase) -- -13 104 435 484 Expenses (GH¢ million) 450 409 532 770 1,077 Benefits 311 355 443 692 944 Administrative expenses 51 50 86 112 132 Others 88 50 85 69 111 187 Saleh 2013, 119. 188 Saleh 2013, 119, estimated administrative costs at 8 percent for 2008/09, which he considered reasonable. 189 Marques 2013b. 190 Marques 2013a. 191 Marques 2013c. 72 Administrative expenses/Contributions (%) 10 7 7 11 15 Administrative expenses/contributions + arrears (%) -- 6.8 6.6 7.6 9.6 Source: SSNIT. Note: --=not available. a. Excludes payment arrears. In sum, the cost-efficiency of the LEAP and LIPW programs is still on the low side, but as the programs mature, they should be brought up to international standards. The School Feeding Program appears to be cost-efficient, but costs were projected to increase in 2015. The administrative costs of NHIS were reasonable during 2008–12, but increased quickly in 2013, the last year for which data are available. This increase was due to the passage of NHIS Act 2012 (Act 852), which consolidated all district mutual health insurance schemes into a unitary scheme. SSNIT appears to be cost-efficient when compared to several social security systems in Sub- Saharan Africa for which information was available, though administrative costs increased in 2013 and 2014. E. Sustainability Delays in release of funds to programs have been severe for most social protection programs in recent years. The exceptions is LIPW in 2015, as it is financed by external resources. Some of the payment arrears are to small microenterprises (Table 4.16).192 Other payment arrears are to the Statutory Funds, which in turn incur payment delays to the programs they are expect to finance. The arrears situation is mainly the result of weak public financial management, as discussed in Chapter V. At the same time, there is heavy reliance on external financing for major programs, which may make some of these programs unsustainable when foreign assistance dries up. Table 4.16: Payment Arrears Programs/Funds Description Amount, 2014 (GH¢ million) Exercise books Books bought in 2012 paid in 2014 -- School feeding Arrears to caterers -125 days overdue as of Feb 2015a -- Scholarships Secretariat Arrears to beneficiary (cumulative); excludes foreign scholarships 96 DACF Difference between transfer due and received in 2013 471 GetFund Difference between transfer due and received in 2014 8 NHIS Payment of claims 5 month in arrears nationwide 299 SSNIT Difference between transfer due and received in 2013 826 DDF Arrears to Shai-Osu Doku district since 2012 -- NAP Arrears to master-craftsmen 0.8 Source: Program managers and Ghana Auditor General. Note: -- = not available. a. These arrears were cleared as of mid-2015. The NHIS faces important challenges to ensure its longer-term sustainability. The NHIS has helped improve access to health care for insured members where services are available, but the limited coverage of the population raises concerns as to the equity and sustainability of the scheme. The scheme's sustainability is threatened because the program is basically voluntary and this may 192 This had been the case, for example, for the caterers that provide school lunches and must incur debt in order to continue operations. All arrears to the caterers of the School Feeding Program were cleared by June 2015. 73 lead to adverse selection: less healthy individuals opt to join the scheme and more healthy individual stay out, which increases average costs. On the other hand, access to service utilization is unlimited (there are no copayments or annual or lifetime limits), and the growth in service provision may outpace the growth in the tax-based financing sources. In a late 2014 communication to Parliament, the Minister of Health raised the alarm that the NHIS was facing severe financial difficulties. He indicated that NHIS is experiencing high and increasing growth in membership; meanwhile, the utilization and cost of providing health care for NHIS subscribers is growing faster than the annual financial resources allocated to the scheme, placing the scheme under severe financial pressure. The NHIS has experienced persistent annual deficits since 2009. Currently, the payment of claims is five months in arrears. As a result of the financial imbalance, the scheme was expected to be confronted with a funding gap in excess of GH¢299 million for 2014. If the financing regime is not reviewed for additional inflows, the funding gap is projected to increase from GH¢347 million in 2015 to reach a projected gap of GH¢887.35 million in 2018. NHIS has taken a series of measures to increase efficiency; nonetheless, these efforts by themselves will not place the NHIS on a sustainable path.193 Government contribution to finance the NHIS exemptions does not cover costs . About 66 percent of NHIS membership is exempted from paying premiums. The GoG is expected to pay NHIS for the cost of their insurance. The GoG’s payment per exempted person rose from GH¢12 per person per year in 2008 to GH¢18 in 2010. This is well short of the average expenditure per member of GH¢58 a year (Table 4.17).194 Therefore, efforts by NHIS to increase the number of exempted population translate into higher operating deficits. Table 4.17: NHIS–Expenditures for the Extreme Poor Exemptions 2010 2011 2012 Overall expenditures (GH¢ thousand) 6,062 24,605 16,548 Per registered extreme poor (GH¢) 51.7 71.9 50.7 GoG transfer per registered indigent 18 Memo: Number of registered indigents 117,295 342,127 326,182 Source: GoG 2014b, 39 and 71. Government contributions in arrears to SSNIT have grown exponentially in the last few years. Table 4.18 shows that the central government arrears have increased tenfold since 2011, to over GH¢1 billion. Total payment arrears to SSNIT, equivalent to 8 percent of its reserves in 2010, now represents 21 percent. Table 4.18: Arrears to SSNIT 2010 2011 2012 2013 2014 Contributions (GH¢ million) 512 749 1,192 1,048 887 Public 140 155 632 335 35 Private 372 594 559 714 852 193 Parliamentary statement by the Hon. Minister of Health on the status of the national health insurance scheme, http://nhis.gov.gh/downloads.aspx. 194 GoG 2014b, 71. Unfortunately, we did not receive the requested information to update these data. 74 Cumulative arrears (GH¢ 232 219 323 758 1,242 million) Public (Controller and Acc Gen) 165 114 202 583 1,083 Subvented (publicly supported) 11 23 37 74 73 Private 55 82 83 101 86 Arrears/Reserves (%) 8 6 8 13 21 Memo: Reserves (GH¢ million) 2,901 3,419 4,281 5,671 6,050 Source: SSNIT. Note: Acc Gen = Accountant General. SSNIT requires additional reforms to become sustainable over the longer term. The 2008 reform reduced contributions to SSNIT by 4 percentage points (earmarked to the privately managed Tier II scheme), but it did not reduce the benefit for SSNIT contributors. According to a recent World Bank evaluation of the Ghana pension system, this, together with increasing dependency rates, will adversely affect the sustainability of the SSNIT Tier I scheme.195 At the same time, the legislation will gradually result in additional levels of income replacement for retirees. The combined replacement rate for Tier I and Tier II pensioners could reach as much 87.5 percent for a full-term worker. The World Bank study used the PROST actuarial tool to estimated SSNIT financial sustainability and conclude that it could begin to incur deficits and be required to draw on reserves to pay benefits as soon as 2034.196 While these projections were made before the recent reforms mentioned in Chapter III, the World Bank study seems to indicate that they will be insufficient to bring longer-term financial sustainability to the system. In sum, the large fiscal disequilibrium has impacted the operations of all social protection programs in recent years—with the recent exceptions of LEAP and LIPW because these programs are financing to a large extent by donor resources . Some programs (including free uniforms, exercise books, some NHIS exemptions) have expanded without taking in due consideration their financial implications and longer-term sustainability. Both NHIS and SSNIT require important reforms to ensure their future financial sustainability. The GoG is fully aware of this and is already taken steps in this direction, but more needs to be done. F. Monitoring and Evaluation The NSPS envisaged the need to develop a monitoring and evaluation (M&E) framework for the social protection system at the national, regional, district, and community level. In the short run, the NSPS called for the development of a “M&E framework and action plan for every aspect of NSPS in order to keep constant and systematic track of sector-wide progress on social protection and the social protection floor.”197 It further called for capacity building for stakeholders for effective monitoring and evaluation; developing and implementing a comprehensive Registry System (Single Register Database) to facilitate the overall management of the implementation of the strategy; liaising with MDAs to collect, collate, analyze, and disseminate information on social protection to stakeholders; and conducting baseline studies of target communities. 198 195 World Bank 2013a. 196 World Bank 2013a, 10. 197 MESW 2012a, 104. 198 MESW 2012a, 130. 75 Information on social assistance programs is quite fragmented, dispersed, and incomplete. The difficulty in obtaining basic information from the institutions responsible for the programs reviewed here provides a clear indication of how inadequate existing information systems are. MoGCSP does not have basic information about social assistance programs. It was very difficult for most programs to produce basic information for this report, such as the cost of the program or the number of beneficiaries. Very few programs could provide their recurrent costs or sources of financing. Few social assistance and active labor market programs have been evaluated to find out whether they are producing the results and impacts expected. Of the 16 social protection programs discussed here, only two had some impact evaluation (LEAP and LIPW). Of the three ALMPs covered here, only one (NAP) is currently being evaluated. With no rigorous evaluations of Ghana’s safety net and ALMPs, the government cannot know what is and is not working and why. In addition to impact evaluations, regular process evaluations are equally important to ensure that programs are being implemented as intended. However, there are a few bright prospects. New rigorous impact evaluations are being planned for LEAP and capitation grants, while an evaluation of the National Apprenticeship Program is ongoing. LEAP and LIPW are developing their MIS and M&E systems. With support from UNICEF, the operationalizing of the LEAP’s M&E system in 2014 led to: (i) the roll-out of reporting tools at the district level and regular data collection on a bimonthly and quarterly basis; (ii) the uninterrupted production of National Quarterly Reports on payment statistics and operational activities at the district, regional, and national level; (iii) production of LEAP’s first annual report; and (iv) a Partnership Cooperation Agreement with a civil society organization to conduct the Independent Monitoring Checks, to provide the LEAP Management Unit with an indication of LEAP beneficiary satisfaction and adherence to LEAP conditionalities across the country over a two-year period. As mentioned, MoGCSP is also developing a single household registry that will gather information on the socioeconomic characteristics of the potential beneficiaries of social protection programs. The registry will also keep track of the beneficiaries of each program. A challenge will be to keep the registry updated to ensure its continued relevance. However, more needs to be done to develop a comprehensive M&E system for the social protection system. This includes clearly establishing the objectives to be achieved by the sector; developing and implementing a medium-term program and action plan and budget; selecting key indicators to track programs and overall progress toward the objectives; training personnel and developing capacities at all levels of government; and providing the inputs required to build and maintain the system. In this context, the social protection sector may take as an example what is already in place in the education and health sectors. In the health sector, the Holistic Assessment of the Health Sector Annual Work Program provides a thorough review of the sector development and challenges against its medium-term development program. The annual Education Sector Performance Report provides similar analysis for the education sector. In sum, progress has occurred in building M&E systems for LEAP and LIPW, as well as the household single registry (GNHR). More needs to be done to develop a comprehensive M&E system for social protection sector, including clearly establishing the objectives to be achieved; selecting key indicators to track progress toward meeting the objectives; and providing the inputs required to build and maintain the system. 76 G. Institutional Arrangements and Coordination Mechanisms 1. Institutional Arrangements NSPS provides an elaborate framework for institutional arrangements and coordination mechanisms. In January 2013, the GoG established the Ministry of Gender, Children and Social Protection (MoGCSP). MoGCSP replaced the Ministry of Women and Children and incorporated the Department of Social Development and the Social Protection Directorate of the then-Ministry of Employment & Social Welfare. This step underscores the priority the authorities have been giving to social protection. The new Ministry, in addition to having direct responsibilities for social services, also has the mandate to coordinate social protection interventions across MDAs as well as implementing NSPS and LEAP. In addition, MoGCSP was given the mandate to formulate and develop a National Social Protection Policy, as well as institutional arrangements at all levels, establish the National Targeting Unit, and ensure sustainable financial arrangements for all social protection interventions. The institutions involved in the sector are the MoGCSP; the Ministry of Finance; NDPC; MDAs that implement social protection programs; MMDAs and their social welfare offices; and the Ghana Statistical Services. MoGCSP has undertaken a restructuring exercise to be able to discharge its responsibilities. With support from Canadian International Development Agency (CIDA) under the Capacity Development Management (CDM) Project, MoGCSP engaged the Management Service Department (MSD) under the Office of Head of Civil Service to provide technical support for the restructuring. A comprehensive restructuring plan has been developed by MSD for three Departments of the Ministry: the Department of Gender, the Department of Children, and the Department of Social Development. The Ministry has also initiated a National Diagnostic assessment of the Department of Social Development to enable it to respond more effectively to the current needs of Ghanaian society.199 In its MTEF 2015–2017 submission, MoGCSP indicated that only half of the staff had received “capacity enhancement training and count on logistics and office equipment for performance and services delivery.” The MoGCSP is responsible for sector policies and therefore plays a leading role in the planning, implementation, and coordination of social protection policies and in the mobilization of resources for the sector. The Ministry's Directorate for Policy, Planning and Monitoring and Evaluation is responsible for assessing the progress in implementing the NSPS and the performance of the social protection system as a whole and the various social protection programs under the MoGCSP. MoGCSP is also responsible for implementing the country’s flagship LEAP and social services in coordination with MMDAs. The Ministry coordinates sector policies and chairs existing coordinating bodies. The MoGCSP currently has two secretariats, one council, six directorates, and three departments. Appendix A presents the current organogram for MoGCSP. The secretariats are the Domestic Violence Secretariat and the Human Trafficking Secretariat. The council is the Council for Disability. The directorates include the traditional civil service function directorates: Administration; Human Resources; Finance; Policy Planning, Monitoring and Evaluation 199 See MoGCSP (2014b) and http://gogcdm.gov.gh/reports/. 77 (PPME); Research, Statistical and Information Management; and Social Protection. The three departments are the Departments of Gender, Department of Children, and Department of Social Development (DSD).200 The Secretariats and the Council were created by an act of the Parliament and are responsible for providing policy guidance, strategic orientation, and coordination of their respective areas. The Directorates of PPME and Social Protection are responsible for formulating policy, while the Departments are responsible for implementing policy and programs. DSD has the primary responsibility for implementing LEAP. There appear to be overlap of functions between Secretariats and Directorates on the one hand, and of MoGCSP Departments and MMDA’s social welfare and community development departments, on the other. Table 4.19 provides details of MGCCSP’s staff distribution. Detailed information on MoGCSP’s Department of Social Development staff distribution between programs or between headquarters and the districts could not be obtained. Table 4.19: MoGCSP Staffing, 2015 Budget programs/Subprograms Total Staff Staff in in the regions/districts Management and Administration 61 General administration Finance Human resources Policy planning, budgeting and M&E Research, Statistics and Information Management Gender Equality and Women Development 25 Gender mainstreaming Women's rights and empowerment Child' s Rights, Promotion, Protection & Development 82 34 Social Development (Dept. of Social Welfare) 1,420 Social Services Securing Inclusion for Disability Social Protection (LEAP) Domestic Violence and Human Trafficking 12 Domestic violence Human trafficking Total 1,561 Source: MoGCSP 2014b. The Ministry of Finance is responsible for most of the sector financing. It provides budget resources to finance the implementation of social protection programs and it also coordinates the mobilization of resources from Statutory Funds and from external sources. MoF plays a key role in ensuring that approved programs are fully financed and the resources are transferred in a timely manner to intermediate and frontline service units. This does not always occur, with adverse effects for program implementation and service delivery. Ministries, departments, and agencies (MDA) are also responsible for implementing social protection programs. Among these are the MLGRD, which has the responsibility for implementing the LIPW and School Feeding Program; the Ministry of Education and Ghana 200 The Department to Social Welfare plans to change its name to the Department of Social Development, which will require legal action. 78 Education Service, which implements the capitation grants, and the uniforms and exercise books programs; the Scholarships Secretariat, which implements scholarships; the Ministry of Health and the NHIS, which implements health insurance waivers; and the Ministry of Employment and Labor Relations, which implements a series of youth programs. Other programs are also implemented by various MDAs, such as Workmen’s Compensation and LESDEP. Some of these programs are implemented through regional offices and MMDAs with the cooperation of civil society organizations or private actors. Within Ghana's decentralization policy framework, the MMDAs play a key role in the delivery of social services to the population. The decentralization policy has devolved a range of functions to the local authorities, including social services and community development. The District Assemblies (DA) are now responsible for matters concerning children, youth, women, the elderly, the disabled, access to health services, and poverty reduction. DAs plan and develop composite budgets based on local needs and national priorities; however, the budget allocations designated by the Ministry of Finance and MLGRD for social interventions through the District Assembly Common Fund (DACF) leave little discretion for local authorities to set their own priorities and develop their own programs (more on this in Chapter V). Ghana Statistical Services (GSS) has the key role of providing information for the monitoring and evaluation of social protection programs, and in enabling the planning, operation and monitoring of social protection policies. The data it produces in the population census and household surveys provide key information on socioeconomic status, demographic developments, and consumption patterns, which is needed to analyze overall development trends and social protection needs in the country. MMAs and MMDAs working on social protection rely on this information to plan and cost their interventions and to identify beneficiaries. Poverty maps and proxy means tests, which help target the programs on the poor, are based on the census and survey data. 2. Coordinating Mechanisms Before 2014, Ghana lacked a high-level body that could dictate policy and establish goals, monitor their achievement, and coordinate the development of social protection programs. As mentioned, many institutions are involved in the sector, and most programs continue to operate in silos, circumscribed by their own agendas, needs, and priorities, with little communication between them and little attempt to explore synergies. At the local level, the major institutional weakness derives from the fact that responsibilities have been devolved to the DAs, but resources have not followed to enable them to discharge the new functions effectively. The delimitation of responsibilities is also not always clear. Coordination with donors and other stakeholders has been done through the Social Protection Sector Group. The group, originally known as the Vulnerability and Excluded Sector Group, was relaunched in April 2013. It is open to all organizations involved in social protection activities, including ministries, departments, and agencies, development partners, civil society organizations, and research institutions. The Group is expected to address challenges in the implementation of the various social protection programs of MDAs and refer issues that require policy direction to the Social Protection Steering Committee (SPSC). In June 2014, the Cabinet approved a new multisectoral governance framework for social protection at the national, regional, district, and community levels. It confers on MoGCSP the 79 responsibility for: (i) developing strategies and strategic policies; (ii) ensuring sufficient financing for social protection in Ghana; (iii) assessing progress against national social protection targets; (iv)leading and providing support to the cross-sectoral committees for coordination; (v) coordinating reporting on social protection to relevant stakeholders; and (vi) approving proposals for new social protection initiatives. The following proposal for a institutional framework was approved.  