Analytical report Solène Rougeaux

Transcription

1 Republic of Senegal Social Protection public expenditure review Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Analytical report Solène Rougeaux World Bank 2017

2 Table of Contents Acronyms and Abbreviations List of Figures List of Tables List of Boxes i iv iv v Chapter 1 Context: Population, Vulnerability, and Poverty 1 I Introduction 1 II Methodology 1 III Population and Vulnerability to Shocks 3 IV Senegal s Poverty Profile 8 Chapter 2 The Social Protection Sector in Senegal and Programs Covered by the Review 9 I Social Assistance Programs 10 II Programs for Emergencies and Response to Shocks 13 III Employment Programs 15 IV Social Insurance Programs 17 Chapter Social Protection Expenditure 21 I Overview of Social Protection Sector 21 II Rising Expenditure on Social Assistance Programs 26 III Funding Sources and Expenditure on Crisis Response Programs 30 IV Very Low Expenditure on Employment Programs 32 V Expenditure on Social Insurance Programs 33 Chapter 4 Analysis of the Coverage of Social Benefits 35 I A Life Cycle Approach to Social Protection Programs 35 II Cross-Cutting Programs Have Highly Unequal Coverage 37 III Healthcare Coverage Is High among Children under 5 and Pregnant Women While Other Groups Are Neglected 39 IV With the Exception of Coverage Provided by Cross-Cutting Programs, the Coverage of Programs Targeting School-Age Children Is Low and Decreasing 41 V Numerous Programs Target People of Working Age, but Many Needs Remain Unmet 42 VI One Quarter of People over 60 (the Target Group) Receive Pension and Health Benefits 45 Chapter 5 Program Efficiency 47 I Social Protection Programs Should Target Poor People as a Matter of Priority 47 i

3 II Matching Benefits to Relevant Objectives 51 III Lack of Financial Viability of the Retirement System 56 Chapter 6 Effectiveness of the Social Welfare System 60 I The RNU: A Tool with High Potential 60 II A Complex Institutional Mechanism 62 III Inefficient Financial Management of the Social Welfare System 67 Chapter 7 Recommendations 73 I Recommendation 1: Improve Program Targeting 73 II Recommendation 2: Strengthen the Efficiency and Effectiveness of the Social Protection System for Fighting Poverty 74 III Recommendation 3: Develop the Efficiency of Shock Response Mechanisms to Improve the Resilience of Poor and Vulnerable Populations to Covariant Shocks 76 IV Recommendation 4: Promote Productive Household Investments in order to Increase the Productivity of the Most Vulnerable Populations 78 V Recommendation 5: Revise the Entire Pension System to Ensure Its Financial Sustainability and Fairness 79 Bibliography 80 Annex 1 Description of social protection programs included in the analysis 83 Annex 2 : Total spending from government s budget from 2010 to 2015 per social protection program in millions of CFA 89 Annex 3 : Total spending from external funding (donors) from 2010 to 2015 per social protection program in millions of CFA 93 Annex 4 : Number of beneficiaries per social protection program from 2010 to ii

4 Acronyms and Abbreviations ACMU ANPEJ ANSD ARC ASPIRE CBO CEC CFAF CMU CNAAS CNCAS CONFEMEN CPIA CRES CSO CSS DCAS DGAS DGCPT DGPSN DHS DP DPEE DSPRV EDS ESPS FNR FSC FSGI FSN GDP HEA IGA IMS Universal Health Coverage Agency (Agence de la Couverture Maladie Universelle) National Agency for Youth Employment (Agence Nationale pour la Promotion et l'emploi des Jeunes) National Statistical and Demographic Agency (Agence Nationale de la Statistique et de la Démographie) African Risk Capacity Atlas of Social Protection Indicators of Resilience and Equity Community-Based Organization Equal Opportunity Card (Carte d'égalité des Chances) CFA Franc Universal Health Coverage (Couverture Maladie Universelle) National Agricultural Insurance Fund of Senegal (Caisse Nationale d'assurance Agricole du Sénégal) National Farm Credit Bank of Senegal (Caisse Nationale de Crédit Agricole du Sénégal) Conference of Ministers of Education of Francophonie States and Governments (Conférence des Ministres de l'éducation des États et Gouvernements de la Francophonie) Country Policy and Institutional Assessment Consortium for Economic and Social Research (Consortium pour la Recherche Économique et Sociale) Operational Monitoring Unit (Cellule de Suivi Opérationnel) Social Security Fund (Caisse de Sécurité Sociale) Directorate of School Feeding programmes (Direction des Cantines Scolaires) General Directorate of Social Action (Direction Générale de l'action Sociale) General Directorate of Public Accounting and the Treasury (Direction Générale de la Comptabilité Publique et du Trésor) General Delegation for Social Protection and National Solidarity (Délégation Générale à la Protection Sociale et à la Solidarité Nationale) Demographic and Health Survey Development Partner Directorate of Forecasting and Economic Research (Direction de la Prévision et des Études Économiques) Directorate of Salaries, Pensions, and Annuities (Direction de la Solde, des Pensions, et Rentes Viagères) Demographic Health Survey (Enquète Démographique de la Santé) Poverty Monitoring Survey in Senegal (Enquète de Suivi de la Pauvreté au Sénégal) National Retirement Fund (Fonds National de Retraite) Food Security Commission (Commissariat à la Sécurité Alimentaire) Special Flood Management Fund (Fonds Spécial de Gestion des Inondations) National Solidarity Fund (Fond de Solidarité Nationale) Gross Domestic Product Household Economy Approach Income-Generating Activity Information Management System i