A Social Protection Inter-Ministerial Committee. It will be chaired by the Minister of Gender, Children and Social Protection and will provide strategic oversight of social protection in Ghana by:  Approving annual plans for the social protection sector developed by the MoGCSP with input from the Technical Committee (see below).  Approving targets for the social protection sector and periodic review of progress against such.  Approving the National M&E Framework for social protection once finalized.  Reviewing and eventually recommending a national social protection policy for approval by the Cabinet.  Approving the implementation of new social protection programs based on recommendations from the MoGCSP.  A Social Protection Technical Committee. This will be chaired by the Chief Director, MoGCSP. Its roles will include:  Developing targets for the social protection sector for approval by the Inter- ministerial Committee and periodic reporting progress towards approved targets.  Discussing technical proposals for the implementation of policies or reforms upon request of the Inter-Ministerial Committee or MoGCSP.  Providing technical inputs into the development of a National Monitoring and Evaluation Framework for Social Protection.  Providing a forum to share experiences, lessons learned, and good practices between social protection programs. According to the Cabinet memorandum, the institutional framework for coordination will extend beyond the national level to include: Regional Social Protection Monitoring Teams; District Social Protection Committees; and Community Social Protection Committees. Once in place, detailed terms of reference (TORs) will be established for each of the committees, providing for a clear delineation of roles and responsibilities between them, the MGSCP, the relevant line ministries, and the Social Protection Sector Working Group. The memorandum also approved the creation of the National Targeting System (NTS) in MoGCSP, which will include the national registry of households to be used by all social protection programs. As MoGCSP is implementing the new institutional framework, it should consider addressing several challenges to make it stronger. It should develop various tools to coordinated the sector, including: (i) the social protection policy and vision, which is being prepared; (ii) a medium-term program with an action plan, a medium-term budget, and indicators (M&E); (iii) annual operational plans and annual budgets; and (iv) information systems (MIS) to keep track of progress in implementing programs. It is also very important to clearly establish the responsibilities of MoGCSP’s Department of Social Development (DSD) and the District Social Development Departments (DSDDs). The DSDDs should be substantially strengthened, as they are now in the 80 front line of social welfare services delivery. The DSDOs already play a key role in implementing LEAP, particularly in the payment of beneficiaries and the referral of LEAP beneficiaries to the NHIS. The DSDOs must become more proactive in following up with vulnerable groups in the community, identifying any family issues in a timely manner, and linking beneficiaries to existing programs and other services. Closer links between social services (referrals, integrated case management, and sensitization) and health, nutrition, and education programs, as well as with skill development programs, could help address existing challenges that families face and provide opportunities for beneficiaries to “leap” from poverty. While there are several ongoing programs to strengthen local government, as discussed in Chapter V, particular attention must be given to the existing capabilities of DSDD, staff composition, qualification, and incentives, and budget resources, in view of their extended responsibilities under the decentralization. Better coordination with civil society organizations is also key, as they play an important role in the delivery of social protection services. In this respect, the experience of the School Feeding Program in collaborating with NGOs and volunteers to deliver benefits is important to consider. Collaboration with civil society organizations in specific social protection activities needs to be carefully considered and, if pursued, should be strategic and pragmatic so as to make best use of their comparative advantage. In sum, to improve the efficiency of Ghana’s social protection system, institutional arrangements and coordinating mechanisms must be strengthened; staff must be trained and motivated; and better coordination is required among MDAs, MMDAs, and services organizations. The recently approved governance framework for social protection at the national, regional, district, and community levels fills in and important gap. As MoGCSP implements the new governance framework, it is important to stress the need to clearly delimit the responsibilities of key players and substantially strengthen the District Social Development Departments (DSDD), as they are now in the front line of social welfare services delivery. MoGCSP needs also to develop the tools to coordinate the sector: social protection vision and policy—which is being developed; a medium-term program and accompanying action plan and budget; and M&E and MIS to keep track of progress in implementing programs. The foregoing assessment of the social protection programs should provide some guidance for the preparation of the social protection policy. The main messages from this analysis are as follows. In recent years, Ghana has developed an impressive social protection system, but coverage and the generosity of key programs are still low. The social protection system covers most of the risks that have been identified, but coverage through LEAP of the most vulnerable groups, such families and children in extreme poverty, is limited, as is the coverage of LIPW, health insurance, and social security. In contrast, there is substantial program fragmentation and much overlap of scholarships and youth programs. Despite doubling in 2012, LEAP transfers to families still cover a small share of their basic consumption needs. On the other hand, current social security pensions are reasonable, but future replacement rates seem too generous and unsustainable. There is room to improve the accuracy of targeting of some social assistance programs. The benefit incidence analysis indicates that apprenticeship programs are not pro-poor. Moreover, scholarships and subsidies for petrol and diesel are also not pro-poor and are also regressive. Ongoing efforts to improve targeting tools are therefore timely. The decision to reduce energy subsidies will help make the social protection systems more equitable. 81 Program cost-efficiency requires further improvement. The cost-efficiency of LEAP and LIPW programs are still on the low side, but as the programs mature they should be brought up to international standards. The School Feeding Program reports very low administrative costs. The administrative costs of NHIS were reasonable during 2008–12, but increased quickly in 2013, the last year for which data are available. SSNIT appears to be cost-efficient when compare to several social security systems in Sub- Saharan Africa for which information is available, though administrative cost have increased in 2013 and 2014. The social protection system faces important sustainability challenges. The fiscal crisis has adversely affected the operations of all social protection programs, with the exception of LEAP and LIPW, as these programs are also financed by donor resources (LIPW is fully funded by donors). Payment arrears in social protection programs are pervasive. Some programs have also been expanded without taking in due consideration their financial implications. Both NHIS and SSNIT require important reforms to ensure their future financial sustainability. The GoG is fully aware of this and has already taken steps in this direction, but more needs to be done. Progress is reported in building the M&E system for LEAP and LIPW and in developing a single national household registry (GNHR). These are important tools. To develop a comprehensive M&E for social protection system, the following steps are necessary: clearly establishing the objectives to be achieved; selecting key indicators to track progress toward meeting the objectives; and providing the inputs required to build and maintain the system. To improve the efficiency of Ghana’s social protection system, institutional arrangements and coordinating mechanism require further strengthening, staff require ongoing training and motivation, and better coordination is required among MDAs, MMDAs, and services organizations. The recently approved governance framework for social protection at the national, regional, district, and community levels fills in an important gap. As MoGCSP implements the new governance framework, it is important to clearly establish the responsibilities of the key public sector actors in social protection and to substantially strengthen the District Social Development Departments (DSDDs), as they are now in the front line of social welfare services delivery. 82 V: Social Protection Public Expenditure and Management This chapter analyzes social protection expenditures in the context of overall public sector expenditure management issues. It first estimates total social protection expenditures from program data and benchmarks them against Ghana’s Sub-Saharan African peers. Taking into account the multiinstitutional characteristic of social protection, it then analyzes the formulation, execution, and auditing of sector expenditures in the context of overall budget management. A. Social Protection Public Expenditures While social protection may also be funded by international aid, NGOs, and private mechanisms, this study analyses only the public provision and financing of social protection. In this section, we estimate social protection spending, and analyze spending composition, trends, and main source of financing. Next, we benchmark Ghana social protection spending with its Sub-Saharan African lower middle-income peers. From the outset, we would like to point out the extreme difficulty we had in obtaining basic information on the social assistance programs, even for 2014. Information is fragmented, dispersed, and incomplete, and many program management units appear not to know their target group, the number of people being reached, their budget and amounts spent, and their sources of financing. 1. Social Protection Spending Ghana’s spending on social protection is limited compared to international standards, which is consistent with the findings of relatively low coverage and modest generosity of major social assistance programs. Ghana spends 1.4 percent of GDP on social protection, 0.5 percent of GDP on social assistance including scholarships (compared to a global average of 1.6 percent of GDP among developing countries),201 0.9 percent of GDP on pensions and 0.1 on ALMPs. In 2014, the government of Ghana (GoG) spent GH¢306 million on social assistance and GH¢258 million on scholarships and subsidies to senior high school students (SHS), and the Social Security National Insurance Trust (SSNIT) spent GH¢1,077 million on pensions. GOG also spent about GH¢75 million (0.1 percent of GDP) on the two active labor market programs (ALMPs) covered in this study (Table 5.1). Social assistance (excluding scholarships) represented only 0.9 percent of total public expenditures. Spending on scholarships and subsidies on SHS is almost commensurate to spending on social assistance. Poorly targeted scholarships and GoG subsidies to SHS students represent 0.2 percent of GDP in 2014 and 0.8 percent of the budget, almost the same amount invested on social assistance programs. Within social assistance school feeding represents 54 percent of social assistance spending, followed by fee waivers (16 percent of total social assistance spending), the Livelihood Empowerment Against Poverty program (LEAP) and labor-intensive public works (LIPW), (13 percent each); and other programs, with 4 percent (Figure 5.1 panel b). LEAP and LIPW, which are well targeted to the poor, are responsible for only 26 percent of total social 201 World Bank 2014e, 2015a. Scholarships are considered as part of social assistance according to international definitions and accounted in global averages. In Ghana, scholarships have not been considered as part of the 5 flagship programs in the recently approved National Social Protection Policy, as well as other programs considered in this study. Without scholarship, spending on social assistance was about 0.3 percent of GDP in 2014. 83 assistance expenditures. Table 5.1: Social Protection Actual Expenditures, 2008–14 (GH¢ thousand) Program 2008 2009 2010 2011 2012 2013 2014 SOCIAL ASSISTANCE 60,963 54,884 91,851 165,072 244,316 252,123 305,614 Percent of GDP 0.2 0.1 0.2 0.3 0.3 0.3 0.3 Unconditional cash transfers 0 6,155 6,113 16,693 28,598 34,869 38,761 LEAP 6,024 6,000 15,875 27,050 34,869 38,761 Social Inclusion Transfer 131 113 818 1,548 School feeding 36,243 5,100 4,824 63,217 98,425 96,363 164,687 Other in-kind assistance 0 12,728 43,310 40,218 50,072 31,655 13,663 Take-home rations for girls 2,410 3,440 3,240 850 2,983 6,041 School uniforms/exercise books 7,556 15,777 9,957 29,222 28,672 7,622 Elimination of child labor 2,762 24,093 27,021 20,000 Public works (LIPW) 5,991 22,973 37,537 38,530 Fee waivers 24,720 30,901 37,098 38,339 43,860 51,119 48,386 Education capitation grants 15,000 23,528 23,766 23,923 24,891 17,224 16,720 Supplementary grant (GEP) 15,100 18,700 b NHIS exemptions 24,720 30,901 37,098 38,339 43,860 51,119 48,386 Social welfare servicesc 506 614 388 580 1,587 SCHOLARSHIPS 29,528 30,772 54,147 88,160 147,034 186,827 257,971 Percent of GDP 0.1 0.1 0.1 0.1 0.2 0.2 0.2 SS scholarships 15,067 13,011 33,400 21,075 57,058 61,255 117,449 GoG subsidy to SHS 14,461 17,761 20,747 44,627 48,838 75,478 67,296 GETFund scholarships 22,458 41,138 33,742 37,467 Girls-PASS scholarships 16,352 35,759 PENSIONS (SSNIT) 198,100 303,100 450,000 409,000 532,000 770,000 1,077,000 Percent of GDP 0.7 0.8 0.9 0.7 0.7 0.8 0.9 ALMPs 0 0 6,000 63,000 84,934 75,000 75,000 Percent of GDP 0.0 0.0 0.0 0.1 0.1 0.1 0.1 National Apprentice Programd 934 LESDEP e 6,000 63,000 84,000 75,000 75,000 TOTAL 288,591 388,756 601,998 725,232 1,008,284 1,283,950 1,715,585 Percent of GDP 1.0 1.1 1.3 1.2 1.3 1.4 1.5 Percent of public expenditures 3.7 4.4 5.2 5.4 4.8 4.7 5.3 Memo: GDP (GH¢ million) 30,179 36,598 47,554 61,274 74,959 93,867 114,654 Total expenditures (GH¢ million) 7,894 8,752 11,550 13,379 20,944 27,462 32,368 Source: Data sources include administrative data and qualitative information, official government records and statistics, and bilateral interviews with program management and staff. Note: -- = not available. b. Estimated by multiplying the lowest premium (GH¢ 7.2) by the number of exempted beneficiaries. c. Includes support to orphans and other vulnerable children, counseling to children in conflict with the law, welfare services to the elderly in need, and rehabilitation of persons with disabilities. d. Includes payments in arrears to master craftsman; the cost of tools in 2014 is estimated. GETFund and District Assemblies are financing the program and financing from the Districts is not included. 84 e. Expenditure in 2014 assumed to be similar to 2013. f. Includes only exercise books Figure 5.1: Distribution of Social Assistance Spending by Type of Program with and without spending on Scholarships for SHS, 2014 (Percent) Social Social LEAP Fee welfare welfare LEAP 7% waivers services services 13% 16% 0% 0% Scholar- ships/ LIPW Subsidies 13% SHS 46% Other School in-kind feeding assis- 29% tance Other 4% School Fee LIPW in-kind feeding waivers 7% assistance 54% 2% 9% Note: The pie chart on the right includes scholarships in total social assistance spending while panel b does not include scholarships. Source:Data sources include administrative data and qualitative information, official government records and statistics, and bilateral interviews with program management and staff. The amount spent on social assistance is a fifth of the amount spent on energy subsidies. Figure 5.2 shows the amount spent (as a percent of GDP) on social assistance, pensions, and ALMP compared to the amount spend on energy subsidies in 2014, as reported by the Ministry of Finance (MoF). Energy subsidies absorbed twice as much of the budget as social assistance program and even more than pensions. Figure 5.2: Comparison of Spending on Social Protection and Energy Subsidies (Percent of GDP, Average 2012–14) 1.2 1.0 Percent of GDP 0.8 0.6 0.4 0.2 0.0 Social Scholarships Social ALMP Energy Assistance Insurance subsidies (budget outturn) Source: Data sources include administrative data and qualitative information, official government records and statistics, and bilateral interviews with program management and staff. ALMP spending refers to NAP and LESDEP only. 85 2. Social Protection Spending Trends Both social assistance and pension spending have increased relatively rapidly since 2008. Scholarship spending almost tripled in real terms since 2008, but from a low base, while social assistance and social insurance spending doubled during the same period (Figure 5.3). The most rapidly growing social assistance programs were school feeding, which nearly doubled in constant value terms. Figure 5.3: Social Protection Real Expenditure, 2008–14 (GH¢ 2006) 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 2008 2009 2010 2011 2012 2013 2014 SOCIAL ASSISTANCE SCHOLARSHIPS SOCIAL INSURANCE ALMP Source: Table 5.1 and Bank of Ghana. 3. Benchmarking of Social Protection Expenditures Ghana spends a fraction of its Sub-Saharan African lower-middle-income (LMI) peers on social assistance. Table 5.2 presents a comparison of social assistance spending (including scholarships for international comparisons) in Ghana with spending in five other lower-middle- income Sub-Saharan African countries for which data are available as well, with other 35 lower- middle-income countries worldwide, compiled by the World Bank and presented in the State of Social Safety Nets 2015.202 Ghana spends 0.5 percent of GDP on social assistance (including scholarships), compared to 2.1 percent of GDP spend by its Sub-Saharan African peers, and to 1.6 percent of GDP by the larger group of lower-middle-income countries. Table 5.2: International Comparison of Social Assistance Expenditures Public expenditures 202 World Bank 2015a. 86 (% of GDP) Ghana (2014) 0.50 Average of lower-middle-income countries (35 countries) 1.60 Average of Sub-Saharan African lower-middle-income 2.10 countries (5 countries) Lesotho (2010) 6.60 Swaziland (2010) 2.10 Bolivia (2013) 1.36 Mauritania (2012) 1.30 India (2014) 0.72 Philippines (2013) 0.57 Vietnam (2010) 0.52 Zambia (2011) 0.50 Source: World Bank 2015a, 21, and Annex D. Table refers to social assistance only (noncontributory benefits), not to spending on pension and labor market programs. Note: Excludes general subsidies. Most recent year available. B. Social Sector Public Spending and Sources of Financing This section contrasts spending on social protection with total and social spending and sources of financing. This should provide some sense of the priority accorded to social protection. The discussion also provides important background information to better understand the analysis of public expenditure management issues that follows. 1. Total Expenditures and Sources of Financing Ghana has experienced large fiscal imbalances in recent years, which the government is addressing. The overall budget (cash) deficit in the last three years have surpassed 9 percent of GDP as total expenditures increased faster than revenues, despite windfall receipts from oil exploration after 2011 (Figure 5.4). Public spending increased sharply in 2008 and 2012. For 2015, the government expected to reduce the overall (cash) budget deficit to below 8 percent of GDP by keeping a lid on expenditure growth and implementing a series of revenue-enhancing measures. The public wage bill has increased and has been a major driver of the deficit. The wage bill increased from 7.8 percent of GDP in 2008 to 9.8 percent of GDP in 2014 (Figure 5.5). It is now the single largest component of government expenditure, surpassing capital spending in 2010, accounting for around 9 percent of GDP, and claiming the lion’s share of tax revenue—more than half (World Bank 2016b). The rapid increase of the wage bill has been a result of new hires; the introduction in 2007 of a new salary structure (which took effect in January 2010) that seeks to rationalize public sector salary administration (the single spine salary structure), but which increased wages across the board; and ad hoc wage increases for some groups, in some cases retroactively. 