5 IPAR IPM IPRES LEAP LFI LFR MDG MEFP MYBP NGO NSER OCHA ORSEC OSB PAPA PASEC PLASEPRI PMT PNBSF PRBC PRNIA PRODAC PRODES PRP PRSP PSE RNU SAR SE/CNSA SENELEC SIGFIP SISBEN SMART SNPS SO TSA Agricultural and Rural Forecasting Initiative (Initiative Prospective Agricole et Rurale) Health Insurance Fund (Institut de Prévoyance Maladie) Retirement Savings Fund of Senegal (Institut de Prévoyance de Retraite du Sénégal) Livelihood Empowerment Against Poverty Initial Finance Bill (Loi de Finance Initiale) Amending Finance Bill (Loi de Finance Rectificative) Millennium Development Goal Ministry of the Economy, Finance, and Planning (Ministère de l'économie, Finances, et Planning) Multi-Year Budget Planning Non-Governmental Organization National Socio-Economic Register Office for the Coordination of Humanitarian Action Public Safety Response Planning (Organisation de la Réponse de Sécurité Civile) Livestock Protection Program (Opération de Sauvegarde du Bétail) Elderly Support Program (Project d'appui à la Promotion des Aînés) Program for the Analysis of Educational Systems in CONFEMEN Countries (Programme d'analyse des Systèmes Éducatifs de la CONFEMEN) Private Sector and Valorization of the Senegalese Diaspora in Italy Support Platform (Plateforme d'appui au Secteur Privé et à la Valorisation de la Diaspora Sénégalaise en Italie) Proxy Means Test National Family Security Grants Program (Programme National de Bourses de Sécurité Familiale) Community-Based Re-Adaptation Program (Programme de Réadaptation à la Base Communautaire) National Food Insecurity Response Program (Plan de Riposte Nationale à l'insécurité Alimentaire) Community Agricultural Programs (Programme des Domaines Agricoles Communautaires) Economic and Social Development Dynamics Support Program (Programme de Renforcement des Dynamiques de Développement Économique et Social) Poverty Reduction Program Poverty Reduction Strategy Paper Emergent Senegal Plan (Plan Sénégal Émergent) Single National Registry (Registre National Unique) African Refinery Corporation (Société Africaine de Raffinage) Executive Secretariat of the National Food Security Council (Secrétariat Exécutif du Conseil National à la Sécurité Alimentaire) National Electricity Company of Senegal (Société Nationale d'éléctricité du Sénégal) Integrated Public Finances Management System (Système Intégré de Gestion des Finances Publiques) Social Subsidies Beneficiary Identification System (Sistema de Identificación de Beneficiarios de Subsidios Sociales) Standardized Monitoring and Assessment of Relief and Transitions National Social Protection Strategy (Stratégie Nationale de Protection Sociale) Strategic Objective Targeted Social Assistance ii

6 UNICEF USD WFP United Nations Children s Fund United States Dollar World Food Program iii

7 List of Figures Figure 1: Chronology of the main shocks to Senegal s agricultural production, ( = 100)... 5 Figure 2: Expenditure on social services sectors, in millions of CFAF and as share of government s total social expenditure Figure 3: Comparing remittances, cash transfer programs, and total social protection expenditure Figure 4: Expenditure as share of total, by type of social protection Figure 5: Total expenditure on social protection as share of GDP Figure 6: Average expenditure on three types of social protection as share of GDP by country Figure 7: Total social protection expenditure, by funding source Figure 8: Change in external funding as share of total expenditure, by program type Figure 9: Total expenditure on social assistance programs Figure 10: Total expenditure on social assistance programs, excluding student grants Figure 11: Expenditure on social assistance programs as share of total social assistance spending in Africa Figure 12: Total social assistance expenditure by funding source Figure 13: Comparison of expenditure on subsidies and social assistance, excluding higher education grants Figure 14: Total expenditure on crisis response programs Figure 15: Total expenditure on employment programs, in millions of CFAF.33 Figure 16: Expenditure on social insurance, in millions of CFAF Figure 17: Expenditure on pensions as share of GDP Figure 18: Beneficiaries of the school feeding program Figure 19: Recipients of retirement and survivors pensions Figure 20: Percentage of the population over 60 receiving benefits under national pension systems Figure 21: FNR coverage and cost (% of GDP) Figure 22: Percentage of programs, by targeting method Figure 23: Percentage of expenditure, by targeting type (2015) Figure 24: Breakdown of students receiving university scholarships, by wealth quintile Figure 25: Pension recipients over 60, by wealth quintile Figure 26: Percentage of active population with access to a healthcare system, by wealth quintile Figure 27: Use of social safety nets as shock response mechanism in Africa.53 Figure 28: IPRES demographic profile Figure 29: Population breakdown by age group and projection Figure 30: Rate of contribution to public sector retirement systems Figure 31: Construction of social net systems in Africa Figure 32: Availability of financial resources and response to food insecurity (food security stock) List of Tables Table 1: Occurrence of self-reported shocks, in percentages... 6 Table 2: Shock response mechanism reported by households, in percentages... 7 Table 3: Strategic objectives of the SNPS... 9 Table 4: Social assistance programs Table 5: Crisis-response programs Table 6: Employment programs Table 7: Social insurance programs iv

8 Table 8: Description of pension schemes Table 9: Total expenditure on social protection, in millions of CFAF Table 10: Social protection programs through the life cycle Table 11: Coverage of cross-cutting programs Table 12: Coverage of programs targeting children under 5 and pregnant women Table 13: Coverage of programs benefiting children aged Table 14: Coverage of programs targeting people of working age Table 15: Coverage of programs targeting people over the age of Table 16: Breakdown of vulnerable individuals and households, by poverty level Table 17: Total school cafeteria expenditure, per beneficiary (in CFAF millions) Table 18: Total expenditure per beneficiary from preschooler item (in CFAF millions) Table 19: RNU population and percentage in poverty Table 20: Filters that could be used in the RNU Table 21: Description of powers transferred in social sectors Table 22: Social welfare budget for the city of Pikine (in CFAF millions) Table 23: Change in budget allocation between initial finance bill (LFI) and amending finance bill (FLR) 70 List of Boxes Box 1: Choosing politically appropriate program parameters Box 2: Impact of money transfer programs in Africa Box 3: How Brazil, Colombia, Ghana, and Senegal created a social register for better efficiency and greater impact Box 4: Institutional change and anchoring in law v

9 Acknowledgements: This report benefited from the contribution of Aline Coudouel, Serigne Moussa Dia, Ameth Faye, Melis Guven and Emma Montiel. The insightful comments of the peer reviewers Julio Ricardo Loayza, Jamele Rigolini et Eric Zapatero Larrio are greatly appreciated. Standard Disclaimer This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. vi

10 Chapter 1 Context: Population, Vulnerability, and Poverty I Introduction 1. For the past fifteen years, Senegal has been committed to protecting vulnerable groups. Having become aware of this issue in the early 2000s thanks to the diagnosis conducted during the participative formulation process for Poverty Reduction Strategy Paper (PRSP) I implementation, Senegal made the protection of vulnerable groups a strategic priority in each of its development plans. In 2013, an interministerial monitoring committee working as part of the National Social Protection Strategy (SNPS) was established to improve coordination in the sector and enhance the efficiency of the social protection system, and more specifically, the social safety net system. 2. The Government based the Emerging Senegal Plan (PSE) on the drafting of a Social Safety Net System capable of responding to the issue of chronic poverty and the protection of populations vulnerable to shocks. The strategy is based on three pillars: (i) growth, productivity, and the creation of wealth; (ii) human capital, social protection, and sustainable development; and (iii) governance, institutions, peace, and security. In October 2013, the Prime Minister began the Government s General Policy Declaration by defining how the social agenda would address the issue of vulnerability and by underlining the special focus on improving access to basic, quality social services, the expansion of the national monetary transfers program, access to universal health coverage, support for vulnerable populations, and the implementation of job creation programs. 3. Although an analysis of social sector expenditure was conducted, it did not include the social protection sector. For the first time, the Government, with the support of UNICEF, analyzed social sector public expenditure from 2006 to 2013 (UNICEF, 2016). The purpose of this analysis was to aid policy makers in the planning and management of public finances. This analysis is now to be conducted on a yearly basis. Although this represents a major step forward in the planning and management of the social sectors in general, it does not include a detailed review of social protection programs as it focuses primarily on health and education expenditure. This is due in part to the fact that social protection is multisectoral and that collecting information about the sector and determining its exact scope is particularly difficult. 4. Faced with a lack of budget-related information about the sector, the inter-ministerial SNPS steering committee recommended in December 2015 that an expenditure review of the sector be carried out. The objectives of this review were to: (i) evaluate the budgets spent with regards to the needs and allocated resources; (ii) analyze the coverage of social protection programs; and (iii) evaluate the efficiency of key programs and of the system as a whole. This endeavor is the first of its kind in Senegal and may serve as a stepping stone for the Government to conduct an annual review of the sector s expenditure and thus help improve its management. II Methodology 5. The review concerns only the list of programs approved by the technical support committee of the inter-ministerial SNPS steering committee. These programs consist of the social protection programs 1

11 financed or implemented by the Government and considered the most important in the current social protection system. In addition to these programs, the committee decided to include programs contributing to job creation and promotion. Although they do not usually fall within the scope of social protection in Senegal, they are considered internationally to be part of the sector as they directly contribute to the objective of empowering and promoting social protection as defined in the PSE and Strategic Objective (SO) 2 of the SNPS. This list is not exhaustive and was established based on the SNPS technical committee s consensus and knowledge (see Annex 1 for the list of programs included in the review). Flagship programs implemented by the World Food Program (WTF) have been included in this list given the size of the allocated budget and the number of intended beneficiaries. 6. Data was collected using the ASPIRE 1 nomenclature for the period As a result, a large amount of data was collected in standard format and in as uniform a way as possible while also contributing to the global effort to systematically collect data on social protection. Specifically, the data collected included: (i) annual program data on number of beneficiaries, amounts spent under each program, administrative costs, expenditure on services, and expenditure by technical and financial partners; and ii) information on program characteristics, including the program start year, a description of the target mechanism, program conditionality and implementation, etc., which are instrumental to understanding how these programs operate. 7. The review focuses on four categories of social protection instruments: (i) social assistance, such as monetary transfers, student loans, school feeding programs, and nutrition programs; (ii) emergency and shock response programs; (iii) job market programs; and (iv) social security programs, such as retirement benefits and health coverage. The author of the present study classified the social protection instruments into these four categories based on the broad categories defined in the ASPIRE framework. The emergency and shock response instruments were given their own category in order to highlight them in the analysis as their connection to social assistance programs as well as their funding are topics of current interest in Senegal. Although the social protection expenditure reviews often include subsidies in these instrument categories, they were not included in this analysis due to the limited amount of data available. However, they will be referred to in order to contribute to the examination and analysis of social protection programs. 8. The financial and program data were collected from four main sources, firstly by using the budget implementation figures generated by the Integrated Public Finances Management System (SIGFIP) administered by the Ministry of the Economy, Finance, and Planning s (MEFP) Directorate of Informatics, secondly by consulting with the implementing ministries and agencies and their annual reports, and finally by using the spreadsheet on external funding of the Directorate of Investment of MEFP (2017). Total expenditure for Education and Health data were obtained directly from the Directorate of Forecasting and Economic Studies (DPEE). 1 Within the framework of the ASPIRE project, the Global Practice of the World Bank s Social Protection and Employment sector initiated the systematic collection of administrative data on total and disaggregated government expenditure on social assistance, social insurance, and employment programs as well as on social services and general grants throughout the world. The purpose of this comprehensive exercise is to create a transparent and comparable database that will enable the identification of long-term trends in social protection expenditure by program type as well as various criteria pertinent to policy outcomes. 2

12 9. This review is based on the social protection expenditure review manual (World Bank, 2009). The documents consulted for this analysis included recent public expenditure reviews conducted in Africa and South America as well as agricultural expenditure reviews (CRES et al., 2012), education expenditure reviews (World Bank, 2015a), the social sector expenditure review (World Bank and UNICEF, 2016), the food security in Senegal expenditure review (DAI/European Union, 2015), the World Bank Poverty Assessment Report (2015b), and the IPAR agricultural subsidies review (2015). 10. Data collection was limited by several constraints. Weaknesses in the archiving system, institutional changes, and the lack of a strong accounting, monitoring, and evaluation system led to a significant loss of information over time. Annual reports dating back more than three years were difficult to obtain, and the results these reports provided were not always exploitable. These constraints are neither new nor specific to this review, being often mentioned in previous public expenditure reviews. III Population and Vulnerability to Shocks According to the 2013 general census, Senegal has a population of 13,508,715 inhabitants. 54.8% of the population lives in rural areas and 23% lives in and around Dakar. From 2001 to 2011, the population increased by 2.5% (World Bank, 2015c). The population comprises 1,618,363 households. Average household size is 8 people. Life expectancy is 63.2 years for men and 66.5 years for women, with disparities between rural and urban areas as well as between men and women. 5.9% of the population reports suffer from a disability (mild disability: 4.2%; severe disability: 1.7%). More women are disabled compared to men. Furthermore, 16.6% of the population states that they do not have a birth certificate, and this rate reaches over 30% in some regions (Tambacounda, Kaffrine, Kolda). 12. Senegal is estimated to have a working-age population 3 of 7,728,868 individuals. However, 49.7% of individuals in this age group is inactive. Housewives represent 43.8% of the inactive population, while students represent 30.4%. The potentially active population comprises 3,704,369 individuals. The activity rate is slightly higher in rural areas (51.3%) than in urban areas (49.6%). Fully 25.7% of the potentially active population are unemployed, yet nearly all of these unemployed (9 out of 10) are first-time job seekers. Only 6.4% of the population has earned a bachelor s degree or attained an equivalent level of education. A total of 755,532 of Senegal s households, or 49.5% of the population, report being agricultural. However, only 11.4% of these households are affiliated with a producer s organization. A very large majority of these households are small producers who farmed less than 5 hectares during the crop year. 13. According to the recent analysis conducted by the World Bank (2016b) based on data from the Senegal National Employment Survey, the situation of young people of working age is alarming. On the one hand, the number of young people in the labor market increases every year. 200,000 young people entered the labor market in 2000, compared to 300,000 in 2015, and this number will increase to approximately 400,000 in 2025 and 670,000 in Meanwhile, economic conditions have worsened, 2 All of the data in this section are taken from the general census (ANSD, 2013) unless explicitly stated otherwise. 3 Aged 15 years or older. 3