87 Figure 5.4: Total Revenues and Expenditures and Fiscal Balance 35.0 -12.0 30.0 -10.0 25.0 -8.0 20.0 % GDP % GDP -6.0 15.0 -4.0 10.0 5.0 -2.0 0.0 0.0 2008 2009 2010 2011 2012 2013 2014P 2015B Total Revenue & Grants Total Expenditure Overall cash balance (right) Source: Appendix B, TableB.1. Note: Central government. P = projected; B = budgeted. Figure 5.5: Central Government Wage Bill, Interest Payments, Energy Subsidies, and Public Investment (Percent of GDP) 10.0 9.0 8.0 7.0 6.0 % of GDP 5.0 4.0 3.0 2.0 1.0 0.0 2008 2009 2010 2011 2012 2013 2014P 2015B Compensation of employees Interest payments Subsidies Public Investment Source: Appendix B, TableB.1. Note: Central government. P = projection; B = budget. Interest payments on Ghana’s public debt and energy subsidies also contributed to the fiscal deficit. Interest payment on public debt increased from 2.2 percent of GDP in 2008 to 3.3 percent in 2012 and almost doubled to reach almost 7 percent in 2014. Energy subsidies were another driver of fiscal deficits in 2012 and 2013. As discussed in Chapter III, according to MoF budget 88 figures, energy subsidies peaked at 1.3 percent of GDP in 2013 and declined to 0.6 percent of GDP in 2014. Subsidies to the utilities and fuels each accounted for half of total energy subsidies in 2014. External resources have financed two-thirds of capital expenditures in recent years. The share of external financing (loans and grants) of total budget expenditures has declined since 2011 and has remained at about 15 percent in recent years, while the wage bill—an item that is usually ineligible for foreign financing—began increasing. On the other hand, the share of external financing of total capital expenditures, after dropping to 47 percent in 2011, has increased, reaching 77 percent in 2014 (Figure 5.6). For 2015, the budget anticipates a decline in the share of external financing of planned capital expenditures to 63 percent. The GoG is committed to continue with the fiscal consolidation to correct existing disequilibria while protecting the most vulnerable from any potential adverse impact of the fiscal adjustment. In early April 2015, the Board of the International Monetary Fund (IMF) approved a three-year arrangement under the Extended Credit Facility. The program will support the fiscal consolidation effort initiated by the GoG in mid-2013 and structural reforms to strengthen public finances and fiscal discipline. The World Bank—with a recently approved Macroeconomic Stability for Competitiveness and Growth Credit—and other development partners are also supporting these efforts. The GoG is committed to using part “of the resulting fiscal space created to safeguard social and other priority spending under the program, including expanding the targeted social safety nets—such as the Livelihood Empowerment Against Poverty (LEAP) program.”203 Figure 5.6: External Financing of the Budget 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 2008 2009 2010 2011 2012 2013 2014P 2015B ext. finance/total expenditures (%) ext. finance/public investment (%) Source: Appendix B, Table B.1 Note: P = projection; B = budget. 203 http://www.imf.org/external/np/sec/pr/2015/pr15159.htm. 89 2. Social Spending and Sources of Financing Public expenditure management is fragmented in Ghana, which makes it difficult to analyze the allocation of public resources. The central government budget covers only the Consolidated Fund and the transfers from this fund to the earmarked funds. The five major earmarked funds (District Assemblies Common Fund, Ghana Education Trust Fund, National Health Insurance Fund, Road Fund, and Petroleum Funds) accounted for 21 percent of total central government noninterest current expenditures in 2014. The financial statements that show the executed budget prepared by the Controller and Accountant General include the accounts of the Consolidated Fund and other funding sources. Similarly, the Ghana Auditor General (GAG) report on the Accounts of the Public Sector covers only the Consolidated Fund. The GAG prepares separate audit reports for the statutory funds and MDAs and MMDAs. There is no consolidated account of the public sector or of the general government. Generally, there are no readily available accounts of actual sector spending (outturns) or financing. While all MDAs are supposed to prepare progress reports annually, not all do. The Ministries of Education and Health appear to be the exception, as they prepare comprehensive annual reports that include actual spending and sources of financing. Budget allocations from discretionary sources to the social sector have increased in the last few years, but they declined in 2014 and are projected to decline again in 2015. They increased from 5.4 percent of GDP in 2011 to a peak of 8.7 percent in 2013 and are projected to decline to 7.4 percent in 2015 (Table 5.3). As a percentage of total government discretionary spending, social spending, after dropping to a low of 36 percent in 2012, increased to 49 percent in 2014 and is budgeted at 47 percent in 2015. Over 95 percent of social spending has been on education and health, with the exception of 2010 and 2011, when the MLGRD was included as part of social spending. In 2013 and 2014, education and health absorbed 98 percent of the social budget and were planned to account for a similar share in 2015. Table 5.3: Budget Allocations from Discretionary Sources 2008 2009 2010 2011 2012 2013 2014 2015 Social (percent of GDP) 6.8 7.5 5.8 5.4 6.4 8.7 8.2 7.4 Social (percent of total 40 43 42 42 36 40 49 47 expenditures) a Education 62 62 62 60 60 54 62 68 Health 37 34 26 30 37 43 36 31 Other 1 5 11 10 3 2 2 2 Administration 16 18 11 9 15 12 8 10 Economic 16 12 13 13 11 9 13 10 Infrastructure 15 11 10 12 10 8 8 10 Public Safety 7 6 7 7 5 7 10 11 Multisector and other 6 11 16 18 24 24 12 12 MDA total (discretionary) 100 100 100 100 100 100 100 100 Memo: MDA total (GH¢ million) 4,770 5,762 5,517 6,480 10,322 15,651 16,951 18,534 Source: Appendix TableB.2. Note: Budget classification in 2008 and 2009 somewhat differ from following years; MLGRD is included in social spending in 2010 and 2011. a. GoG classification in budget documents. 90 Ghana’s expenditures on education compared well with its peers. Education expenditures grew rapidly until 2012, reaching 7.6 percent of GDP and 27 percent of total public expenditures (Table 5.4). This compares with an average of 5.7 percent of GDP for 13 middle-income Sub- SaharanAfrican countries from 2008 to 2011.204 Since then, education expenditures in Ghana have declined and were expected to reach 5 percent of GDP and 17 percent of total expenditures in 2015. Most of the expenditures are for wages, which have absorbed over 97 percent of the budget in recent years. As expected, the larger financing source is the Consolidated Fund, followed by internally generated funds (IGFs), which in 2013 accounted for 13 percent of total expenditures. IGFs are projected to continue to increase and reach 19 percent of total spending in 2015. Table 5.4: Education Expenditures and Financing (GH¢ million) 2008 2009 2010 2011 2012 2013 2014 2015 appropr. appropr. Total 1,744 1,950 2,564 3,566 5,704 5,697 5,817 6,740 GoG 1,219 1,462 1,826 2,563 4,587 4,504 4,561 4,944 GetFund 212 151 313 518 361 197 104 IGFs a 164 210 302 354 631 718 952 1,307 ABFAb -- -- -- -- 11 9 -- 342 Development partners 101 95 65 127 114 269 200 156 HIPC/MDRI 47 31 59 2 -- -- -- -- Memo: % of GDP 5.8 5.3 5.4 5.8 7.6 6.1 5.1 5.0 % of total expenditures 22.1 22.3 22.3 26.7 27.2 20.7 18.0 17.2 GDP 30,179 36,598 47,554 61,274 74,595 93,867 114,654 135,011 Total expenditures 7,894 8,752 11,500 13,379 20,944 27,462 32,368 39,132 Source: Education Sector Performance Report of the Ministry of Education. Note: appropr. = appropriation a. Internally generated funds. b. Annual Budget Funding Amount. Ghana’s health expenditures also used to be higher than its peers. Health expenditures in Ghana grew rapidly until 2012, reaching 3.3 percent of GDP and 12 percent of total public expenditures (Table 5.5). This compares with an average of 3 percent of GDP for 13 middle- income Sub-Saharan African countries from 2008 to 2012.205 Since then, health expenditures in Ghana have declined andwere expected to reach 2.3 percent of GDP and 8 percent of total expenditures in 2015. Wages absorbed about 57 percent of the budget in 2013 and were expected to decline to about half in 2015. As expected, the largest financing share is from the Consolidated Fund, followed by internally generated funds (IGFs), which in 2013 accounted for 20 percent of total expenditures. IGFs are projected to continue to increase and reach 32 percent of total spending in 2015. Table 5.5: Health Expenditures and Financing (GH¢ million) 2008 2009 2010 2011 2012 2013 2014 2015 appropr. appropr. Total 602 546 718 1482 2479 2717 3354 3069 GoG 286 344 400 771 1750 1521 1209 1308 204 World Development Indicators. 205 World Development Indicators. 91 NHIA 42 23 15 6 -- -- IGFa 132 108 208 392 442 569 1364 1004 ABFAb -- -- -- -- -- -- -- 44 DPc /budget support/credits 141 83 110 296 272 621 781 713 HIPC/MDRI 1 11 -- -- - -- -- -- Memo: % of GDP 2.0 1.5 1.5 2.4 3.3 2.9 2.9 2.3 % of total expenditures 7.6 6.2 6.2 11.1 11.8 9.9 10.4 7.8 GDP 30,179 36,598 47,554 61,274 74,595 93,867 114,654 135,011 Total expenditures 7,894 8,752 11,500 13,379 20,944 27,462 32,368 39,132 Source: Ministry of Health, Holistic Assessment of Health Sector Work Program and appropriation bills. Note: appropr = appropriation. a. Internal generated funds. b. Annual budget funding amount. c. Development partners. The budget for the Ministry of Gender, Children, and Social Protection (MoGCSP) is a tiny share of total public expenditures. The MoGCSP was established in January 2013. It replaced the Ministry of Women & Children’s Affairs (MOWAC) and incorporated the Department of Social Development and the Social Protection Division of the then-Ministry of Employment & Social Welfare. Therefore, we present MoGCSP accounts only for the 2013–15 period. As can be observed in Table 5.6, the MoGCSP budget amounted to 0.04 percent of GDP and 0.15 percent of total public expenditures in 2014. Table 5.6: MoGCSP Expenditures (GH¢ million) 2013 2014 2015 appropriation Total 24.4 47.9 43.6 Compensation 2.1 17.3 -- Goods and services 18.1 20.6 -- Capital expenditures 4.2 4.2 -- Memo: % of GDP 0.03 0.04 0.03 % of total expenditures 0.09 0.15 0.11 GDP 93,867 114,654 135,011 Total expenditures 27,462 32,368 39,132 Source: GIFMIS and appropriation bills. The growth of social assistance spending has outpaced the increase in public expenditures since 2010. In 2013, social assistance spending in real terms declined, but in 2014 it recovered as result of a large increase in the school feeding and scholarship programs. For the 2008–14 period, real social assistance spending grew by more than 2.5 times and outpaced the growth of education and health spending (Figure 5.7).206 However, the growth of social assistance spending is from a very low base. 206 There is overlap between social assistance and education and health spending; thus the trends are not independent. 92 Figure 5.7: Trends in Total Public, Education, Health, and Social Assistance Real Spending 300 250 200 2010=100 150 100 50 - 2008 2009 2010 2011 2012 2013 2014 Total public Education Health Social Assistance Source: Tables 5.1, 5.4, and 5.5. Nearly one-fifth of social assistance expenditures is financed from external sources. Table 5.7 shows the spending of social assistance programs and their financing from local and external sources. In 2014, external resources financed 19 percent of total social assistance expenditures. Most programs are financed with local resources, with the exception of LEAP, LIPW, and the Girls-PASS scholarships (DFID). In 2014, about half of LEAP expenditures were financed with external resources (DFID). This share should increase in 2015, when the World Bank Additional Financing starts disbursing. According to the program management unit, in 2014 LIPW was financed completely by external resources. In this context, it is important to keep in mind that a significant amount of external support to GoG has been through general budget support, and has not been earmarked to any particular sector. This budget support is consolidated and channeled through the so-called Multi-Donor Budget Support (). These funds supported the implementation of the government’s poverty reduction strategies and the Ghana Shared Growth and Development Agenda (GSGDA). Development partners (DP) contributing currently to the MDBS include: the African Development Bank, Canada, Denmark, the European Union, France, Germany, Japan, Switzerland, the United Kingdom, and the World Bank. Table 5.7: Social Assistance and Scholarships Financing, 2014 Program Spending Source of financing Source of financing (GH¢ 000) (GH¢ 000) (%) GoG DP GoG External LEAP 38,761 20,000 18,761 51.6 48.4 School feeding 164,687 158,063 6,624 96.0 4.0 Other in-kind assistance 13,663 7,622 6,041 55.8 44.2 Take-home rations for girls 6,041 -- 6,041 0.0 100.0 School uniforms/exercise books 7,622 7,622 -- 100.0 0.0 93 Public works –LIPW 38,530 -- 38,530 0.0 100.0 Fee waivers 48,386 48,386 -- 100.0 0.0 Education capitation grants 16,720 16,720 -- 100.0 0.0 Supplementary grant 18,700 -- 18,700 0.0 100.0 NHIS exemptions 48,386 48,386 -- 100.0 0.0 Scholarships/Subsidies to SHS 257,971 222,162 35,759 86.1 13.9 Scholarships (SS) 117,449 117,399 -- 100.0 0.0 GoG subsidy to SHS 67,296 67,296 -- 100.0 0.0 GETFund scholarships 37,467 37,467 -- 100.0 0.0 Girls-PASS scholarships 35,759 -- 35,759 0.0 100.0 Social welfare services 1,587 1,587 -- 1000.0 0.0 TOTAL 563,585 457,820 105,715 81.2 18.8 Source: World Bank staff estimates. Note: Does not take into consideration budget support. Dependence on external financing is higher for social assistance than for education and overall public expenditures, but lower than for health (Figure 5.8). Programs such as LEAP and LIPW rely to large extent on external financing. Figure 5.8: External Finance as a Share of Total Expenditures (Percent) 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 Total Public Education Health Social Assistance Source: TablesB.1, 5.2, 5.5, and 5.6.Special Funds Budget formulation at both the national and program level is hindered by the existence of five major statutory funds that received earmarked funds by legislation. These funds represent about 20 percent of central government revenues.  District Assemblies Common Fund. The DACF was established in 1993 to give the Municipal, Metropolitan, and District Assemblies financial autonomy to make decisions. The DACF has since become an important tool for fiscal decentralization. The District Assemblies Common Fund Act 1993 (Act No. 455) regulates allocations to the MMDAs on the basis of a formula approved annually by Parliament.207 In 2008 the proportion of the 207 Every year, the Administrator of the DACF is required by law to prepare a formula for the approval of Parliament indicating how the Fund should be distributed to the various DAs. The formula takes account of the basic services (45 percent of total allocation) (education, health, water, and tarred roads) in the districts, as well as the MMDA efforts to generate their own revenues; pressures put on facilities as a result of rural/urban migration, measured by population density; and a percentage of the Fund shared equally among the MMDAs before the other factors are applied. The 94 revenues allocated to the DACF was increased from 5 to 7.5 percent of total central government revenues.208 Act 455 gave the Minister of Local Government and Rural Development the authority to issue guidelines for the utilization of the DACF. Accordingly, only part of the DACF funds is disbursed directly from the DACF to District Assemblies. This is referred to as direct transfers. Another part of the resources (called indirect transfers) is used to finance other GoG programs, including school lunches, youth programs, programs for persons with disabilities, Members of Parliament for constituency projects, and the operations of the Regional Co-ordinating Councils (Table 5.8).  Ghana Education Trust Fund. The main objective of GETFund, as stipulated in the Ghana Education Trust Fund Act 2000 (Act No. 581), is to supplement the resources available in the general budget for expenditures on education infrastructure, scholarships and loans schemes for underprivileged students. Most areas of spending are for higher education. Some 2.5 percent points of the annual collections of the value-added tax (VAT) is earmarked for the GETFund. It also receives revenues from other sources, including interest earned on invested reserves.  National Health Insurance Fund. The objective of the NHIF is to support the implementation of the National Health Insurance Act, which seeks to guarantee access to basic health care services for all Ghanaian residents. Revenue derives from a 2.5 percent point earmarked share of VAT revenues, 2.5 percentage points of the SSNIT contribution, insurance premiums (for those who are not exempted), and interest earned on reserves.  Road Fund. The objective of the Road Fund is to finance periodic maintenance and rehabilitation of public roads. Revenue derives from levies and user fees based on actual use of services. The Road Fund is sometimes a source of funding for the Scholarships Secretariat scholarship program.  Petroleum Holding Fund. Since December 2010, Ghana has been extracting oil, and this has since become an important source of revenue. Oil revenues are subject to special financial management arrangements and are governed by the Petroleum Revenue Management Act 2011. This includes the establishment of a Petroleum Holding Fund. The Fund’s resources are transferred to (i) the Annual Budget Funding Amount (ABFA) held in the Consolidated Fund to implement the annual budget; (ii) the Ghana Stabilization Fund, from which resources may be mobilized in response to revenue shortages emanating from price fluctuations; and (iii) the Ghana Heritage Fund, to be used once the oil and gas deposits are exhausted. The forecasts on oil revenue are included in the overall revenue forecasts in the three-year medium-term fiscal framework. MDA and MMDA also have internally generated funds (IGF) that help finance their operations. Some of these resources are retained and use by the MDAs and MMDAs according to MDA Retentions Act 735 of 2007. Others are transferred to the MoF and transferred in part back to them. These resources are included in the appropriation bill. formula for 2014 has a cap of 10 percent of total funds allocated that can be used for recurrent expenses. The remaining 90 percent is for development projects. See http://www.commonfund.gov.gh/. 208 Revenues for this purpose are defined as total revenue transferred to the Consolidated Fund, excluding oil and mineral revenues and donor support. 95 3. Financing Arrangements Two governing systems operate at subnational level in Ghana: traditional authorities209 and local government.  Traditional authorities. This system of administration is regulated by the Chieftaincy Act of 2008. The system comprises Traditional Councils, which in turn oversee the Stools (traditional kingdoms and chiefdoms throne). A key area of authority of Traditional Councils is over land and land use issues. At central government level, the responsibility for these issues lies with the Ministry of Chieftaincy Affairs.  Local government. There are 10 regions that coordinate decentralization issues through the Regional Coordinating Councils (RCCs). The Regions comprise four level of governance: (i) Municipal, Metropolitan and District Assemblies (MMDAs); (ii) sub- Metropolitan Councils; (iii) Zonal, Urban, Town and Area Councils; and (iv) Unit Committees. There are 216 MMDAs coordinated through the RCCs which, at the central government level, are under the responsibility of the Ministry of Local Government and Rural Development (MLGRD). The main source of funds for Traditional Councils are from personnel emolument payments and funds collected and distributed through the Office of the Administrator of Stool Lands (OASL). These include mineral development royalties, timber royalties, and land rents. Royalties and land rents are collected by central government agencies on behalf of the Traditional Councils. MMDAs have three main sources of funding: (i) transfers from the central government; (ii) financing from development partners; and (iii) internally generated funds from property taxes, levies, tolls, licenses, fines, and the like. Transfers of resources to MMDA are made through the District Assemblies Common Fund (DACF), the District Development Facility (DDF), and the Minerals Development Fund, and sector-specific grants from the Consolidated Fund to cover salaries, goods and services, and purchases of assets. Fiscal decentralization has been under way for several years in Ghana and some progress has been made. The 1992 Constitution called for the decentralization of government responsibilities and activities, including fiscal decentralization.210 As mentioned, in 1993, the District Assembly Common Fund was established to give the MMDAs financial autonomy to make decisions locally. The Local Government Instrument (LI 1961) of 2009 transferred several thousand civil servants to the Local Government Service. In 2010, the government’s Decentralization Policy Framework identified fiscal decentralization as critical to enable MMDAs to have the appropriate funding and enable local governments to improve public service delivery, and established the Fiscal Decentralization Unit (FDU) within the Budget Department. In 2011, several functions were devolved to the MMDAs, but most related financing continued to be centralized. Beginning in 2012, the MMDAs began preparing the “composite budget,” which 209 Traditional authorities are not recognized within the intergovernmental framework of Ghana. 