13 and the job market is unable to provide work for everyone. Approximately 35% of young adults aged participate in the labor market, with disparities between men and women. In the age bracket in rural areas and the bracket in urban areas, the majority of young women fall into the neither/nor category in that they neither work nor study. However, this group is much smaller among young men, ranging from 20 to 25% for those 25 or older. In addition to the low participation rate in the labor market, approximately 27% of young people are underemployed (reporting that they work less than 40 hours a week). 14. Despite the progress made in the past decade to achieve the Millennium Development Goals (MDG), Senegal continues to earn low scores when it comes to infant, child, and maternal mortality rates. The infant mortality rate is 39 per 1,000, and the neonatal mortality rate is 23 per 1,000 live births (ANSD, 2015). Overall, 59 children out of 1,000 do not reach their fifth birthday (ANSD, 2015). The maternal mortality rate shows large disparities depending on the region, varying from 921 deaths per 100,000 live births in Kedougou to 271 in Thiès, with the national average being 392 (ANSD, 2012). 15. The nutritional status of children aged 0 to 5 has improved since the 2000s but remains a major concern. In 2014, the acute malnutrition rate was 5.9%, while the chronic malnutrition rate was 18.7% for children aged 0 5 (ANSD, 2014). 16. According to the 2013 general census, the literacy rate has decreased over the past ten years. Only 42.6% of people 6 years of age or older are literate, compared to 59% ten years ago, while 58.15% of those aged and 64.1% of those aged are literate. The literacy rate then decreases with age. The literacy rate is higher for urban populations compared to rural populations, and men are more literate than women. 17. The level of education remains low. The latest results of the PASEC survey (2015) show that approximately 40% of children cannot read at the end of their elementary education. Furthermore, the dropout rate is high as only 40% of children finish elementary school (42% of boys, 37% of girls). Overall, only one third of children of elementary school completion age can read and do basic math. In addition, the census reveals discrimination in access to education for girls, beginning at age 13. From this age on, the proportion of girls in total enrollment decreases through to the end of tertiary studies. 18. The vulnerability of children is very high. In 2014, 27.3% of children under 5 had not been registered at birth, and this rate reached over 50% in some areas such as Kaffrine and Tambacounda (UNICEF, 2016). The census revealed that 6.9% of children aged 6 14 participate in the labor market (a rate that is slightly higher for boys). According to Human Rights Watch (2016), over 30,000 children (primarily boys) beg on the streets of Dakar. Girls are victim of practices such as early marriage and genital mutilation. According to the data recently published by UNICEF (2016), in 2013, 3.7% of girls aged and 24.5% of girls aged were married. In 2014, according to the same publication, it was estimated that 12.9% of girls under 15 had undergone genital mutilation. These statistics vary greatly depending on the region. For example, it is highly prevalent in Matam (52.6%), Sedhiou (50.5%), and Kolda (45.7%), while it is close to zero in regions such as Diourbel, Thiès, and Kaolack. 4

14 Shocks Affecting Senegal and Its Households Senegal s economy is often affected by major shocks that affect its economic performance. Exogenous shocks, notably the increase in the price of imported goods or the effects of the global economic recession, have a significant impact on Senegal because of the nature of its economy, as was clearly demonstrated during the financial, food, and oil crises. Senegal imports all of its oil (which serves to produce the majority of the electricity consumed), and 80% and 100% of its consumption of rice and wheat, respectively. In , the local market price of rice tripled, grain prices increased by 50%, and that of other staples such as sugar, wheat, and milk increased by 30%. The rise in fuel prices had a significant impact on the energy sources of the poorest households, such as butane gas. However, in recent years, Senegal s economy has benefited from lower oil prices and stable food prices. 20. Flooding represents a particular threat for urban and suburban areas, while droughts primarily affect rural areas. Given the dependence on subsistence agriculture of many households, the cultivation of rain-fed seed varieties and the high poverty levels in rural areas, droughts, and locust infestations can have a significant impact on the country s food security. For example, as a result of the 2011 drought, agricultural production fell by 20%, and 25% of households considered themselves food insecure in 2013 compared to 15% in This also had an impact on malnutrition, as the number of stunted children increased to 26.5% (compared to 19.6 in 2005) (ANSD, 2012; ANSD, 2015). Figure 1, which is taken from the Senegal Agricultural Sector Risk Assessment (World Bank, 2015d), shows the chronology of the main shocks that affected agricultural production from 1980 to Over 33 years, Senegal s agricultural sector faced 10 significant shocks. FIGURE 1: CHRONOLOGY OF THE MAIN SHOCKS TO SENEGAL S AGRICULTURAL PRODUCTION, ( = 100) Source: World Bank, Senegal Agricultural Sector Risk Assessment, 2015d 4 The following paragraphs are taken from two World Bank documents: République du Sénégal: Évaluation des filets sociaux (2013) and Revue diagnostique du financement de la réponse aux catastrophes (2016 a). 5

15 21. In addition to this variability in rainfall over the short term, Senegal faces a decrease in rainfall over the long term. From 1960 to 2000, annual rainfall decreased regularly, dropping from 650 to 600 millimeters. Over the same period, peanut production fell from 1 million to 750 thousand tons. Climate change affects the variability of rainfall and climate patterns and compounds the impact of flooding and drought. Given the current climate change trajectory in Senegal, agricultural production modeling research suggests a yield loss of 5 25% by Senegalese households are affected by covariant and idiosyncratic shocks and have few strategies for confronting them Senegalese households are affected by covariant and idiosyncratic shocks. 6 According to data from ESPS II, households are most frequently affected by the following shocks: 7 illness or serious accident, loss of livestock, and loss of harvest through fire, drought, or flooding. These shocks are much more frequent in rural areas compared to urban areas. Twice as many poor households report having been affected by at least one shock the year before (more than 40% for the two poorest quintiles compared to 16.5% for the richest quintile). TABLE 1: OCCURRENCE OF SELF-REPORTED SHOCKS, IN PERCENTAGES Quintile Place of Residence Type of Shock Total Urban Rural Death of family breadwinner Illness or serious accident Loss of employment Bankruptcy of family business Loss of harvest (fire, drought, flooding, etc.) Loss of livestock (fire, disease, theft, etc.) Significant loss of income (temporary layoff, etc.) Partial or complete loss of home (fire, flooding, etc.) Loss of main means of production At least one of these shocks Source: Échevin (2012) based on ESPS data (ANSD, 2011) 5 The following paragraphs are taken from République du Sénégal: Évaluation des filets sociaux (2013) 6 Covariant shocks affect all households in a given zone or group, while idiosyncratic shocks affect only the household concerned. 7 Based on household self-reporting. 6