210 The MMDAs are regulated by the Constitution, the Local Government Act (1993), the Local Government Service Act (2003), and the Local Government Service Regulations (2008). 96 integrates the deconcentrated expenditures of MDAs in MMDAs and their main source of financing.211 The District Assemblies (DAs) are at the center of intergovernmental transfer system. The DA is the highest authority at the district level, and has deliberative, legislative, and executive functions. MMDAs’ major functions include public health and disease control, primary and junior high school education, water supply, waste management and sanitation, agriculture, agricultural extension, animal health and veterinary services, arts, crafts and culture, local power supply, town and country planning, regulation of buildings, road and street construction and maintenance, maintenance of highways and public buildings, public libraries, regulation of markets, disaster relief and management, forestry control and management, tourism, traffic management, fire service, and sports development, community development, and social welfare. However, the degree of discretion granted to MMDAs by the central government differs. In some sectors, such as health and education, functions are deconcentrated and control remains in the center. In others areas, such as planning and construction of roads, parks, and cemeteries, functions have been devolved. The MMDAs are now responsible for social welfare services and community development. Staff to deliver the devolved function have been moved to the Local Government Services and are paid centrally. The other expenditures for goods and services are included in the composite budget of the MMDA and financed by transfers from the Consolidated Fund. A thorough assessment of existing human and physical capabilities at the local level and those required to discharge the devolved functions should be made. It appears that the devolution of functions was not followed by the required resources (see Box 5.1 for an example of MMDA budget for social services). Most MMDA resources continue to be under the control of the central government. As mentioned, personnel emoluments continue to be paid centrally. The amount of DACF funds distributed indirectly to MMDAs is greatly reduced, as various amounts are taken off by MDAs before being distributed to the districts. Table 5.8 indicates that in 2008 about 50 percent of the DAFC resources were transferred directly to the MMDAs; by 2013 direct transfers to MMDAs 211 As part of its decentralization strategy, the GoG has developed an Intergovernmental Fiscal Framework and Action Plan for 2014–18, which is being implemented. The Plan covers revenues and internally generated funds, intergovernmental transfers, borrowing, public financial management (planning and budget formulation, budget implementation), budget reporting, external auditing, and budget monitoring and evaluation. Among the activities considered in the plan are: (i) developing a performance benchmark for the internally generated funds to encourage revenue mobilization, with possible linkages to performance-based transfers through the Functional Organizational Assessment Tool (FOAT) system; (ii) realigning DACF to receive a constitutionally mandated share of central government revenues released directly to MMDAs for funding development; (iii) designing and implementing a new, harmonized formula-based transfer system for all government and development partner transfers to MMDAs, along with an implementation transition plan; (iv) designing and implementing an intergovernmental transfer monitoring system; (v) developing and strengthening the Public Investment Management (PIM) framework for district assembly- 97 level capital investment; (vi) developing and implementing a strategy to implement program-based budgeting at the district level; and (vii) designing and implementing training programs on procurement, contract monitoring, supervision and management, accounting, internal controls, and internal audits for leaders and staff at the MMDA level (http://www.mofep.gov.gh/?q=divisions/fdu/Intergovernmental-fiscal-framework-action- plan(http://www.mofep.gov.gh/?q=divisions/fdu/Intergovernmental-fiscal-framework-action-plan-). had been reduced to only 18 percent to total DACF. Direct transfers in 2013 in nominal terms were less than those in 2008. Table 5.8: District Assembly Common Fund Disbursements 2008 2009 2010 2011 2012 2013 Total 217.0 188.6 340.4 487.9 571.7 648.1 Direct Transfers 116.4 78.2 107.2 168.1 84.0 115.5 Indirect Transfers 100.6 110.3 233.2 319.6 487.4 532.6 Constituency labor (MP) 12.1 9.3 7.7 10.4 36.8 29.7 RCC 2.7 3.5 5.4 8.4 9.1 7.7 People with Disability -- -- 3.6 12.4 7.7 8.8 NYE 63.0 43.5 95.0 117.1 -- -- DDF 14.8 14.1 22.7 -- 50.4 24.0 Cured Lepers -- -- 0.9 0.5 1.7 1.7 Reserve fund 5.0 3.0 -- 15.5 28.4 35.7 Monitoring fund 1.3 0.8 -- -- -- 3.4 National Ass. of Local Authorities 0.8 0.7 1.3 -- -- -- Training fund 0.9 1.2 -- 0.5 2.6 3.4 Waste management -- 9.6 -- -- 37.3 36.9 CIP (MLGRD) -- 1.7 7.0 -- -- -- School feeding -- -- -- -- 54.5 77.3 Sanitation guards -- -- -- -- 10.4 14.1 Seed capital for new MMDA -- -- -- -- 27.5 24.6 National sanitation program -- -- -- -- -- 6.4 Other -- 22.9 89.6 154.8 221.0 258.9 Source: District Assembly Common Fund Annual Reports. Note: CIP = Community Initiative Projects [[define]]; DDF = [[define]]; MP= [[Members of Parliament?]]; NYE = national youth employement . MMDAs spend a small and declining share of the DACF direct resources on social services. MMDAs’ spending on social services has declined from 25 percent to only 15 percent in recent years (Table 5.9). It is not possible to determine how much MMDAs spent on traditional welfare services because there is no budget line item for this category. Social services line items are: basic, secondary, and tertiary education; sponsorships; sports and culture; water; health; and Community Initiative Projects (CIP). Table 5.9: District Assembly Common Fund by Type of Spending 2008 2009 2010 2011 2012 2013 Utilization of MMDA direct resources (%) 100 100 100 100 100 100 Economic ventures 18.2 16.0 15.2 12.9 10.9 12.5 Social servicesa 24.8 25.9 27.4 22.5 15.2 15.0 Administration 36.5 43.1 42.8 48.3 48.9 53.7 Environment 10.5 8.7 10.0 10.0 12.0 10.9 Constituency Labor 9.9 6.4 4.7 6.3 12.9 8.0 Source: District Assembly Common Fund Annual Reports. a. Includes basic and secondary education, sponsorships, sports and culture, water, health, and Community Initiative Projects (CIP). 98 The decentralization reform efforts are being supported by the development partners. The World Bank financed Local Government Capacity Support Project (US$175 million, from 2012 to 2017) builds on the incentive-driven approach to promote MMDA performance that was initiated in 2008 through the District Development Facility performance grant. It focuses on local government public financial management (PFM) systems. The specific reform areas being supported are: (i) budget reform; (ii) reporting and auditing systems; (iii) revenue management; (iv) asset management; and (v) social accountability. The project complements existing development partner support for decentralization. Support for administrative decentralization is being provided by the governments of Denmark (through DANIDA), France (through AFD), and Germany (through GTZ and KfW) , with further support planned by the European Union (EU) for personnel decentralization and civil service reforms (Local Government Service Secretariat). 212 C. Budget Management While the process of public expenditure management in Ghana has improved in recent years, major challenges remain in the areas of budget formulation, execution, and auditing. Public expenditure management seeks to achieve three objectives: fiscal discipline, strategic allocation of resources, and efficient service delivery. Taking into account the multiinstitutional characteristic of social protection, this section analyzes the main challenges with the formulation, execution, and auditing of sector expenditures in the context of overall public financial management and ongoing reforms. 1. Formulation Revenue projections have been quite off target in recent years. This discrepancy affects expenditure allocations, creates uncertainty among the budget executing agencies, and reduces the credibility of the budget process. Revenues in 2011 and 2012 were above the projected levels because of the beginning of oil revenues and windfall from the mining sector. In 2013 and 2014, revenue projections were overly optimistic. As Figure 5.9 shows, recent deviations result mainly from overoptimistic projections of central government direct and indirect taxes. In 2013, actual revenues were GH¢1.88 billion or 11 percent below the amount budgeted. 212 World Bank 2011b, 4. 99 Figure 5.9: Deviation between Actual and Budgeted Revenues (GH¢ million) 2008 2009 2010 2011 2012 2013 2014P 2500 2000 1500 1000 500 0 -500 -1000 -1500 -2000 -2500 Tax revenue Non tax revenues Source: MoF Budget Statements. Budget formulation is complicated by the existence of five major statutory funds that received earmarked funds by legislation. As discussed, these statutory funds represent about 20 percent of all revenue of the central government. The statutory funds transfer resources to the respective sectors, but these allocations are not detailed in the budget proposal. Budgeting for expenditures from the statutory funds follows separate rules as set out in their respective legislation. The budget preparation guidelines issued in June of each year instruct MDAs and MMDAs to prepare their budget according to their annual action plans, which should be consistent with the policy orientation and priorities established in the GSGDA II (2014 –17). MDAs should prepare their budget on a program basis. In June they are given spending ceilings on compensation, goods and services, and capital expenditures financed by the Consolidated Fund, and expenditure ceiling from internally generated funds and development partners. Allocations from statutory funds are not included. Once approved by the Cabinet, revised budget ceilings are issued toward the end of October and may deviate significantly from earlier guidelines. MMDAs receive budget guidelines for 2015–17 budget preparation from MoF aligned to GSGDA II. The MMDAs are instructed to prepare their composite budgets by activity, and starting in 2016 by program. MMDAs prepare their budget within the ceiling provided by the MoF, which include transfers from the Consolidated Fund to pay for devolved functions, namely compensation of employees, goods and services, and assets, projected internally generated funds, and funding from the DACF (7.5 percent of projected revenues). MoF communicates revised ceilings to MMDAs once they are approved by the Cabinet. The detailed sector-disaggregated ceilings (for agriculture, roads, community development and social welfare, and the like) for each MMDA are provided and should be strictly adhered. However, MMDAs are encouraged to allocate funds from the internally generated revenues to complement the allocations from the central government for devolved functions, in accordance with the MMDA’s mission and vision. The MMDA budgets are approved by MMDA Assemblies and submitted to the Parliament at the same time as the overall central government budget. 100 The budget guideline for 2015–17 includes a freeze on employment and disclosure of all payment arrears. MDAs and MMDAs are asked to take into consideration all outstanding commitments and arrears and budget appropriately within the ceilings provided. MDAs are also to provide details of all outstanding commitments and arrears. In the budget implementation guidelines, MDAs and MMDAs are reminded that all recruitments and replacements are suspended until the Public Services Commission (PSC) concludes its ongoing Human Resource (HR) Audit of the Public Service. The social protection sector does not have a budget per se. The formulation of the social protection programs is made by MDAs and MMDAs responsible for the specific programs and is incorporated in their institutional or composite budget. There is no formal consolidated or summary budget of the social protection sector, as there is, for instance, for poverty reduction, as in most cross-cutting issues. The formulation of social protection programs is expected to be informed by the GSGDA and the National Social Protection Strategy. The formulation of the social protection programs suffers the same problems as other sectors: namely, the unpredictability of budget allocations from MoF and statutory funds. It is not possible to identify social protection program in the program-based budgeting sector submissions. For the 2015–17 Medium-Term Expenditure Framework prepared by the MDAs and published by the MoF, it is not possible to identify any of the social assistance programs. While in a very few cases social protection programs are identified as a subprogram (such as the School Feeding Program), not even in these cases are the budgeted expenditures shown. In the MMDA composite budget, is possible to identify the allocation to social welfare services as well as the key activities and their financing source (Box 5.1). 101 Box 5.1: Budget Formulation and Execution in the District Assembly of Shai-Osu Doku The District of Shai-Osuduko (formerly Dangme West District) forms part of the 16 MMDAs in the Greater Accra Region and is the largest. It is located in the southeastern part of Ghana. The capital town is Dodowa. It had a population of 52,170 as of 2010. The district is predominantly rural and has three area councils and one town council. Shai-Osuduko ranks 46 out of 216 districts (a lower rank is better) in the UNICEF-CDD Ghana’s District League Table 2014, which takes into consideration the district situation in six indicators (BECE pass rates; skilled delivery at birth; rural water coverage; open defecation free certification; police personal coverage; and fulfillment of district administrative FOAT).a Among other devolved functions, Shai-Osuduko has been given the main responsibility for the delivery of social services. The District Social Welfare and Community Development Unit has 21 social workers who are paid by the Local Government Service Budget. District of Shai-Osuduko budget preparation is made on the basis of the guidelines received from the MoF and after extensive consultations throughout the District. According to the budget guideline for 2015–17, budget hearings should be conducted for all MMDAs between September and October at both the district and regional level before the budgets are approved by the District General Assembly. The district hearings are coordinated by the District Budget Committees. The regional hearings should provide quality assurance and also verify that programs and projects outlined in the budget are in line with the GSGDA II, District Medium-Term Development Plan and annual action plans. The Regional Ministers chairs the budget hearings. After the regional budget hearings, all MMDAs are required to present the budget to the General Assembly for approval by October 31. The budget is then submitted to the Parliament together with the national budget. It is the District Assembly that has the last word concerning the budget, though there are exceptions in emergencies where the central authorities may revise the approved appropriations. According to District of Shai-Osuduko authorities, budget execution has been affected by lack of disbursement of DACF funds to the District. In 2014, the District received only three of the four quarterly disbursements (one was in arrears from 2013 and two were in arrears from 2014). In addition, the District has not received the grants from the District Development Facility since 2012. Budget auditing is done by the Auditor General every year. At local level, officials explain what was accomplished with the execution of budget during the consultations for the preparation of the new budget. The District Assembly budget for the District of Shai-Osuduko for 2015 is GH¢12.9 million. The budget for social services and community development GH¢132,287, or 1.1 percent of the District’s total budget. This is equivalent to less than one dollar a year per habitant. Over 85 percent of the social services budget is for wages, which are paid by the central government. The remainder (US$6,779) is distributed through several activities, mostly training and sensitization activities. Very few resources appear to be available to provide direct support of vulnerable groups. Table B5.1.1District of Shai-Osuduko 2015 Budget for Social Services GH¢ US$ US$/habitant % Total 12,211,658 4,210,917 80.72 Social welfare 132,287 45,616 0.87 100.00 Wages (GoG) 112,592 38,825 0.74 85.11 Goods & Services (source of funding) 19,659 6,779 0.13 14.86 GoG 14,845 5,119 0.10 11.22 IGF 2,450 845 0.02 1.85 DACF 2,400 828 0.02 1.81 102 Box 5.1: Budget Formulation and Execution in the District Assembly of Shai-Osuduko (Continued) The 2015 working plan for social welfare appears well articulated to the MCGSP priorities. Among other things it includes activities related to children, probation, the disabled, youth, and LEAP. District of Shai-Osuduko, Social Welfare Work Plan Estimated Cost ACTIVITY STRATEGIES OUTPUT RESOURCES GH¢ 1.Establish 4 child Conduct baseline survey 4 C.P centers Stationery, T&T, 1,245 panel systems for established in the venue, and snack the Area councils Organize sensitization for district WDO by Dec 1,500 stakeholders 2015 Selection & training of Stationery, files and 1,500 panel members snack 2. Visit Day Care Identify and register all 100 DCC registered T&T, files and 1,945 Centers DCCs district wide and inspected by Dec stationery Embark on inspection on 2015 T&T and stationery quarterly basis 3.Probation Work Conduct Sec Enquiry cases, At least 12 Enquiry Stationery, files & 3,000 and making reports and 12 juvenile snack recommendations put on probation Follow up on probationers in T&T 1,400 their homes. 4. Train 120 Register all PWDs youth 120 PWDs youth T&T, stationery, unemployed and equipped with income- venue, snack physically Conduct needs assessment generating skills T&T challenged and Select and train 30 PWDs T&T, stationery, 1,400 youth in IGAs, June quarterly venue, and snack 2015 5. Monitor LEAP List homes of 50 20 case management T&T 2,400 Activities beneficiaries quarterly cases handled 6. Train 20 daycare Sensitize operators of 60 daycare operators T&T 800 centers operators daycare centers on their roles and attendants trained and 40 attendants as stakeholders by end of Dec 2015 by Dec 2015 Identify and register T&T 800 attendants 7. Set up child Sensitize stakeholders Four Area Councils T&T, snack, files, 1,214 protection with 20 child stationery, venue committees in the protection four Area Councils Identify and train 5 committee members by Dec 2015 committee members trained by end of 2015 quarterly Source: Information obtained during a meeting in Dodowa with MMDA authorities, March 5, 2015 and MoF. Note: The original work plan includes the person responsible for each activity. IGAs = Income Generating Activities; PWD = [define/persons with disabilities; a. UNICEF and CDD 2014. FOAT = Functional Organizational Assessment Tool. 2. Execution The overly optimistic budget projections and lack of fiscal discipline lead to significant differences between the approved budget and outturns. Midyear budget revisions lead to unanticipated cuts in programs, with negative consequences for program implementation. 103 Transfers to statutory funds suffer significant delays, which in turn affect budget execution by MDAs and MMDAs and payments to contractors and suppliers. Fiscal discipline in budget execution has been weak. This results from the unpredictability of funds and delays in the release of funds, as well as ineffective expenditure controls, reflected in overspending and in the accumulation of expenditure arrears, which undermine fiscal discipline and hamper service delivery. Figure 5.10 shows that actual expenditures on wages and subsidies have systematically exceed the amounts budgeted, particularly since 2012. In 2012, the actual wage bill was 19 percent greater than budgeted, mainly because of wage hikes associated with the introduction of the Single Spine Pay Policy. This increase was partially compensated by a sharp drop in investment. In 2013 and 2014, the wage bill was 5 percent higher than budgeted. These increases in the wage bill were compensated by another cut in investment and in goods and services spending relative to the amount budgeted, which surely affected most social spending, including social protection. Figure 5.10: Difference between Actual and Budgeted Expenditures 1,500 1,000 500 0 -500 -1,000 -1,500 -2,000 -2,500 2008 2009 2010 2011 2012 2013 2014P Wages G&S Subsidies Investment Source: MoF Budget Statement. MoF authorizations to purchase goods and services are subject to significant delays, which have a negative impact on MDA activities and lead to large arrears. Indeed, the large accumulation of arrears underpins the lack of fiscal discipline. In 2010, the stock of government payment arrears reached the equivalent of 9.1 percent of GDP, including: 2.8 percent to private contractors; 1.6 percent to statutory funds; 3.8 percent to public enterprises (Tema Oil Refinery, Volta River Authority, Electricity Company of Ghana, and other state-owned enterprises); 0.3 104 percent to pensions; and 0.6 percent to wages.213 Since then, the GoG has been trying to reduce the stock of arrears, but additional domestic arrears were accumulated in 2014.214 The implementation of social protection programs suffers from these same problems. At the national level, three major issues are reported by program managers: late disbursement of funds; incomplete disbursement of funds from the MoF; and unpredictability of disbursement from statutory funds. These problems are particularly worrisome in the case of cash transfers provided by LEAP because the beneficiaries count on steady and predictable transfers to finance their basic needs and avoid other impoverishing means of financing. LEAP management reports that payment delays have been mostly eliminated in recent months. At the local government level, the same types of issues prevail: unpredictability and incomplete disbursement of funds from MoF and statutory funds. Expenditure arrears on social protection programs are shown in Table 4.16 in Chapter IV. Since the introduction of the composite budget in 2012, disbursement to MMDAs for the execution of their budget has been simplified. Previously each MDA would submit a separate request for disbursements for their MMDA-related expenditures. Today, the MLGRD makes only one request for disbursements per quarter on behalf of all MMDAs. However, challenges remain in meeting all the amounts appropriated to the MMDAs. For instance, in 2014, only three of the expected four quarterly disbursements from DACFs were made. This disbursement shortfall is expected to be compensated in 2015. Most resources of the DACF are executed at the central level. A provision in the District Assembly Common Fund Act, 1993(Act 455) gave the Minister of Local Government and Rural Development the authority to deduct portions of the DACF at the source for specified purposes, release portion of the DACF to MMDAs as tied grants, and authority to use the remainder of the DAFC only in indicative sectors. The proposed Consolidated Local Government Bill would eliminate this provision (section 9) of the District Assembly Common Fund Act so that all resources would be transferred to the MMDA. The proposed bill would consolidate in one legal instrument several other Acts, including: the District Assemblies Common Fund Act, 1993 (Act 455); the Local Government Act, 1993 (Act 462); the National Development Planning (System) Act, 1994 (Act 480); the Local Government Service Act, 2003 (Act 656); and the Internal Audit Agency Act, 2003 (Act 658). This bill should be approved as soon as possible. 3. Auditing The Ghana Audit Service (GAS) is the supreme audit institution in Ghana, and its external audit practices “broadly adhere to international auditing standards,” according to the latest Public Expenditure and Financial Accountability Review.215 The GAS has access to all public records and has independence in the preparation of the annual audit work plan. The jurisdiction of the GAS includes external audit of public accounts of all public offices, courts, the central and local government offices, public universities and educational institutions, public corporations, and all other bodies or organizations established by an Act of Parliament. The GAS conducts audit of the Consolidated Fund accounts prepared and submitted by the Controller and Accountant General and of the MDA annual financial statements. In addition to the financial audit, each year the GAS 213 World Bank 2011a, 11. 214 http://www.imf.org/external/np/sec/pr/2015/pr15159.htm. 215 Ecorys Macro Group 2013, 121. 105 undertakes a performance audit of selected MDAs and programs. For instance, in 2013, it audited the Ghana National Fire Service program and the Extension of Electricity program. The Public Accounts Committee (PAC) of the Parliament reviews of public accounts are covered live on television and radio. PAC invites ministers of state, heads of departments, and any other officer(s) linked to audit findings for questioning. The accounts of the MMDAs are reviewed and audited every year. At the end of the fiscal year, the MoF Decentralization Unit produces a report for MoF management reviewing the execution of the MMDA budget. On the other hand, the Auditor General audits the accounts of the MMDAs, the DACF, and other statutory funds each year, and submits the reports to the Parliament. Lack of follow-up by MDAs and MMDAs on the recommendations of the Ghana Auditor General is a recurring issue. The Audit Service Act of 2000 establishes Audit Report Implementation Committees (ARIC) in MDAs and MMDAs to implement GAG’s recommendations. However, these committees have not been functioning appropriately. Therefore, the same problems keep on reoccurring: (i) misappropriation of funds; (ii) fraudulent payroll deals; (iii) nonadherence to existing public procurement laws and process; (iv) irregularity in pension schemes; (v) pure embezzlement and theft; and (vi) inappropriate stores procedures and inappropriate record keeping. GAG is issuing new guidelines to address existing problem with the ARIC. The fragmentation of resource appropriation through the budget and various funds and the fragmentation of expenditure reporting make oversight of expenditures very difficult. For instance, the Financial Statement by the Controller and Accountant General (CAG) reports only on the Consolidated Fund and does not report on expenditures incurred from retained internally generated funds, development partners, or statutory funds. Similarly, information on payment arrears is not easily evaluated because there is not a single definition of expenditure arrears and no systematic and consistent mechanism for monitoring and measuring expenditure arrears.216 The audited financial statements prepared by the CAG are presented in a different classification from the budget documentation, and information on some categories of expenditures (IGFs and transfers) is lacking in the financial statements. While statutory funds accounts are not audited by the Controller and Accountant General, they are audited annually by the Ghana Auditor General (GAG), and the reports are submitted to the Parliament. Formal auditing of social protection programs is part of MDA and MMDA auditing and therefore faces the same constraints they face. Some of the social protection programs financed by development partners are required to contract periodic external auditing in addition to internal auditing. An important auditing function is provided by the public expenditure tracking survey (PETS). As discussed, the PETS on the distribution of capitation grants and books revealed very poor recordkeeping at the district and school levels and significant leakage of resources (Chapter IV). Social accountability is an important approach to monitor programs and evaluate whether they are delivering what they promised. Social auditing, community score cards, citizens score cards, and citizens satisfactory surveys are all social accountability tools.217 Social accountability 216 Ecorys Macro Group 2013, 45. 217 https://saeguide.worldbank.org/what-social-accountability https://saeguide.worldbank.org/what-social- accountability. 106 involves citizens’ oversight of government activities, including management of public finances and service delivery. Social audits may go beyond the oversight of program finances and procurement to examine all aspects of the program, including level of access to information, accountability, public involvement, program outputs, and outcomes. In Ghana, for instance, report cards are being used in the education sector to keep track of teacher absenteeism, a major problem that the GoG is trying to address. At the local government level, social accountability is in place for some social protection programs and is being strengthened. Social protection programs such as school feeding, LEAP, and LIPW have strong community participation. The establishment of complaint and redress mechanisms and strong communication with the community and beneficiaries are distinct characteristics of LEAP and LIPW. The World Bank-backed Local Government Capacity Support Project is also assisting the development of social accountability mechanisms at local levels. MLGRD has established a Social Accountability Unit, which is charged with the responsibility for implementing this component. The project is supporting the introduction of a social audit approach, which will stimulate civil society engagement with local governments on public financial management issues. This will be done through the formation and facilitation of community-level Social Public Expenditure and Financial Accountability (SPEFA) groups, which will receive ongoing practical training and assistance from civil society organizations in engaging effectively with MMDAs. The design of this social audit approach is informed by lessons learned from Ghana School Feeding Program. The project is also supporting the development of a media network (radio, print, television) on local government public financial management issues.218 4. Ongoing Reforms Several initiatives are under way that should help improve resource allocation. In 2009, the MoF launched the Ghana Integrated Financial Management Information System (GIFMIS) with the objective of improving automating government transactions.219 GIFMIS can streamline the accounting and reporting of social protection expenditure and includes several modules: accounts receivable, accounts payable, general ledger, purchasing, and cash management.220 As of early 2015, the general ledger and purchasing modules had been installed in all 23 ministries in Accra, covering only Consolidated Fund transitions. The next phases of the project involve including IGF, statutory funds, and operations funded by development partners, and rolling out the system to MMDAs. As part of the project, a new chart of accounts has been introduced and has been adopted by all MDAs and is progressively being adopted by MMDAs. The multitude of banking accounts of budget implementation units are being consolidated in a unified treasury account. MoF organization and planning tools are also being strengthened. In 2011, the MoF created a Public Investment Division, which is seeking to establish a policy framework to guide public investment decisions. In 2014, the MoF introduced program-based budgeting (PBB) to replace the existing line-item budgeting, with all MDAs moving to the program-based approach. The implementation of PBB should improve the linkages between the policy objectives as articulated in the GSGDA and the budget, and provide for a better budget monitoring. In 2016, the MMDAs 218 World Bank 2011b, 8. 219 World Bank 2011a, viii. 220 http://www.cagd.gov.gh/gifmis/index.php/faqs/16-what-are-the-electronic-financial-modules-being-introduced- under-gifmis. 107 will also begin preparing their budgets on a PBB basis. Nonetheless, the PBB-MTDF for 2015–17 prepared by the MDAs does not provide information on social assistance program spending because most programs or subprograms do not report the amount spent or budgeted, or subprogram are defined so broadly that specific social assistance programs cannot be identified. Similarly, GIFMIS could not provide with information on the social assistance programs being implemented. Recently, the GoG has stated that its expenditure rationalization priority measures include the following steps:221  Deepening the implementation of GIFMIS through completion of the budget modules and roll out to all MDAs and MMDAs  Establishing of the unified treasury account  Rationalization of the wage bill, pensions, gratuities, and social security payments  Rationalization of public sector staff in MDAs; elements include the ongoing payroll upgrade and installation of new Human Resource Management System in the Office of Head of Civil Service (OHCS) and Public Services Commission (PSC)  Restructuring DACF and other statutory funds to reduce rigidities in the budget, and aligning them to priority programs  Continuing the policy of regular adjustment of fuel and utility prices to keep expenditure on subsidies within budget constraints. In sum, there continues to be a large information gap on public expenditures outcomes. As mentioned in the 2011 Joint Review, “the regular and comprehensive review of public expenditure, from an effectiveness, efficiency, equity and sustainability perspective, is lacking in Ghana. Reports on expenditure outturns remain infrequent, and do not comprise functional classifications. Annual progress reports (APRs) from the National Development Planning Commission do not establish a link between expenditure and outcomes, beyond the alignment of expenditure to the successive Poverty Reduction Strategies broad priorities. At the sector level, most Ministries do not publish APRs, although all are required to do so.” 222 D. Implications for Managing Social Protection Expenditures The foregoing analysis of the allocation and management of public resources to social protection should provide some guidance for the preparation of the social protection policy. The main messages from this analysis follow. Ghana spends little on social protection, both in absolute terms and as a share of GDP, and most of the resources are not well targeted to the poor. A large share of spending goes to scholarships and subsidies for senior high school students, which are not well targeted to the poor. If Ghana wants to consolidate a strong social protection system, it needs to target existing programs better and increase the coverage of well-targeted programs. The formulation and execution of social protection expenditures suffers from the same problems as other sectors: unpredictability of budget allocations from the MoF and statutory 221 Government of Ghana 2014a, 8 and 9. 222 World Bank 2011a, viii. 108 funds; late disbursement of funds; incomplete disbursement of funds from the MoF; and unpredictability of disbursements from statutory funds. The fiscal adjustment program being implemented by the GoG may create fiscal space for well-targeted social protection programs. Increasing the coverage and generosity of well- targeted transfers program such as LEAP may help mitigate any potential adverse impact of fiscal adjustment on the poor. At the same time, there is a need to increase the cost-efficiency of social protection programs and better target them to the poor. Large program such as the Scholarship Secretariat programs should be better focused on the poor. The responsibilities of the MMDAs with the devolved functions, particularly with respect to welfare services, should be clearly defined. Social welfare and community development functions, among others, have been devolved to MMDAs, but the MoGCSP’s Department of Social Development still discharges some related functions, while the “center” still controls “hiring and firing” of MMDA staff and payment of salaries. Once the functions of MMDAs are clearly defined, there is a need to evaluate MMDAs’ requirements in terms of human resources and budget allocations to discharge these function in an effective way. Personnel decisions should be transferred to MMDAs. Intergovernmental financial transfers should be geared to providing the MMDAs with the resources to discharge their responsibilities effectively. DACF resources should be transferred directly to the MMDAs in their entirety to help finance the devolved functions and other local priorities. The introduction of the program-based budget and GIFMIS has not made it possible to identify the social assistance programs in sector budgets. In the 2015–17 budget, it is not possible to identify any of the expenditures on the social assistance programs. For instance, it is not possible to find out how much was spent or is going to be spent on LEAP or even on a much larger program, such as the School Feeding Program. Similarly, GIFMIS is not currently able to provide information about the budget or actual expenditures for social assistance programs. MoGCSP should work with the MoF to ensure that all social protection programs are explicitly presented as programs or subprograms in the budget, including its recurrent expenditures and benefits, as well as corresponding financing sources. This will make it possible to build a social protection budget to keep track of the amount of resources being allocated to the sector, their composition, and sources of financing. This will also facilitate monitoring of the execution of social protection expenditures. The recommendations of the Ghana Auditor General concerning social protection expenditures should be implemented promptly and the different tools of social auditing (report cards, public expenditure tracking surveys, and so on) should be deployed when considered cost effective. The social protection policy under preparation should clearly define social protection and the core social protection programs it comprises. The MoGCSP should coordinate budget formulation, execution, and evaluation of these programs with the MDAs responsible for their implementation. Program-based budgeting (PBB) being implemented by MoF should explicitly present these core social protection programs, and GIFMIS should include a functional code for social protection that will aid in appropriately tracking social protection spending. 109 Support should continue to be provided to carry out the financial modernization and financial decentralization framework with strong local participation. Among the priorities areas is training in financial management for leaders and staff at the MMDA level. In sum, Ghana spends a fraction of its Sub-Saharan African lower middle-income peers on social assistance; therefore key social assistance programs have small coverage and limited generosity and are to a large extent financed from external sources. Ghana is committed to continue reducing its large fiscal deficit while protecting expenditures in well-targeted social protection programs. While the process of public expenditure management in Ghana has improved in recent years, weaknesses remain in the areas of budget formulation, execution, and auditing. Public expenditure management is highly fragmented, which makes it difficult to analyze the allocation of public resources. The decentralization program devolved many responsibilities to the local government, including welfare services, but not the control over resources. Social protection program managers report three major issues: late disbursement of funds; incomplete disbursement of funds from the MoF; and unpredictability of disbursement from statutory funds. Several initiatives are under way that should help improve public expenditure management, including the establishment of the Ghana Integrated Financial Management Information System (GIFMIS), the introduction of program-based budgeting (PBB), and the fiscal decentralization action plan. 110 VI. Conclusions and Recommendations This final chapter summarizes the main findings of the study and offers some suggestions to further increase the impact, efficiency, and sustainability of the social protection system in Ghana. Ghana has developed an impressive social protection framework. It has a set of social assistance programs that address the principal risks facing Ghanaians; it is developing the tools to implement and monitor social assistance programs in an efficient way; it is developing a governance framework for the sector at the national, regional, and local levels; and the Ministry of Gender, Children and Social Protection (MoGCSP) is developing an overarching social protection policy and vision for the sector. But challenges remain to consolidate a social protection system consisting of a “coherent portfolio of programs that work together to deliver resilience, opportunity and equity” and to make the portfolio effective and sustainable.223 The analysis in this report indicates four challenges that need to be addressed: consolidating the substantial gains achieved to date and building a stronger portfolio of programs; strengthening management of expenditures; expanding well-targeted programs and rationalizing poorly targeted programs; and ensuring the sustainability of key programs. The following considerations are intended to support MoGCSP as it develops a strong policy and medium-term program for the sector. A. Consolidating Gains and Building a Stronger Program Portfolio To help consolidate gains and build a stronger social protection system, the social protection policy under development should set specific goals to be achieved and provide the elements for the development of a medium-term program and action plan to achieve the goals. Currently, there are no clear, galvanizing, rallying, but realistic goal(s) for the social protection system that can be monitored. The government of Ghana (GoG) should consider setting specific goal(s) to be achieved by the social protection system to guide the design of interventions and the allocation of resources for the sector. The goal(s) should be set with the participation of all key stakeholders. They could include, for example, goals like of the Millennium Development Goals (MDGs), such as the elimination of extreme poverty or cutting the poverty rate in half—which could help mobilize Ghanaians and development partners (DPs) to support the social protection programs along the lines of widespread backing of the MDGs. The elimination of extreme poverty in Ghana will require a minimum investment of 0.5 percent of GDP in a well-targeted transfer program. This estimate is based on Round 6 of the Ghana Living Standards Survey (GLSS-6) and the expenditures gap. This latter measures the distance between the average expenditure of the extreme poor to the extreme poverty line as a percentage of the extreme poverty line.224 The expenditure gap of the extreme poor is 27 percent, 223 The World Bank’s Social Protection and Labor Strategy 2012 –2022 (World Bank 2012b, xiv). 224 The expenditure (or consumption) gap should not be confused with the poverty gap. Both measure the average distance of the poor to the poverty (extreme poverty) line. The expenditure gap is averaged by the number of poor (extreme poor), while the poverty gap is averaged by the total population. Both are presented as a percentage of the poverty line. The expenditure gap is equal to the poverty (extreme poverty) gap divided by the poverty (extreme poverty) incidence. For Ghana, the expenditure gap for poverty is 0.077/0.241=0.32. For extreme poverty, it is 0.023/0.084=0.27. 