16 23. Over half of households lack a specific adaptation strategy for dealing with shocks. As shown by Table 2, nearly one quarter of households reported having used their savings to deal with a shock, especially if it concerned their health (risk of illness or death) or bankruptcy. Selling personal property was the most common alternative strategy, despite the fact that it can lead to poverty in the long term due to the loss of assets. Family support within the country or from abroad was mentioned by 27% of the respondents. Other sources of support were minimal. Only 2% of cases received support from an NGO, and government aid was only mentioned by 1% of households. This shock response pattern reveals the high vulnerability of households. In effect, the majority of them have no formal adaptation strategy, and when they do have one, it primarily relies on selling assets or using savings, both of which are less accessible options for the poor. Since the loss of livestock and harvest were the shocks the poorest households mentioned most often and are also the shocks for which there are the fewest adaptation and aid strategies, the situation could have serious repercussions on adaptation strategies that are negative in the short term and perpetuate poverty traps in the long term. TABLE 2: SHOCK RESPONSE MECHANISM REPORTED BY HOUSEHOLDS, IN PERCENTAGES Types of Shocks Govt. Aid NGO, OBC Support Sale of Assets Use of Savings Loan Family Support within Country Family Support from Abroad Aid from Friends No Strategy Death of the family breadwinner Illness or serious accident Loss of employment Bankruptcy of family business Loss of harvest (fire, drought, flooding, etc.) Loss of livestock (fire, disease, theft, etc.) Significant loss of income (temporary layoff, etc.) Partial or complete loss of home (fire, flooding, etc.) Loss of main means of production Source: Échevin (2012) based on ESPS data (ANSD, 2011). 7

17 IV Senegal s Poverty Profile Despite a marked decrease in Senegal s poverty rate in the early 2000s, this began to stagnate in 2005, reaching 46.7% in 2011, and extreme poverty has not decreased significantly since Recent analyses using a new consumption aggregation show that the poverty rate decreased between 2 and 6 percentage points over the period (World Bank, 2016d). Although the poverty rate fell from 55.2% to 46.7% over 10 years, the number of poor people continued to increase, exceeding 6.3 million in 2011, while the poverty rate in rural areas is twice as high as in Dakar. Extreme poverty stands at around 14.8%. The depth of poverty from the poverty line fell slightly, from 17.2% in 2005 to 14.5% in 2011, while the poverty severity index fell from 7.3 to 6.6 over the same period. However, the slight increase in these indicators in urban areas (not including Dakar) shows that the well-being of the poorest populations in these zones has deteriorated. The decrease in the acceleration of poverty over the period can be explained by the low growth that began in Moreover, this growth was not inclusive and did not benefit the poor or the lowest 40% of the population. 25. Poor households generally consist of several individuals who lack education and work in agriculture or the informal sector. In 2011, 80% of poor households had 10 or more members, and over one quarter had over 20 members. 83% of poor heads of households had not received a formal education. The number of poor people living in households headed by a woman was also much lower (34.7%) compared to that of households led by a man (50.6%). 26. Poverty more often affects rural areas but is concentrated in Dakar in terms of population size. Nearly 70% of the poor and 84% of the extreme poor lived in rural areas in 2011, while 13% of the poor and 2% of the extreme poor lived in Dakar. Regionally, poverty is concentrated in Dakar and its neighboring regions due to their significant population density. Casamance also has a strong concentration of poverty, with a higher poverty rate than the national average. The lowest poverty rates are found in Niayes region and the Senegal River valley, which have irrigation infrastructures and horticultural production. 27. Many Senegalese households are highly vulnerable to idiosyncratic and covariant shocks, and populations around the poverty line frequently move in and out of poverty. Idiosyncratic shocks such as illness or the death of the head of household can affect up to 20% of households each year. Agricultural households are highly vulnerable to climate-related shocks due to their lack of assets and limited access to loans or insurance. Selling what little property they own is often their only way to limit the impact of the shock, even though it compromises their future income. Education appears to be an essential factor in helping households move out of poverty and in reducing their risk of falling into poverty. Disabilities are also a factor associated with chronic poverty. 8 All of the data in this section are taken from the Poverty Assessment (World Bank, 2015b) unless explicitly stated otherwise. 9 The next poverty assessment is scheduled for 2017 and will further refine these analyses. 8

18 Chapter 2 The Social Protection Sector in Senegal and Programs Covered by the Review 28. The Emerging Senegal Plan (PSE) defines social protection as a set of measures that aim to protect citizens from social risks. Social protection includes public social security and private or community-based schemes. It involves three principles: assistance, insurance benefits, and empowerment of social categories. This definition aligns with that articulated by the World Bank in its strategy, which states that social protection and labor systems, policies, and programs help individuals and societies manage risk and volatility and protect them from poverty and destitution through instruments that improve resilience, equity, and opportunity. 29. Written in 2005, the first National Social Protection Strategy (SNPS) covered a 10-year period. The new SNPS adopted in 2017 will orient social protection policy through to This new national strategy adopts the PSE s definition of social protection. The long-term objective of the SNPS is to build a social protection system that benefits all Senegalese men and women, guarantees a minimum income for all, and provides health insurance and safety nets in order to strengthen the resilience of individuals experiencing shocks and crises that could cause them to slip into poverty. The strategy includes five strategic objectives based on the lifecycle. Each objective aligns with one or more of the social protection objectives outlined in the PSE (Table 3). TABLE 3: STRATEGIC OBJECTIVES OF THE SNPS PSE social protection objectives SNPS Strategic Objectives (SO) Assistance Insurance Empowerment SO1: Support integrated social protection for all children X X SO2: Set up programs and schemes for people of working age SO3: Set up a system of guaranteed minimum income and health coverage for the elderly X X SO4: Set up an integrated social security system for those who are disabled or unable to work X X X SO5: Strengthen community resilience to shocks and disasters Source: Author, compilation of collected data X X X X 30. This review covers 50 social protection programs implemented during the period. All 50 programs were administered by the government, with the exception of the WFP food vouchers and school feeding program and private sector social insurance programs. All programs meet at least one of the social protection objectives outlined in the PSE (assisting, insuring, and empowering individuals and households against social risks). 31. These programs are heterogeneous and have few overlaps. Some are new, while others are older. Some are implemented nationwide, others in only a few regions, and some are entirely governmentfunded while others are funded primarily by external resources. (See Table 4 for a complete list of X 9