111 or GH¢215. Therefore if the expenditure capacity of the extreme poor was increased by the estimated expenditure gap, the incidence of extreme poverty would be eliminated. A of GH¢215 to all the extreme poor in Ghana would cost GH¢479 million, which is equivalent to 0.5 percent of GDP as of 2013. It is important to note that this assumes that the transfer is perfectly targeted to the extreme poor households and the transfer is tailored to each household expenditure gap. A transfer of GH¢215 across the board to those in extreme poverty would still leave those with expenditure gaps below the average in extreme poverty. Only an across-the- board transfer of GH¢754 would ensure that all extreme poor would exit extreme poverty; this amount is equivalent to 1.8 percent of GDP. But in this case, this redistribution of resources would also reduce overall poverty by 20 percent and the poverty gap by 56 percent. Ghana spends 1.4 percent of GDP on social protection, a fraction of its lower middle-income peers in Sub-Saharan Africa. Only 0.9 percent of the budget is spent on social assistance programs. To have an impact on poverty and improve the lives of the most vulnerable groups, social protection spending must increase and reach the poor. Spending on scholarships and subsidies to senior high schools, which the analysis show that are poorly targeted, is almost commensurate to spending on social assistance and represent 0.8 of the budget and 0.2 of GDP (compared to social assistance which represents 0.9 percent of the budget and 0.3 percent of GDP). The social protection policy under preparation clearly defines social protection and the core social protection programs it comprises. These programs will be the focus of the MoGCSP’s coordination efforts. The MoGCSP should coordinate budget formulation, execution, and evaluation of these programs in cooperation with the Ministry of Finance (MoF) and the Ministries, Departments, and Agencies (MDAs) responsible for implementing the various social protection programs. While setting specific goal(s) should help galvanize support for social protection, the social protection policy should also seek to achieve broader objectives to consolidate an effective social protection system in Ghana. As mentioned, the broader objective of the social protection system is to minimize the impact of risks and vulnerabilities facing Ghanaians using a “coherent portfolio of programs that work together to deliver increased resilience, opportunity, and equity” and achieve the desirable objectives, particular for those most disenfranchised. Table 6.1 summarizes the strengths and issues with the existing portfolio of programs and presents recommendations to further strengthen it. Key administration tools to manage the social protection systems are being developed and should gradually be mainstreamed throughout the system. These tools included the single registry, a revised targeting mechanism (including a poverty map and proxy-means testing, PMT), the e-payment system, grievance and redress mechanisms, monitoring and evaluation, and management information systems (MIS) for the different programs. Efforts are also under way to strengthen social auditing of programs at different levels. The social protection medium-term program and action plan should be supported by a budget and monitoring and evaluation (M&E) framework. Once clear goals for the social protection have been established, policies and programs should be geared toward achieving those goals. The multiyear action plan should be costed out and annual budgets should be prepared. For the M&E framework, a few indicators should be selected to ensure that program inputs are being transformed in the desired outputs and that outputs are supporting the desirable 112 outcomes/goals. It is quite useful to have a logical framework for the social protection medium- term program. Information should then to be collected and processed to feed the monitoring indicators and make it possible to track whether the programs are achieving what it is expected. For a selected number of programs, process evaluations, targeting accuracy evaluations, public expenditure tracking surveys (PETS), or beneficiary satisfaction surveys may be conducted when considered cost effective. Similarly, as already contemplated, impact evaluations should be conducted for a selected number of programs, particularly those that can inform policy and program design. Table 6.1: Issues to Be Addressed in the Social Protection Portfolio Social Protection Function Strengths/Issues Recommendationsa Programs Main/Secondary Social Assistance Unconditional cash Equity Support to extreme poor elderly poor, OVC, Increase coverage and keep benefits transfer (LEAP) chronic ill, pregnant women. levels in real terms (short term, ST). Issues: Low coverage, low generosity to Target all families in extreme poverty ensure adequate minimum security. (medium term, MT). In-kind transfers Equity/opportunity Helps increase and maintain enrollments. Keep expenditures under control and (school feeding) Issues: Payment delays to providers. payments up to date (ST). Cash and in-kind Equity/opportunity Help keep students in class. Target uniform and textbooks to poor and transfer (capitation, Issues: Leakages, payment arrears, lack of vulnerable children (ST). uniforms, and targeting. textbooks) Waivers/NHIS Equity/opportunity Free access to health care for the poor. Target exemptions only for the poor and exemption Issues. Low coverage of the poor, low rate of vulnerable (MT). Establish a mechanism renewals, lack of targeting. to ensure automatic renewals of LEAP beneficiaries (ST). Public works/LIPW Equity/resilience/ Insure against drop in income, involve Continue to expand the program to the opportunity community in selection of workers. most food vulnerable districts (ST) Issues: Low coverage; oversubscription in Keep compensation at a level to ensure work sites; working conditions of women with some degree of self-selection (ST). children; fully financed from external sources. Ensure appropriate working conditions of women with children (ST). Link with skills/income- producing opportunities (pilot underway) (MT). Start allocating local resources to the program (ST). Scholarships/Subsidi Opportunity/equity Promote human capital accumulation. SS should publish and disseminate es to SHS Issues: poor selection mechanisms, lack of regulations of different programs (ST). targeting, payment arrears, lack of monitoring. Target awards to the poor (ST). Potential duplication and overlapping of Monitor beneficiaries of tertiary existing programs. scholarships, ensure relevance for the country of tertiary scholarships, and revise bonding arrangements (MT). Rationalize all scholarships programs (MT). Social Insurance Resilience Insure against drop in income during old age Continue to increase social security Issues: Low coverage, future benefits too coverage and adopt reforms to make it generous, not sustainable. sustainable (MT). Active Labor Market Opportunity Enhance youth employability and workers’ Establish labor market information Programs skills. system, job search assistance, career Issues: Lack of labor market information counseling, evaluate youth programs, and services, little attention to basic skill formation consolidate youth programs (MT). and second chance education programs, overlapping and duplication of youth programs, no impact assessment of programs. Note: LIPW=Labor Intensive Public Works program; NHIS=National Health Insurance Scheme; OVC=orphans and vulnerable children; SS=Scholarship Secretariat. 113 a. Timeline of implementation: ST=short term (within one year); MT = medium term (within three to five years). As MoGCSP implements the recently approved governance framework, it is important to stress the need to strengthen the District Social Development Departments (DSDD) substantially, as they are in the front line of delivering social welfare services. The recently approved governance framework for social protection at the national, regional, district, and community levels is filling an important gap. Institutional arrangements and coordinating mechanism must be strengthened, staff must be trained and motivated, and better coordination should be established among Ministries, Departments, and Agencies (MDAs); Municipal, Metropolitan, and District Assemblies (MMDAs); and services organizations. As the governance framework is implemented, there is a need to support the work of the district social workers (DSDOs) through training and better working conditions. The DSDOs must become proactive in monitoring vulnerable groups in the community, identifying any family issues in a timely manner, and linking families to existing programs and services. Closer links between social services (referrals, integrated case management, and sensitization) and health, nutrition and education programs, as well as closer links to skill development programs, could help address some of the existing family challenges and provide opportunities for beneficiaries to “leap” from poverty. Protocols and memoranda of understanding to promote synergy between programs should be established and implemented, supported by appropriate training of staff. The single registry will help identify the needs of particular households and the potential beneficiaries of existing programs. To better use this tool to increase the synergy between programs, social workers will need to be trained to identify problems within the households and to address them or refer them to other services and/or institutions. Referral protocols and cooperation agreements between institutions need to be established. A training program for the DSDOs also needs to be developed and implemented. B. Strengthening Management of Expenditures While public expenditure management in Ghana has improved in recent years, major challenges remain in the areas of budget formulation, execution, and auditing. Public expenditure management is highly fragmented, which makes it difficult to analyze the allocation of public resources. The central government budget covers only the Consolidated Fund and the transfers from this fund to the earmarked funds. The financial statements that show the executed budget prepared by the Controller and Accountant General include only the accounts of the Consolidated Fund. Similarly, the Ghana Auditor General (GAG) report on the Accounts of the Public Sector covers only the Consolidated Fund. There is no consolidated account of the public sector or of the general government. And there are no readily available actual accounts (budget outturns) of sector spending and financing. The well-established practice to be conservative about revenue projections and realistic about expenditures projections should be adhered to in preparing the budget, while earmarking revenues should be avoided. Revenue projections have been quite off target in recent years; this complicates budget preparation and budget execution. While a Medium-Term Expenditure Framework (MTEF) has been formally in place since early 2000, it has not contributed much to further fiscal discipline. Budget formulation is complicated by the existence of five major statutory funds that received earmarked funds. Earmarking treasury funds reduces 114 the flexibility of fiscal management and makes program-based budgeting less effective. To the extent possible, earmarking should be avoided. The MoGCSP, in collaboration with the Ministry of Finance (MoF), should spearhead the development of a social protection “portfolio” of interventions to be prioritized in the budget allocation process. This will require: (i) establishing the coordinating procedures, determining priorities, consolidating information, and providing the information to MoF; (ii) building capacity in MoF and training MoF staff on social protection; (iii) sensitizing ministries to report back on actual spending; and (iv) building capacity on data recording and reporting at different levels to ensure that institutions know how to capture social protection information and report it to be included in the national budget. All major social assistance programs should be shown explicitly in the budget. The program- based budgeting being implemented by the Ministry of Finance represents a first important step for coordinating the financing and accountability of social protection programs. However, the introduction of the PBB has not made it possible to identify the major social assistance programs in the budget nor can the Ghana Integrated Financial Management Information System (GIFMIS) currently provide program information. MoGCSP should work with the MoF to ensure that all social assistance programs are explicitly presented in the budget, including their recurrent expenditures, transfer or payments, and corresponding financing sources. This will make it possible to build a social protection budget to keep track of the amount of resources being allocated to the sector, their composition, and sources of financing. It will also facilitate coordination, monitoring and auditing of sector expenditures. The program-based budgeting (PBB) being implemented by MoF should explicitly present the core social protection programs. GIFMIS should include a functional code for social protection that will aid in appropriately tracking social protection spending. Ongoing reforms to improve public sector financial management should improve the execution of social protection programs. The implementation of social protection programs suffers from the same general weakness of overall budget implementation: late disbursement of funds; incomplete disbursement of funds from the MoF; and unpredictability of disbursement from statutory funds. The implementation of GIFMIS, the adoption of the new harmonized chart of accounts, and the establishment of a unified treasury account should help with budget execution. Budget sector outturns should be prepared, published, and evaluated annually. The adoption of program-based budgeting (PBB) should help the reporting and evaluation of public expenditure outcomes. This should include regular public reports on MDA budget outturns, using economic and functional classifications, and covering the different sources of financing (Consolidated Fund, statutory funds, internally generated funds, funds from development partners). It should also include reports on actual expenditures of MMDAs and sources of finance. The Audit Report Implementation Committees (ARIC) in MDAs and MMDAs should be made responsible for ensuring that the recommendations of Ghana Auditor General (GAG) are implemented in the institutions. Recommendations to correct the problems identified in the audits are included in GAG’s reports submitted to the Parliament. However, there is no follow up on the report’s recommendations among the MDAs and MMDAs, and no accountability; the same problems keep repeating year after year. The Audit Service Act of 2000 established ARICs in 115 MDAs and MMDAs to implement the GAG’s recommendations. However, these committees have not been functioning appropriately. The GAG is issuing new guidelines to address this problem. The functional responsibilities of different levels of government should be better defined and intergovernmental financial transfers should be geared toward providing MMDAs with the resources to discharge their responsibilities effectively. Social welfare and community development functions have been devolved to MMDAs, but the MoGCSP’s operational departments (such as the Department of Social Development and the Department of Social Development) still discharge some related functions. At the same time, the “center” still controls “hiring and firing” of MMDA staff and payment of salaries. New legislation (L1971) has been reviewed to support the retooling of MMDAs’ functions; there is now a need to expedite efforts to provide MMDAs with the human and budget resources to discharge these functions in an effective way. Personnel decisions should be transferred to MMDAs. District Assembly Consolidated Fund resources should be transferred directly to the District Assemblies (DAs). Support should continue to be provided to implement the financial decentralization framework with strong local participation. Financial management training program for leaders and staff at the MMDA should be considered as a priority. C. Expanding Well-Targeted Programs and Rationalizing Poorly Targeted Programs 1. Expanding LEAP and LIPW Poor parents continue to have difficulties accessing basic services and sending their children to school. Despite government efforts to ease demand constraints to access education and other basic services, poor parents still need to spend an important share of their meager budgets to access these services when they are available (see Chapter II). There is thus a need to continue to support programs such as the Livelihood Empowerment Against Poverty Program (LEAP), which provides cash transfers to the poorest families, with the adequate reach and generosity, to further help parents overcome some of the constrain they face. The expansion of LEAP to cover all extreme poor families appears feasible within the confines of the ongoing fiscal consolidation. As mentioned, the GoG is committed to using part of the resulting fiscal space created by the ongoing fiscal consolidation to expand well-targeted programs such as LEAP. Microsimulation projections indicate that the gradual expansion of LEAP to cover all extreme poor families within the next five years would cost 0.12 percent of GDP in 2019. Table 6.2 presents the projections, which consider the current annual average transfer to families of GH¢432 adjusted for inflation (by the previously year rate), as well as 20 percent overhead. By the 2019, the transfer would cover all extreme poor households, as estimated in the GLSS-6. The projected LEAP expenditures are less than one-half of 1 percentage point of the projected public expenditures for the next three years. Table 6.2: Simulated Cost of LEAP Expansion 2014 2015 2016 2017 2018 2019 Number of households 88,932 150,785 200,000 250,000 302,000 357,822 New households 61,853 49,215 50,000 52,000 55,822 Average annual transfer/household GH¢ (2013 prices) 432 -- -- -- -- -- Annual average inflation (%) 15.5 12.2 10.2 8.4 5.0 5.0 116 Average annual transfer/household (GH¢) 432 499 560 617 669 702 Annual cost of transfers (GH¢ million) 38 75 112 154 202 251 Overhead (20%) 8 15 22 31 40 50 Total cost of transfers (GH¢ million) 46 90 134 185 242 302 Total cost of transfers (% of GDP) 0.04 0.07 0.09 0.10 0.11 0.12 Total cost of transfers (% of projected expenditures) 0.1 0.3 0.4 0.4 Memo: Projected expenditures (% of GDP) 27.8 26.7 25.4 23.7 -- -- GDP (GH¢ million) 113,436 133,344 155,570 184,952 212,066 243,155 Source: World Bank staff estimates. Data on GDP, expenditures, and inflation (2014–17) from IMF. The increase in LEAP coverage will have a visible impact on extreme poverty. The microsimulation using the GLSS-6 in Table 6.3 indicates that the incidence of extreme poverty would decline by 22 percent and the extreme poverty gap would decline by 28 percent from their 2013 levels, if the current transfer of GH¢432 was given to all extreme poor families (357,822 families).225 Table 6.3: Simulated Impact on Poverty Indicators of an Expended LEAP Extreme poverty Poverty P0 P1 P0 P1 Baseline (2013) 8.424 2.291 24.230 7.759 LEAP transfers perfectly targeted 6.553 1.630 24.148 7.284 Percentage points change 1.82 0.644 0 0.442 Percent change -21.74 -28.32 0.00 -5.72 Source: World Bank staff estimates based on GLSS-6. Note: Simulated impact on poverty of expanding LEAP to all families in extreme poverty with a transfer of GH¢432 using GLSS-6. P0= poverty headcount; P1= poverty gap. Consideration should also be given to expanding LIPW as resources permit. As mentioned, the Labor-Intensive Public Works program (LIPW) has shown positive results and serves as an important safety net function in areas where seasonal unemployment is a major problem. 2. Promoting graduation pathways and productive inclusion of the poor and vulnerable More efforts and resources need to be allocated to promote the inclusion of the poor and vulnerable in productive activities as a medium-term strategy to lift them out of poverty. Although there is evidence226 that recipients of social assistance benefits invest in livelihood activities, graduation from SP programs typically require complementary support such as productive assets, grants, training, and financial inclusion. There is a need to promote sustainable livelihood of social assistance beneficiaries who have the capacity for labor and the potential to graduate as a medium-term strategy to lift them out of poverty and reduce dependence on public transfers. The pilot—Complementary Income- Generating Skills—under preparation will provide 225 This of course assumes that the transfer would be perfectly targeted to all households in extreme poverty. The simulation evaluates the impact on poverty of transferring GH¢69.3 to 2,228,720 extreme poor persons (as the average extreme poor household has 6.2 members, and the average transfer to beneficiary households is GH¢430). 226 Protection to Promotion project, FAO, UNICEF, DFID. 117 LIPW and LEAP beneficiaries with skills to pursue income-generating activities, and is a welcome initiative, mostly in rural areas, that could help them eventually graduate from the program. More generally, the government needs to invest in promoting the productive inclusion of the poor and low-productive workers— not only social assistance beneficiaries—to more productive jobs in urban areas as well. 3. Rationalizing Scholarships The Scholarship Secretariat senior high school (SHS) program should be restructured and placed on an equitable footing. The potential inequity of scholarships has already been flagged in the Development Education Plan 2010–20. Indeed, the Ministry of Education (MOE) indicated in the Plan that most of the students attending public senior secondary schools “actually come from private schools, so they are not poor, and should not receive boarding subsidies. However it is important to subsidize poor students, especially those from northern and other hard-to-reach areas.”227 The Plan’s recommendations continue to be valid and should be implemented:  Rationalizing some boarding facilities, except for those from disadvantaged areas and for girls  Increasing means testing and cost-recovery from parents and guardians (especially those who can afford to pay)  Providing grants and/or loans to poor students;  Providing free transport for day-students living within 20 km of schools, where feasible  Expanding student self-help schemes (cooking and cleaning).228 More generally, the GoG should address the overlaps and poor targeting of scholarships and student subsidies to SHS. The recently released poverty maps229 should be used for further analysis to improve the targeting of secondary scholarships. Scholarships at the tertiary level should be based on the requirements of the economy to fill existing gaps in skills. These scholarships should be awarded to support a national human resources development policy. The Ghana Shared Growth and Development Agenda (GSGDA) II (2014–2017) recognizes the need to develop such policy.230 All beneficiaries of the scholarships should be required to enter into effective bond agreements that ensure that the country is able to recover the investment made. Student loans with appropriate guarantees could replace scholarships for most students. 4. Rationalizing Active Labor Market Programs Effective labor intermediation services in Ghana should be created. These services include (i) labor market information; (ii) job search assistance; (iii) career counseling; and (iv) special programs for the unemployed, including second chance programs. According to the 2014 National 227 Ministry of Education 2010, 30. 228 Ministry of Education 2010, 30. 229 GSS 2015. 230 NDPC 2014, 124. 118 Employment Policy, “an effective Labor Market Information System (LMIS) will depend on the improved capacity of the Employment Information Unit (EIU) of the Labor Department of this Ministry (MELR), the Ghana Statistical Service (GSS), and other labor organizations. Funding and human resource capacities currently do not exist to support the development of an efficient and effective LMI system.”231 The World Bank and the International Labor Organization (ILO) plan to assist the Department of Labor in setting up a modern labor market information system that can meet the needs of Ghanaian firms and job seekers. Job search assistance and career counseling services should be built up gradually. There is significant overlap and duplication among youth programs, and these programs should be consolidated and rationalized. It is beyond the scope of this report to present a detailed analysis of youth programs. Nonetheless, since apprenticeship and entrepreneurship programs are usually part of the portfolio of programs of the social protection and labor systems, a few considerations are warranted here. Best practice indicates that the design of apprenticeship and entrepreneurship programs have common elements. First, growth sectors and potential jobs must be identified. Second, there is a need to : (i) assess beneficiaries’ capacities and to design an individual development and training plan with an eye on potential jobs; (ii) train beneficiaries in basic skills (life skills such as discipline, creativity, and team work); and (iii) train beneficiaries in professional skills (in such areas as tourism, construction, and business management). Third, the beneficiary must be placed in an apprenticeship position or be supported in the development of an entrepreneurial activity. Apprenticeships must be monitored, and the beneficiary must be supported in obtaining a job or be provided with technical and financial assistance, in the case of business initiatives. Some apprentices may decide to establish a microenterprise or become self- employed. Finally, financing by results may be used when considered appropriate, and programs must be evaluated for their impact. One of the most successful experiences in apprenticeship programs are the Jovenes programs first implemented in Latin America. These programs combine vocational training courses with work experience in private sector firms. Private sector participation in the program is essential and is sometimes coordinated by the local Chambers of Commerce and Industry. The programs usually provide a per diem payment for food, transport, and extras; cover training costs; and provide work risk insurance (during the apprenticeship). The selection of institutions implementing the program (nongovernmental organizations, training institutions, and so on) is competitive by public bidding and based on their results (measured in terms of the rates that beneficiaries are placed in apprenticeships in participating businesses). Ghana should consider such a model. Despite the multitude of youth programs, it is not clear what works in Ghana because programs have not been evaluated for their impact. One exception is the National Apprentice Program, which is currently being evaluated. The rationalization of programs should take into consideration the existing large informal apprenticeship system that is a specific feature of the Ghanaian economy. Some argue that to strengthen the informal apprenticeships system, there is a need to address the following challenges: (i) eliminate its gender bias; (ii) make it more pro-poor; (iii) modernize its technologies; and (iv) train and improve the standards of the trainers (master crafts-persons).232 231 MELR 2014, 14. 232 Schwegler-Rohmeis and others 2014, 50 and 55. 119 However, youth apprenticeship and entrepreneurship programs should not be geared toward providing youth with low-productivity jobs in the informal sector. In Ghana, it appears that most youth are placed in youth programs without the required evaluation and appropriate development plan. Most youth programs appear geared toward the creation of microentrepreneurs, without the proper evaluation of the capabilities of individual young people or market needs. Qualitative evidence shows that in most cities, one beauty salon after another is open for business, but most lack clients, and all make very little money, if any. To support the GoG to understand how better provide opportunities for youth, the World Bank is preparing a study on a youth employment programs in Ghana. D. Ensuring Sustainability of Existing Programs 1. Social Assistance and Scholarships Programs Better targeting, increased cost-efficiency, and reduced dependence on external financing would help make social assistance and scholarship programs more sustainable. While the GoG will continue to require the support of development partners during the fiscal consolidation period to finance the operations and expansion of key social assistance programs, provisions must be made for the future reduction in their external dependence of these programs. The amount of resources the country is allocating to social protection is a fraction of its peers. The creation of a legal basis for key social protection programs may also contribute to their sustainability. The GoG should start allocating budget resources to finance program. As discussed, LIPW is currently fully funded by external sources. To ensure its sustainability over the longer term, external resources should be replaced gradually by domestic funding. Continuing efforts to reduce poorly targeted energy subsidies should help create fiscal space to strengthen social protection programs. As discussed, spending on electricity and fuel subsidies is five times the amount of public resources spent for social assistance, and the amount spent on poorly targeted scholarships and SHS subsidies is almost the same as the amount spent on social assistance. Reducing these energy subsidies will make it possible to reallocate some of these resources to well-targeted social assistance programs, including better targeted scholarships. 2. National Health Insurance A strategy and action plan to implement the reforms required to ensure the sustainability of the National Health Insurance Scheme (NHIS) should be developed. Adverse selection, lack of restrictions on demand, lack of copayments, and cost escalation are all characteristics of the current system that make it financially unsustainable. Arrear payments to providers impairs service delivery and discourages new enrollments and membership renewals. These problems and reform options have been studied233 and discussed at length during the ten-year celebrations of the NHIS, whose theme was “Towards Universal Health Coverage: Increasing Enrolment whilst Ensuring Sustainability.”234 Among these reform options are: containing costs (such as capitation), medicine lists, and prescription levels; raising additional funding (such as a levy on tobacco and alcoholic beverages); intensifying clinical audits; scaling up the issuance of instant identification cards; increasing coverage of the poor; improving computerization of operations; and shortening claims 233 See, for example, Saleh (2013) and Schieber and others (2012). 234 National Health Insurance Scheme 2013. 120 processing and payment time. 235 Recent studies have also clearly presented other additional reforms options, including redefining the benefit package (core services package and other services) and introducing copayments for the nonpoor and noncore services.236 3. Social Insurance The Social Security National Insurance Trust (SSNIT) requires additional reform to ensure its longer-term sustainability. A World Bank evaluation of the current system suggests that the following reforms should be considered:237  Establish a sustainable, linear accrual rate. Replacing the current nonlinear SSNIT Tier I accrual rate with a linear accrual rate of 1.5/year (the current rate is 3.33 percent for the first 15 years and then 1.125 percent), along with other parametric reform measures, would support the sustainability of the SSNIT Tier I scheme, improve the incentives for workers to contribute over their entire work life, and still result in an adequate pension for a full- term worker.  Increase the age for eligibility for normal retirement benefits. Gradually eliminating reduced early retirement benefits at age 55 and increasing the normal retirement age from 60 to 65 will improve the sustainability of SSNIT and increase the incentives for individuals to work longer, as life expectancy at retirement age increases. Introducing actuarially fair penalties for early retirement and actuarially fair supplements for retirement after the retirement age will enable the scheme to have the flexibility to accommodate individual circumstances.  Introduce automatic indexation to adjust pension benefits to wage inflation.  Extend the reference period for benefits. Gradually increasing the reference period for benefit determination from the best three years to ten years prior to retirement and adjustment or “valorization” of the reference wage base to determine benefits can reduce that inequality and disparity in replacement rates between cohorts with comparable work histories. Eventually the reference period should be extended to the lifetime work history, provided there is supporting data to accomplish this. 235 PowerPoint presentation by NHIS Chief Executive, http://www.nhis.gov.gh/conference.aspx. 236 Bitran 2014. 237 World Bank 2013a, vii. 121 Glossary Active labor market programs. Programs aimed at increasing the skills, employment, and long-term earning potential of participants, particularly unemployed youth, through training, apprenticeships, job search assistance, subsidized job placements, and the like. Cash transfer programs. Programs that transfer cash to eligible people or households. Common variants include child allowances, social pensions, needs-based transfers, and conditional cash transfers. Categorical targeting. A targeting method in which all individuals in a specific category (for example, a particular age group, geographic location, gender, or demographic composition) are eligible to receive benefits. Conditional cash transfers. Provide money to poor families contingent on them making investments in human capital, such as keeping their children in school or taking them to health centers on a regular basis. Equity. In the context of a social protection strategy, equity means reducing poverty and destitution and promoting equality of opportunity. Key interventions include social assistance programs (also known as safety net programs—including cash transfers and in-kind transfers, such as school feeding and targeted food assistance), alleviating chronic poverty, and protecting against destitution. Error of exclusion. The exclusion of a person who meets eligibility criteria from a program. Error of inclusion. The inclusion of an ineligible person in a program. Expenditures gap. Measures the distance between the average expenditure of the extreme poor to the extreme poverty line as a percentage of the extreme poverty line. The expenditure (or consumption) gap should not be confused with the poverty gap. Family allowances. Cash transfer for families with children. They can take various forms, such as means- tested child benefits, birth grants, or universal transfers for all children under a fixed age. Generosity. The level of a program benefit as a share of the poverty line or other type of indicator, such as the minimum wage, the average wage, or the total consumption of beneficiary households. Geographic targeting. A targeting method in which location determines eligibility for benefits or allocates budget to concentrate resources on poorer areas. In-kind food transfers. Provide additional resources to households by making food available when they need it most, in the form of food rations, supplementary and school feeding programs, or emergency food distribution. Last resort programs. Needs-based, usually means-tested programs designed to help those who are not assisted, or not assisted enough to prevent poverty, by social insurance (pensions, unemployment insurance) or universal programs (child allowances, education, and the like). Leakage. In discussions of targeting, the leakage rate is the proportion of those who are reached by the program who are classified as nonpoor (errors of inclusion). In discussions of accountability, the term is 122 often used more broadly to include funds that, through various forms of negligence or malfeasance, are diverted from legitimate beneficiaries to other uses. Management information system (MIS). Includes all the databases kept by the various program units in the performance of their functions—registry of beneficiaries, payments, and so on. Means test. A targeting method based on income that seeks to collect comprehensive information on household income and/or wealth and verifies the information collected against independent sources. Noncontributory pensions (or social pensions). Benefits paid to the elderly from sources finances through taxes (rather than contributions) and without regard to past participation in the labor market. Opportunity. In the context of social protection strategy, opportunity means building human capital, assets, and access to jobs and freeing families to make productive investments because of their greater sense of security. Key intervention include promoting better health, nutrition, education, and skills development, along with helping men and women access productive work. Institutions that promote opportunity are often integrated with those supporting resilience and equity. Orphans and vulnerable children. Orphans (children who have lost one or both parents) and other groups of children who are more exposed to risks than their peers, such as children with HIV/AIDS and those with sick caregivers, street children, children in institutions, child soldiers, child prostitutes, and others who are not cared for in a family setting or who are involved in the worst forms of child labor. Poverty gap. The mean difference between the poverty line and household income divided by the poverty line calculated over the entire population. The nonpoor have a poverty gap of zero. The poverty gap should not be confused with the expenditure gap. Poverty lines. Cutoff points separating the poor from the nonpoor. They can be monetary (for example, a certain level of consumption) or nonmonetary (for instance, a certain level of literacy). The use of multiple lines can help in distinguishing among different levels of poverty. For monetary measures, the poverty lines are often based on estimates of the cost of basic food needs: that is, the cost of a nutritional basket considered minimal for the health of a typical family (extreme poverty line), to which a provision is added for nonfood needs (upper poverty line). Proxy means test. A targeting method by which a score for applicant households is generated based on fairly easy-to-observe household characteristics, such as the location and quality of the household’s dwelling, ownership of durable goods, demographic structure, education, and so on. Public works programs (or workfare). Where income support for the poor is given in the form of wages (in either cash or food) in exchange for work effort. These programs typically provide short-term employment at low wages for unskilled and semiskilled workers on labor-intensive projects such as road construction and maintenance, irrigation infrastructure, reforestation, and soil conservation. Generally seen as a means of providing income support to the poor in critical times rather than as a way of getting the unemployed back into the labor market. Quintile. One-fifth of an ordered population; for example, the poorest or richest one-tenth of the population. Replacement rate. The value of the pension entitlement relative to individual earnings, usually calculated for a worker with average earnings. 123 Resilience. In the context of social protection strategy, resilience means helping people insure against drops in well-being from different types of shocks. Key sources of resilience are social insurance programs that minimize the negative impact of economic shocks on individuals and families—such as unemployment and disability insurance, old-age pensions, and scalable public works programs. Risk management strategies. Strategies introduced by individuals, households, or communities dealing with risks that may temporarily or permanently affect their well-being. Ex ante strategies look to avoid the risk’s occurrence (prevention strategies) or, if this is not possible, to reduce its impact (mitigation strategies). Ex post strategies are aimed at dealing with the shock once it occurs (coping strategies). Safety nets. Noncontributory transfer programs targeted in some manner to the poor and those vulnerable to poverty and shocks. Analogous to the U.S. term “welfare” and the European term “social assistance.” Safety net system. A collection of programs, ideally well-designed and well-implemented, complementing one another as well as complementing other public or social policies. School feeding programs. In-kind food transfers that provide meals or snacks for children at school to encourage their enrollment and improve their nutritional status and ability to pay attention in class. Social accountability. An approach to governance that involves citizens and civil society organizations in public decision making. Social accountability interventions can enable citizens and civil society actors to articulate their needs to governments and service providers. Social accountability also brings the perspective of citizens and civil society organizations to government activities, such as policy making, the management of public finances and resources, and service delivery. Finally, social accountability allows civil society to participate in monitoring the public sector and giving feedback on government performance. Social assistance (or social safety nets) programs. Noncontributory transfer programs targeted in some manner to the poor and those vulnerable to poverty and shocks. Social audit. A monitoring process through which project information is collected, analyzed, and shared publicly in a participatory fashion. Social audits may go beyond the oversight of project finances and procurements to examine all aspects of the project, including level of access to information, accountability, public involvement, project outputs, and outcomes. Social audits are typically carried out by community volunteers (social audit teams/committees) and findings are presented at a public forum/hearing. Social insurance. Contributory programs designed to help households insure themselves against sudden reductions in income. Types of social insurance include publicly provided or mandated insurance against unemployment, old age (pensions), disability, the death of the main provider, and sickness. Social pensions. Noncontributory pensions. Social protection. The set of public interventions aimed at supporting the poorer and more vulnerable members of society, as well as helping individuals, families, and communities manage risk. Social protection includes safety nets (social assistance), social insurance, labor market policies, social funds, and social services. Target group (or target population). The intended beneficiaries of program benefits. Targeting. The effort to focus resources among those most in need of them. 124 Targeting assessment (or benefit incidence analysis). Describes how public spending is distributed across population groups, whether defined as deciles, poor versus nonpoor, geographic areas, ethnic groups, and so on. Targeting errors. Errors that occur when program officials or the targeting rules they use may mistakenly identify nonpoor people as poor, or poor people as nonpoor. When the former are admitted to a program, it is an error of inclusion; when the latter are denied access to the program, it is an error of exclusion. Targeting method. Approach taken to identify the target group and thus determine eligibility for program benefits. Vulnerable groups. Typically including the elderly, orphans, widows, people with disabilities, people with HIV/AIDS, refugees or internally displaced persons, among others. Vulnerable groups face special difficulties in supporting themselves because of some particular aspect of their situation. Source: This glossary is based partly on World Bank (2013c, 2014e, 2015a) and Grosh and others (2008). 