19 programs included in this review.) I Social Assistance Programs 32. This review covers 20 social assistance programs. These can be grouped into six sub-categories: (i) cash transfer programs; (ii) student grants; (iii) targeted subsidies to improve healthcare access by vulnerable groups; (iv) school feeding programs; (v) nutrition programs; and (vi) social services for children. 10 These social assistance programs vary in terms of budget and number of beneficiaries. All are nationwide (except the school feeding program, which was designed on a national scale but implemented by different actors, and the nutrition programs, which use geographic targeting based on malnourishment data). However, despite being national programs, most reach a limited number of beneficiaries, with only 3 programs impacting more than one million beneficiaries yearly. The others reach from a few hundred individuals (3 programs) to a few hundred thousand (5 programs), while 7 programs involve targeted subsidies for improving healthcare access. While 4 programs involve childcare services, 3 of them have between 0 and a few hundred beneficiaries. These 20 programs are implemented by 11 different agencies or ministries. Table 4 lists and describes these programs. TABLE 4: SOCIAL ASSISTANCE PROGRAMS Program Program description Start date Implementing agency Cash transfer programs 1 National Family Security Grants Program (PNBSF) Quarterly cash transfers of CFAF 25,000 to the poorest households for a period of 5 years 2013 General Delegation for Social Protection and National Solidarity 2 3 Program to reduce child vulnerability Support program for wards of the state Cash transfers to households with vulnerable children to help them keep children in school and provide them with healthcare services Cash transfers to orphans (many orphaned by the 2002 Joola capsizing disaster) Student grants Ministry of the Family Directorate of Child Protection National Office for Wards of the State 4 Grants for higher education Grants for higher learning (full or partial tuition) for all university students in Senegal or abroad 2007 Ministry of Higher Education Directorate of Higher Education Grants Targeted subsidies to improve access to health care 10 The author based these categories on ASPIRE groupings and on what was important to emphasize about Senegal. For instance, student grants are not an independent category in ASPIRE but are included in the Cash Transfers category. Given the budget these programs receive in Senegal, the author opted to distinguish them from cash transfer programs. 10

20 5 Medical care for the destitute Provides medical coverage for the destitute 2009 Ministry of Health and Social Action General Directorate of Social Action 6 HIV/AIDS program Provides screening and treatment for people affected by HIV 2001 National AIDS Council 7 SESAME Plan 8 Free healthcare for children under 5 Guarantees access to healthcare for people over the age of 60 Children under 5 receive medical care at no cost Universal Health Coverage Agency Universal Health Coverage Agency 9 Medical insurance through communitybased health plans. Plan enrolment fees Insurance branch of Universal Health subsidized by the government (50% Coverage 100% of cost, or CFAF 3,500 7, Universal Health Coverage Agency 10 Free cesarean sections Covers labor and cesarean sections in health centers and hospitals 2004 Universal Health Coverage Agency 11 Dialysis Covers the cost of dialysis services School feeding programs Directorate of School feeding programs School feeding programs Offers one hot meal per day in elementary and secondary schools in rural and peri-urban areas 2006 Universal Health Coverage Agency Ministry of Education Directorate of School feeding programs 13 WFP School feeding Program Offers one hot meal per day in elementary and secondary schools in rural and peri-urban areas 1970 World Food Program Nutrition programs 14 Community nutrition program Aims to provide each Senegalese citizen with adequate nutrition and promotes nutritional habits to maximize well-being and community development. Screening, awareness, and distribution of therapeutic foods National Committee to Combat Malnourishment 15 Social transfers to combat malnourishment Temporary cash transfer programs to combat malnourishment and promote prenatal care for pregnant women 2009 National Committee to Combat Malnourishment Services for children 11

21 16 Program for disadvantage d children Assists vulnerable children (talibés, street children, orphans, etc.): education kits, subsidies for Islamic schools, training, etc Ministry of Health and Social Action General Directorate of Social Action 17 Community daycare Community daycare for children under Ministry of the Family Poverty Monitoring Committee 18 Program for street children Helps get children off the street 2004 Ministry of the Family Poverty Monitoring Committee 19 Early childhood education program Education centers for children aged 3 6. Includes learning, medical care, and nutritional support 2002 National Agency for Early Childhood Education Other 20 Equal Opportunity Cards (CEC) CECs help people with disabilities access equipment, training, healthcare, etc Ministry of Health and Social Action General Directorate of Social Action Source: Author, compilation of collected data 33. Four of these programs merit further description due to their scope and budget. The National Family Security Grants Program (PNBSF) began in December It provides conditional cash transfers to Senegal s poorest households. Beneficiaries receive CFAF 25,000 per quarter for five years and must attend all sensitization workshops offered by the program. The program identifies the eligible beneficiaries through the unique registry. At the end of 2016, the program had 300,000 beneficiaries. 34. Senegal has adopted a strategy aiming to provide health coverage to 75% of the population by the end of This strategy proposes multiple mechanisms for improving healthcare access, including: (i) free healthcare services for children under 5; (ii) free healthcare services for people over 60; (iii) extending health insurance coverage to the informal sector and rural areas; (iv) a private sector health insurance system; and (v) healthcare coverage for public sector employees. To provide coverage to 75% of the population and to extend medical coverage to informal sector and rural workers, the government established the Insurance Branch of Universal Health Coverage in 2014, which provides healthcare to the most vulnerable households by subsidizing health plan enrolment fees. This program covers 100% of the cost of health plans (CFAF 7,000) for beneficiaries of family security grants and CECs and 50% of the cost for others. Only households without current medical coverage are expected to enroll in the program, and the subsidy will not be available to individuals working in the formal sector with private sector health coverage often the wealthiest or to public sector employees. 35. The grants program for higher education began in 1982, and the decree creating it has not been revised since then. Unlike many grants programs throughout the world, this program was not designed to 12