125 Appendix A. Organogram of the Ministry of Gender, Children, and Social Protection (MoGCSP) DV SECRETARIAT HON. MINISTER MINISTERIAL ADVISORY HT SECRETARIAT Deputy Ministers BOARD COUNCIL FOR DISABILITY DEPT OF GENDER INTERNAL AUDIT PUBLIC AFFAIRS CHIEF DEPT OF CHILDREN DIRECTOR DEPT OF SOCIAL DEVT. LEGAL UNIT PPBME RSIM SOCIAL HUMAN ADMINISTRATION FINANCE PROTECTION RESOURCE POLICY RESE2012bAR PROGRAMMES SECTOR HR PERSONNEL COORDINATION CH & DEVELOPMENT PLANNING ACCOUNTS STATISTICS & RECORDS MGT PLANNING COORDINATION DOC/DATA SECTOR TREASURY DEVELOPMENT TRANSPORT CENTER/ LIBRARY INSPECTORATE &TRAINING BUDGETING & REGULATION ESTATES RESOURCE MOBILISATION IT/IM SECTOR SECURITY M&E PERFORMANCE MANAGEMENT PROCUREMENT Note: DV= domestic violence; HT= human trafficking; IT = information technology; M&E = monitoring and evaluation STORES 126 Appendix B. Supporting Tables TableB.1: Central Government Accounts (GH¢ million) 2008 2009 2010 2011 2012 2013 2014 2015 Rev. Budget Actual Budget Actual Budget Actual Budget Actual Budget Actual Budget Actual Budget proj. Budget Total revenues and grants 5,616 5,622 7,474 7,215 9,628 8,829 10,601 12,825 15,614 16,668 22,533 20,800 26,057 24,739 32,406 Revenue 4,763 4,802 6,172 5,998 8,264 7,657 9,299 11,836 14,458 15,508 21,275 19,542 24,926 23,937 30,855 Grants 853 820 1,302 1,217 1,364 1,172 1,302 989 1,156 1,160 1,258 1,258 1,131 802 1,551 Project grants 514 401 683 683 833 650 784 450 690 512 454 454 805 775 1,018 Program grants 339 419 619 534 531 522 518 539 466 648 804 804 325 27 533 Total expenditures 6,430 7,894 9,692 8,752 10,900 11,550 12,670 13,379 17,513 20,944 28,163 27,462 31,803 32,368 39,132 Employee compensation 1,877 2,348 3,060 2,265 3,703 3,620 4,472 5,304 6,020 7,178 9,004 9,479 10,597 11,160 12,313 Use of goods and services 505 647 415 648 635 962 604 724 967 1,322 1,742 1,449 1,550 1,085 1,970 Interest payments 466 678 824 1,067 1,346 1,439 1,831 1,611 1,883 2,436 3,194 4,397 6,178 7,845 9,557 Subsidies 0 20 0 0 0 131 27 0 0 808 1,022 1,158 50 619 50 Transfers to other 1,063 1,124 1,509 1,290 1,221 1,037 1,295 1,101 1,825 3,765 6,209 4,547 6,498 5,243 7,409 government units Social benefits (Lifeline) 10 4 19 -- 44 0 29 0 31 0 39 2 49 49 60 Other expenditures 636 969 853 816 796 876 666 964 1,089 1,851 1,798 1,639 864 897 816 Capital expenditures 1,873 2,104 3,012 2,666 3,155 3,485 3,746 3,675 5,698 3,584 5,155 4,791 6,017 5,472 6,957 Domestic-financed 746 1,003 305 991 1,315 1,136 1,587 1,963 2,666 1,049 1,304 1,646 1,491 1,242 2,557 Foreign-financed 1,127 1,101 2,707 1,675 1,840 2,349 2,159 1,712 3,032 2,535 3,851 3,145 4,526 4,230 4,399 Overall balance -814 -2,272 -2,218 -1,537 -1,272 -2,721 -2,069 -554 -1,899 -4,276 -5,630 -6,662 -5,747 -7,629 -6,725 (commitment) Arrears and tax refunds -42 -142 -533 -673 -673 -687 -268 -2,186 -1,470 -3,830 -2,380 -1,922 -3,224 -3,301 -2,091 Overall balance (cash) -856 -2,414 -2,751 -2,210 -1,945 -3,408 -2,337 -2,740 -3,369 -8,106 -8,010 -8,584 -8,971 -10,930 -8,816 GDP 30,179 30,179 36,598 36,598 47,554 47,554 61,274 61,274 74,959 74,959 93,867 93,867 114,654 114,654 135,011 GDP non-oil 30,179 30,179 36,598 36,598 47,554 47,554 57,528 57,528 69,310 69,310 86,532 86,532 104,726 104,726 123,099 Cash deficit (non-oil) -2.8 -8.0 -7.5 -6.0 -4.1 -7.2 -4.1 -4.8 -4.9 -11.7 -9.3 -9.9 -8.6 -10.4 -7.2 Source: MoF budget statements. Note: Rev. proj. = revised projection; 127 Table B.2: Central Government Budgeted and Actual Spending by Sector (GH¢ million) 2008 2009 2010 2011 2012 2013 2014 2015 Budge Actu Budget Actual Budget Actual Budget Actual Budget Actual Budget Actual Budget Actual Budget t al Social 2,042 2,748 2,775 2,407 3,291 3,643 4,805 8,135 9,385 9,984 Ministry of Education 1,264 1,693 1,729 1,826 1,983 2,656 2,871 4,413 5,816 6,740 Ministry of Employment 38 61 39 44 and Labor Relations Ministry of Youth and 91 44 66 46 142 54 54 36 36 Sports National Commission for 15 23 9 18 27 29 Civic Education Ministry of Chieftaincy and 12 17 16 19 20 20 Traditional Affairs Ministry of Health 752 922 727 464 988 771 1,799 3,529 3,354 3,068 Ministry of Gender, 16 39 91 44 Children and Social Protection National Labor 2 2 2 3 Commission Ministry of Employment 21 33 33 20 35 28 and Social Welfare Ministry of Women and 5 9 9 3 13 5 Children Affairs Ministry of Local 233 226 Government and Rural Development Administration 789 1,148 745 635 707 1,107 2,009 2,552 1,500 2,101 Economic 819 762 884 341 1,002 707 1,427 1,913 2,592 2,020 Infrastructure 743 693 631 264 942 985 1,356 1,549 1,494 2,055 Public safety 377 411 482 674 538 1,212 725 1,502 1,980 2,374 MDA total 4,770 5,762 5,517 6,480 7,654 10,322 15,651 16,951 18,534 Multisector and other 290 699 1,068 170 1,446 69 3,207 4,911 2,319 2,545 Total expenditures 5,060 6,461 6,585 4,491 7,926 7,724 13,529 20,562 19,270 21,079 (discretionary sources)a Source: Budget statements and actuals from Reports of Ghana Auditor General, The Public Accounts of Ghana (Consolidated Fund). Note: Budget qualification in 2008 and 2009 are somewhat different from following years; MLGRD is included in social spending in 2010 and 2011. MDA = Ministries, Departments, and Agencies. a. Excludes social contributions (pensions, gratuities and social security); interest payments, subsidies, grants to other government units (NHF, GETFund, Road Fund, petroleum-related funds, DACA, transfers for GNPC from oil revenue), and social benefits (lifeline electricity subsidy). 128 TableB.3: Central Government Budget Allocation from Discretionary Sources (GH¢ million) Sectors/Ministries 2008 2009 2010 2011 2012 2013 2014 2015 Social 2,042 2,748 2,775 3,291 4,805 8,135 9,385 9,984 Ministry of Education 1,264 1,693 1,729 1,983 2,871 4,413 5,816 6,740 Ministry of Employment and Labor 38 61 39 44 Relations Ministry of Youth and Sports 91 44 46 54 54 36 36 National Commission for Civic Education 9 18 27 29 Ministry of Chieftaincy and Traditional 16 19 20 20 Affairs Ministry of Health 752 922 727 988 1,799 3,529 3,354 3,068 Ministry of Gender, Children and Social 16 39 91 44 Protection National Labor Commission 2 2 2 3 Ministry of Employment and Social Welfare 21 33 33 35 Ministry of Women and Children Affairs 5 9 9 13 Ministry of Local Government and Rural 233 226 Development Administration 789 1,148 745 707 2,009 2,552 1,500 2,101 Economic 819 762 884 1,002 1,427 1,913 2,592 2,020 Infrastructure 743 693 631 942 1,356 1,549 1,494 2,055 Public safety 377 411 482 538 725 1,502 1,980 2,374 MDA total 4,770 5,762 5,517 6,480 10,322 15,651 16,951 18,534 Multisector and other 290 699 1,068 1,446 3,207 4,911 2,319 2,545 Total expenditure (discretionary sources)a 5,060 6,461 6,585 7,926 13,529 20,562 19,270 21,079 Source: MoF budget statements. Note: Budget classification in 2008 and 2009 somewhat different from following years; MLGRD included in social spending in 2010 and 2011. a. Excludes social contributions (pensions, gratuities and social security); interest payments, subsidies, grants to other government units (NHF, GETFund, Road Fund, petroleum-related funds, DACA, transfers for GNPC from oil revenue), social benefits (lifeline electricity subsidy), and tax exemptions. TableB.4: Budget Allocation to Social Assistance Programs from the Consolidated Fund (GH¢ million) 2008 2009 2010 2011 2012 2013 2014 2015 Capitation grants 15.7 28.0 35.5 36.0 25.4 25.8 25.4 25.1 Subsidy-BECE 3.8 6.7 9.5 14.1 15.5 School feeding 17.1 17.0 50.0 50.0 60.0 199.0 School uniforms 12.2 10.0 28.8 28.0 10.0 8.0 Exercise books 29.0 28.7 10.0 8.0 LEAP 2.2 7.5 12.0 12.0 10.0 30.0 38.0 38.0 GoG scholarshipsa 21.8 22.0 22.0 30.0 15.0 Subsidy-SHS 11.4 20.0 22.8 30.0 48.2 50.7 60.8 66.1 LESDEP 84.0 75.0 Source: Budget financial statements. a. Additional financing from GETFund. 129 References Ampratwum, Edward, and Daniel Armah-Attoh. 2010. “Public Expenditure Tracking Survey– Tracking Capitation Grant in Primary Schools in Ghana. ” Results for Development and Ghana Center for Democratic Development (CDD Ghana). March. Ampratwum, Edward, Daniel Armah-Attoh, and Maxwell Agyei Ashon. 2012. “Public Expenditure Tracking Survey in Education–Tracking Possible Leakages in the Supply and Distribution of Textbooks in Public Primary Schools in Ghana.” Research Paper No. 20, Ghana Center for Democratic Development (CDD-Ghana). Bitran, Ricardo. 2014. “Field Report: Rapid Assessment of the Current Status of the Ghana’s National Health Insurance Scheme.” Britan Associados (June 12). Black, Maureen, Susan Walker, Theodore Wachs, Nurper Ulkuer, Julie Meeks Gardner, Sally Grantham-McGregore, Betsy Lozoff, Patrice Engle, and Meena Cabral de Mello. 2008. “Policies to Reduce Undernutrition Include Child Development.” The Lancet 371 (9611): 454–55. Bundy, Donald, Carmen Burbano, Margaret Grosh, Aulo Gelli, Matthew Jukes, and Lesley Drake. 2009. “Rethinking School Feeding: Social Safety Nets, Child Development, and the Education Sector.” World Bank and FAO. Coady, David, Margaret Grosh, and John Hoddinott. 2004. “Targeting of Transfers in Developing Countries: Review of Lessons and Experience.” World Bank, Washington, DC. Cooke, Edgar F. A, Sarah Hague, John Cockburn, Abdel-Rahmen El Lahga, and Luca Tiberti. 2014. “Estimating the Impact on Poverty of Ghana’s Fuel Subsidy Reform and a Mitigating Response.” Working Paper 2014-02, UNICEF and Partnership for Economic Policy (January). Darvas, Peter, and Robert Palmer. 2014. “Demand and Supply of Skill in Ghana–How Can Training Improve Employment and Productivity?” World Bank, Washington, DC. Ecorys Macro Group. 2013. “Ghana–Public Expenditure and Financial Accountability (PEFA) Performance Review” (June). FAO (Food and Agriculture Organization). 2013. Qualitative Research and Analyses of the Economic Impacts of Cash Transfer Programs in Sub-Saharan Africa: Ghana Country Case Study Report. Rome: Food and Agriculture Organization. http://www.fao.org/fileadmin/user_upload/p2p/Publications/Ghana_qualitative.pdf. GAG (Ghana Auditor-General). 2003. “Performance Audit Report of the Auditor-General on Foreign Scholarships.” http://www.ghaudit.org/gas/site/reports/index/page:5. ---------. . 2006. “Performance Audit Report of the Auditor General into the Administration of the Ghana Government Scholarship in the Second Cycle Institutions.” http://www.ghaudit.org/gas/site/reports/index/page:5. Ghana Aids Commission. 2014. “Country Aids Response Progress Report–Ghana–Reporting Period January 2012–December 2013” (March 31). Ghana, Government of (GOG). 2014a. “Economic and Financial Policies for the Medium Term” (April). Ghana, Government of (GOG). 2014b. “Rationalizing Social Protection in Ghana-Technical Notes” Ministry of Finance, Ministry of Gender, Children and Social Protection and World Bank Ghana Social Opportunity Project. Ghana, MELR (Ministry of Employment and Labor Relations). 2014. “National Employment Policy,” Volume1: Policy Document (August). Ghana, MESW (Ministry of Employment and Social Welfare). 2012a. “Ghana National Social Protection Strategy (GH-NSPS). Investing in People for a Better Ghana” (anuary). ---------. 2012b. “Operations Manual for the Livelihood Empowerment Against Poverty (LEAP) Program” (March). Ghana, MICS (Multiple Indicator Cluster Survey) . 2011. Ghana, Ministry of Education. 2010. “Education Strategic Plan 2010 to 2020, Volume 1, Policies, Strategies, Delivery, Finance.” Ministry of Education. Ghana, Ministry of Health. 2014. “Holistic Assessment of the Health Sector Program of Work 2013 (July). Ghana, MLGRD (Ministry of Local Government and Rural Development). 2010. “Draft Decentralization Policy Framework–Theme: Accelerating Decentralization and Local Governance for National Development” (April). Ghana, MLGRD and MoGCSP (Ministry of Local Government and Rural Development and Ministry of Gender, Children and Social Protection). 2014. “Social Opportunity Project– Draft LEAP Operational Manual” (April 14). Ghana, MoF (Ministry of Finance). 2014. “Budget Highlights for the 2015 Budget Statement and Economic Policy of the Government of Ghana” (November). Ghana, MoGCSP (Ministry of Gender, Children and Social Protection). 2014a. “Social Opportunity Project–Draft National Targeting Operation Manual” (April 14). ---------. 2014b. “Medium-Term Expenditure Framework (MTEF) for 2014–2016. Program-based Budget Estimates for 2014.” ---------. 2015. “Ghana National Social Protection Policy”. Ghana, NDPC (National Development Planning Commission). 2010a. “Ghana Shared Growth and Development Agenda (GSGDA) 2010–13, Volume I: Policy Framework.” http://ndpc.gov.gh/GPRS/Final20Draft20Policy20Framework20092010 .pdf. 131 ---------. 2010b. Ghana Shared Growth and Development Agenda (GSGDA) 2010–13, Volume II: Costing Framework.” http://ndpc.gov.gh/GPRS/GSGDA20Costing20Framework2028Vol.20II2 920-Final.pdf ---------. 2013. “The Implementation of the Ghana Shared Growth and Development Agenda (GSGDA) 2010–13. 2012 Annual Progress Report.” National Development Planning Commission, Accra (November). ---------. 2014. “Ghana Shared Growth and Development Agenda (GSGDA II), 2014–2017” (December). Ghana, NHIS (National Health Insurance Scheme). 2013. “National Health Insurance Scheme 10th Anniversary International Conference Report.” International Conference Centre, Accra (November). Ghana Statistical Service (GSS). 2012. Population and Housing Census–Summary Report of Final Results (May). ---------. 2014a. “Ghana’s Poverty Profile 2013–Ghana Living Standards Survey Rounds 5–6” (August). ---------. 2014b. “Ghana Living Standards Survey Round 6–Child Labor” (August). ---------. 2014c. “Poverty Profile in Ghana, 2005–2013. Ghana Living Standards Survey Round 6” (August). ---------. 2014d. “Ghana Living Standards Survey 6 (GLSS-6).” Main Report (July). ---------. 2014e. “Ghana Living Standards Survey 6 (GLSS-6). Key Findings: Labor Force Report” (October). ---------. . 2015. Ghana Poverty Mapping Report (May). Grantham-McGregor, S., Y. Bun Cheung, S. Cueto, P. Glewwe, L. Richer, B. Trupp, and the International Child Development Steering Group. 2007. “Developmental Potential in the First 5 Years for Children in Developing Countries.” The Lancet 369 (9555): 60–70. Grosh, Margaret, Carlo del Ninno, Emil Tesliuc, and Azedine Ouerghi. 2008. “For Protection and Promotion: The Design and Implementation of Safety Nets.” World Bank, Washington, DC. Handa, S, M. Park, R. O. Darko, I. Akoto-Osei, B. Davis, and S. Daidone. 2014. Livelihood Empowerment against Poverty Program Impact Evaluation, 3ie Grantee Final Report. New Delhi: International Initiative for Impact Evaluation (3ie) Horton, Richard. 2008. “Maternal and Child Undernutrition: An Urgent Opportunity.” The Lancet 371 (9608, January): 179. Kedze, Sandra. 2013. “The Distortive Effects of Ghana’s School Feeding Program on Primary Education Delivery: The Case of Adentan Municipality.” Research paper presented to the International Institute of Social Studies, The Hague, Netherlands, December. 132 IMF (International Monetary Fund). 2013. “Energy Subsidy Reform in Sub-Sahara Africa– Experience and Lessons.” African Department, International Monetary Fund, Washington, DC. Lake, Anthony. 2011. “Early Childhood Development–Global Action Is Overdue.” The Lancet 378 (9799, October 8): 1277–78. Lancet, The. 2013. “Maternal and Child Nutrition–Executive Summary.” The Lancet Maternal and Child Nutrition Series. http://www.thelancet.com/series/maternal-and-child-nutrition. Marques, José Silvério. 2013a. “Sierra Leone–Social Protection Assessment.” Draft report prepared for the Government of Sierra Leone and the World Bank (June). World Bank, Washington, DC. ---------. 2013b. “Botswana–Social Protection Assessment.” Draft report prepared for the World Bank and the Government of Botswana (June). World Bank, Washington, DC. [[okay?]] ---------. 2013c. “São Tomé e Príncipe–Politica e Estratégia Nacional de Proteção Social- Proposta.” Prepared for UNICEF and the Government of São Tomé e Príncipe. World Bank, Washington, DC. Molini, Vasco, Fabio Clementi, and Francesco Schettino. 2014. “Ghana: The Fuel Subsidy Reduction and its Impact. Povety and Scoial Impact Analysis.” Unpublished. Otoo, Nathaniel, Evelyn Awittor, Patricio Marquez, and Karima Saleh. 2014. “Universal Health Coverage for Inclusive and Sustainable Growth–County. Summary Report for Ghana.” Japan-World Bank Partnership Program for Universal Health Coverage. Health, Nutrition and Population Global Practice World Bank Group Report, World Bank, Washington, DC (September). Roelen, Keetie, and Helen Karki Chettri. 2014. “Researching the Linkages between Social Protection and Children’s Care in Ghana–LEAP and its Effects on Child Well-being, Care and Family Cohesion.” Centre for Social Protection (CSP), Institute of Development Studies (IDS) (October). Saleh, Karima. 2013. Health Sector in Ghana A Comprehensive Assessment. Directions in Development Series. Washington, DC: World Bank. Schieber, George, Cheryl Cashin, Karima Saleh, and Rouselle Lavado. 2012. Health Financing in Ghana. Directions in Development Series. Washington, DC: World Bank. Schwegler-Rohmeis, Wolfgang, Sarah Schmelzer-Roldán, Dan Baffour-Awuah, and Asamaoh Duodu. 2014. “Ghana–An Employment and Labor Market Analysis.” GIZ (November/December). SNV Netherlands Development Organization (SNV). 2011. “Baseline Report. Independent Civil Society Organization’s Monitoring and Evaluation of Social Accountability within the Ghana Social Feeding Program” (July). 133 UNICEF (United Nation’s Children’s Fund). 2013. “Improving Child Nutrition–The Achievable Imperative for Global Progress.” United Nation’s Children’s Fund, New York (April). UNICEF and CDD (Ghana Center for Democratic Development). 2014. “Ghana’s District League Table 2014–Strengthening Social Accountability for National Development” (December). United Kindgom, DFID (U.K. Department for International Development). 2014. Girls Participatory Approaches for Student Success (Girls-PASS). Annual Review (December). Walker, Susan, T. D. Wachs, J. M. Gardner, B. Lozoff, G. A. Wasserman, E. Pollitt, J. A. Carter, J. M. Gardner, B. Lozoff, and the International Child Development Steering Group. 2007. “ Child Development Risk Factors for Adverse Outcomes in Developing Countries.” The Lancet 369 (January 13): 145–57. WFP (World Food Programme). 2009. “Republic of Ghana–Comprehensive Food Security and Vulnerability Analysis” (May). http://www.wfp.org/content/republic-ghana- comprehensive-food-security-and-vulnerability-analysis-2009. ---------. 2013. “State of School Feeding Worldwide.” WFP, Rome, Italy. Wodon, Q., ed. 2012. “Improving the Targeting of Social Programs in Ghana.” World Bank, Washington, DC. http://wwwwds.worldbank.org/extemal/default/WDSContentServer/WDSP/IB/2013/04/1 0/000445729_20130410125313/Rendered/PDF/764840PUB0EPI001300PUB0DATE090 22012.pdf White, P., A. Hodges, and M. Greenslade. 2013. “Guidance on Measuring and Maximising Value for Money in Social Transfer Programs–Toolkit and Explanatory Text.” U.K. Department for International Development (DFID), London. WHO (World Health Organization) and World Bank. 2011. World Report on Disability. World Health Organization, Geneva, and World Bank, Washington DC. WHO, UNICEF, UNFPA (United Nations Population Fund), and World Bank. 2012. Trends in Maternal Mortality 1990–2010. Geneva. World Bank. 2005. World Development Report 2006: Equity and Development. Washington, DC: World Bank. ---------. 2010. Project Appraisal Document to the Republic of Ghana for a Social Opportunities Project. World Bank Report No. 52841-GH. ---------. 2011a. “Ghana–Joint Review of Public Expenditures and Financial Management.” DANIDA, KFW, and World Bank (October). ---------. 2011b. “Ghana Local Government Capacity Support Project. Appraisal Document” (February). ---------. 2012a. “Ghana Partnership for Education Grant Project. Project Appraisal Document” (October 10). 134 ---------. 2012b. The World Bank’s Social Protection and Labor Strategy 2012–2022—Resilience, Equity and Opportunity. Washington, DC: World Bank. ---------. 2013a. “Pension Reform Evaluation. PREM. Sub-Sahara Africa Region.” Unpublished. ---------. 2013b. “Country Partnership Strategy for the Republic of Ghana—FY13–FY16.” Washington, DC (August 20). ---------. 2013c. ‘‘Botswana: Social Protection Assessment” (December). ---------. 2014a. “Ghana-Maternal and Child Health & Nutrition Improvement Project.” PAD, Country Department AFCW1.Africa Region (April 25). ---------. 2014b. “AF Ghana Social Opportunity Project, Integrated Safeguard Data Sheet.” Report ISDSA7742 (April 30). ---------. 2014c. “Ghana-Social Opportunity Project-Implementation Status and Results” (July). ---------. 2014d. “Ghana-Nutrition and Malaria Control for Child Survival Project. Implementation Completion and Results Report” (January). ---------. 2014e. The State of Social Safety Nets 2014. Washington, DC: World Bank. ---------. 2015a. The State of Social Safety Nets 2015. Washington, DC: World Bank. ---------. 2015b. Ghana Poverty and Inequality Profile. Washington, DC: World Bank. ---------. 2016a (forthcoming). “Expanding Jobs Opportunities in Ghana.” Washington, DC. ---------. 2016b (forthcoming). “Ghana Public Expenditure Review 2016. Fiscal Consolidation for Growth and Employment.” Washington, DC. 135