22 promote equity among students, foster excellence, encourage certain training programs, or attract the brightest students. Instead, it was designed as a universal grant available to all full-time students. The amount (full, two thirds, or one half of grant) is determined by performance and level of education. A full grant amounts to CFAF 36,000 per month. As of 2014, over 70% of students received a grant (full or partial). 36. The other program with broad coverage is the community nutrition program. This includes growth monitoring and promotion, malnourishment screening and treatment, activities to combat micronutrient deficiencies, and behavior modification activities. This program covers 400 municipalities (out of 557). 37. Family support is a critical component of social assistance. Family support is one of the most important strategies for withstanding shocks (see Table 5). While it does not receive government funding, it is a vital mechanism of social assistance and one the population sustains on its own in order to compensate for services the government cannot provide. According to the Poverty Monitoring Survey (ESPS) II, 71% of private transfers (national and international remittances) are used for everyday food consumption. As the next section shows, the amount of this type of support can be several times greater than what households spend on social protection. II Programs for Emergencies and Response to Shocks 38. Senegal has several mechanisms and programs in place for responding to the shocks and emergency situations described below. A total of 6 crisis response programs have been identified. Two programs target drought (or rainfall shortages) and food insecurity (for people and livestock). The National Food Insecurity Response Program (PRNIA) is implemented by the Food Security Commission (FSC), the WFP, and NGOs. Three programs target households affected by floods and fires. Two programs target farmers affected by climate shocks. Data on the exact number of beneficiaries are unavailable, and no overarching coordination mechanism is in place for these programs. TABLE 5: CRISIS-RESPONSE PROGRAMS Shock Program Description Ministry and Agency Beneficiaries Trigger Drought and rainfall shortages National Food Insecurity Response Program (PRNIA) Distributes food assistance through the national response program (PNRIA) Food Security Commission (FSC) Populations facing food insecurity identified by the Executive Secretariat of the National Food Security Council (SE/CNSA) At the discretion of the Prime Minister; no clear trigger thresholds in place 13

23 National Food Insecurity Response Program (PRNIA) Distributes food assistance through the national response program (PNRIA) World Food Program Populations At discretion of facing food the Prime insecurity Minister; no clear identified by trigger thresholds SEA/CNSA or HEA in place Livestock Protection Program (OSB) Subsidy covering 50% of feed costs Ministry of Livestock Vulnerable livestock; farmer selection system unclear Livestock feed is subsidized in targeted areas from May to July, scaled up or down depending on drought conditions Fires, floods As needed Distributes inkind and financial National support to Solidarity Fund affected (FSN) households Vulnerable households affected by shocks requesting support Ad hoc, based on FSN assessments and official requests from governors Floods As needed Finances flood prevention and flood response (generally pumping and assistance for affected households) Special Fund for Flood Management (FSGI) Households affected by floods, populations in severely flooded areas Official request from governors and emergency responders Agricultural Disaster Fund Debt relief and financial support for farmers National Farm Credit Bank of Senegal (CNCAS) Account holders at CNCAS Unknown Agricultural risks Crop insurance Governmentsubsidized (50%) crop insurance from CNAAS National Crop Insurance Fund of Senegal (CNAAS) Mostly commercial farmers (average farm size: 1 hectare) Index insurance based on climate data (rainfall as measured at weather stations) and traditional insurance Source: World Bank, 2016a 14

24 III Employment Programs 39. In Senegal, social protection programs and actors working on coordinating the sector tend to overlook employment. While SO 2 of the new SNPS aims to create programs and plans for people of working age, it only refers to passive labor programs (unemployment benefits and maternity benefits), not to any intermediation or active labor programs. Senegal s National Employment Strategy ( ) is a cornerstone of the fight against poverty. In fact, the PSE s definition of social protection specifically mentions empowerment. However, in Senegal, few if any direct links exist between employment programs and the social protection sector. 40. What little data were collected about employment programs were obtained primarily through document review. (See Table 6 below for details of the programs.) Since Senegal has no intermediation programs or passive labor market programs, the author did not use ASPIRE categories in this section but instead designed categories that seemed appropriate for classifying Senegal s existing programs. These are: (i) Income-generating activities (IGA) programs for the most vulnerable; (ii) programs supporting entrepreneurship; and (iii) vocational training programs. A total of 6 funds or programs expired between 2010 and 2015, and 8 current programs or funds were identified. The 6 expired programs ended due to dependence on external financial resources or the merger of multiple funds targeting youth into a single fund in order to improve efficiency. This review does not cover all programs. Collecting data about these programs was difficult due to their being small, fragmented, poorly represented in terms of social protection, or heavily dependent on external funding. TABLE 6: EMPLOYMENT PROGRAMS Program Description Start date End date (if any) Income-generating activities (IGA) programs for the most vulnerable Implementing agency 1 Support project for the elderly (PAPA) Promotes IGAs for vulnerable elderly persons (60+) through training, microloans, and subsidies; also supports groups of individuals 2007 Ministry of Health and Social Action General Directorate of Social Action 2 Poverty reduction program (PRP) to support implementation of the poverty strategy Grants and micro-loans for poor people Ministry of the Family Poverty Monitoring Committee 3 Economic and Social Development Combats poverty; supports the creation Dynamics Support of IGAs for the poorest populations Program (PRODES) 2013 Ministry of the Family Poverty Monitoring Committee 15

25 4 Program to empower persons and families affected by leprosy Improves living conditions of persons affected by leprosy equipment transfers, micro-loans, IGAs 2010 Ministry of Health and Social Action General Directorate of Social Action 5 Community-based re-adaptation program (PRBC) Social and economic integration of disabled persons via training and funding for IGAs 2006 Ministry of Health and Social Action General Directorate of Social Action Entrepreneurship programs 6 Private sector support platform (PLASEPRI) Promotes participation in the private sector by Senegalese immigrants so they can support their country s economic development Ministry of the Family 7 Financing for young peoples business ideas Finances start-up activities 2013 National Agency for Youth Employment (ANPEJ) 8 Program for the Promotes the hiring of young people in professional insertion the formal sector of young people 2014 National Agency for Youth Employment (ANPEJ) 9 National fund for the employment of young people Finances self-employment projects through a national government-employer agreement 2000 National Agency for Youth Employment (ANPEJ) 10 National fund for the promotion of young people Finances business projects of persons aged without requiring a personal contribution or guarantee 2013 Ministry of Employment and Young People 11 National fund for women entrepreneurs Promotes the training and strengthens the capacities of current and potential businesswomen; Provides support for preparing business plans; Finances projects by women entrepreneurs or led by women; Monitors projects; Guarantees loans from savings and loan associations 2004 Ministry of the Family 12 Agency for youth and Promotes employability, suburban entrepreneurship, and self-employment employment Ministry of Employment and Young People 16

26 13 National Agency for Youth Employment (ANPEJ) Provides information, consulting, and support for the creation of livelihoods Ministry of Employment and Young People 14 Financing support for entrepreneurship 2013 Ministry of Finance 15 Program for community agricultural areas 2013 Ministry of Youth and Recreation Vocational training programs Targets the development of high-quality Support program for vocational and technical training equally vocational training accessible to males and females and meets and professional the needs of social and economic insertion development 2010 Office for the training Promotes employability, entrepreneurship, and employment of and self-employment suburban youth Ministry of Vocational Training, Apprenticeships, and Crafts Ministry of Employment and Young People Source: Author, compilation of collected data 41. The majority of current programs are active labor market programs. These programs primarily target young people, women, and farmers, while 4 specifically target vulnerable persons (elderly, disabled, poor, etc.), but are quite narrow. Only 1 current program consists of a labor market intermediation program. This new program is managed by ANPEJ. At present, its coverage and budget are very limited. The other program targets vocational training for young people and is financed almost exclusively by technical and financial partners. Programs targeting labor-intensive jobs are still underdeveloped. There is currently no passive labor market program (unemployment insurance or early retirement). Current employment programs are implemented by 6 different ministries or agencies. Their annual budgets range from a few thousand to several billion CFAF. 42. Some Senegalese actors view programs classified in this review as employment programs as social assistance programs since they target vulnerable populations. However, this review opted to classify them as employment programs given their objectives (support for entrepreneurship or microproject financing). In most cases, these programs are implemented by the Ministry of the Family or the Ministry of Health and Social Action. IV Social Insurance Programs 43. Social insurance programs include pensions, medical insurance, and family services. Table 7 lists the social insurance programs included in this review. 17

27 TABLE 7: SOCIAL INSURANCE PROGRAMS Program Description Start date Implementing agency Retirement Savings Fund of Senegal (IPRES) National Pension Fund (FNR) Workers compensation and work-related illness (CSS) Pension fund for private sector employees Pension fund for public sector employees Insures private sector employees against work-related accidents 1977 IPRES Ministry of Finance National Pension Fund (FNR) Social Security Fund (CSS) 4 Family benefits (CSS) Families contributing to the CSS can receive CFAF 7,800 per quarter per child (maximum of 6 children) 1975 Social Security Fund (CSS) 5 Health and social welfare services 6 Maternity leave Source: Author, compilation of collected data Health services for private sector employees Maternity leave for private sector employees Social Security Fund (CSS) Social Security Fund (CSS) 44. The Senegalese pension system includes two mandatory schemes: i) the Retirement Savings Fund of Senegal (IPRES), which covers private sector employees, government employees without civil servant status, and local government employees, and ii) the National Pension Fund (FNR), which covers public sector employees and members of the armed forces. IPRES is an autonomous institution operating under the joint technical and financial authority of the Ministry of the Civil Service, Labor, and Institutional Relations and the Ministry of the Economy, Finance, and Planning (MEFP). The FNR is a special treasury fund. Administrative management is provided by the Directorate of Salaries, Pensions, and Annuities (DSPRV) with financial management provided by the General Directorate of Public Accounting and the Treasury (DGCPT). The mandate of the DSPRV is to authorize pension-related expenditures. Pension payments are made by the Public Treasury. The National Pension Fund does not have the authority to independently invest surpluses. However, the government does provide financing in the event of shortfalls. TABLE 8: DESCRIPTION OF PENSION SCHEMES Parameters IPRES FNR Pension type Defined benefits scheme based on a Benefits determined on a pay-asyou-go points system basis Contribution rate General scheme: 14% Supplemental scheme: 6% 35% (12% by employee and 23% by employer) (employee pays 40% and employer pays 60%) Retirement age 60 years of age (55 for certain jobs) and early retirement at (55 for the armed forces) 18

28 Accrual rate 11 Basis for determining pensions Minimum years required to qualify for pension rights Pension amount calculated by multiplying the number of points earned by the base point value First 36 months of contributions At least 1 year and at least 400 points (15 years for full pension) 2% in 2002 and 1.8% since 2002 per year of contribution Average salary during last 3 years of work 30 years (25 for the armed forces) Source: Author, compilation of collected data 45. IPRES is a defined benefits pension system in which contributions by current workers fund the pensions of retired people. The system is based on points. In a points system, a person earns points for each year of contributions. The number of points earned each year is determined by three factors: the salary from which contributions are deducted, average salary, and the length of time during the year the person contributed to the system. Upon retirement, all points earned are added up and multiplied by the value of the base point. The contribution rate for the general IPRES pension scheme is 14% of salary, 8.4% of which is paid by employers and 5.6% by employees. The general IPRES pension scheme covers private sector employees, government employees without civil servant status, seasonal workers, and day laborers. The Supplemental Scheme offers additional coverage to senior-level employees. The pension system covers old age, disability, and survivors benefits for eligible beneficiaries. Conditions for receiving retirement benefits are having reached the retirement age of 60 (with early retirement available at 53) and having accumulated a minimum of 400 points. Survivors pensions total 50% of the amount of the retirement benefit or the disability pension the deceased person was receiving or was entitled to receive. A survivor s pension paid to a widow between 45 and 49 years of age without any dependent children is reduced by 5% yearly until she reaches 50 years of age. If the pensioner designated more than one widow to receive survivors benefits, the pension is divided among them equally. 46. FNR is a defined contribution pension scheme for public sector employees. To receive a pension, claimants must be 60 years of age (55 for the armed forces) and have worked for at least 30 years (25 for the armed forces). The system has a high contribution rate and a relatively high accrual rate. Generally speaking, a 1.8% accrual rate per year with 30 years of service amounts to an income replacement rate of 54% upon retirement. In contrast, the IPRES system has an accrual rate of only 1.3% per year. Pension benefits are determined by the average salary during the last three years of work. Although not unusual in the region, calculating pension benefits on the basis of salary earned during the last three years of work instead of contributions earlier in employees' careers when they earned a lower salary can prove costly for the pension system. 47. For public sector employees, health insurance is provided by the government. This mandatory scheme for government employees with or without civil servant status and their families was set up by Decree of March 7, 1972 relative to the social security coverage of civil servants. This system is government-funded as a budget item. The government covers all transport costs related to medical evacuations. The costs of hospitalizations and treatment services are shared between the government 11 Source: 19

29 and the patient (80%/20% respectively). This system covers roughly 300,000 beneficiaries (66,000 employees and their dependents). However, financial data about this type of insurance plan could not be collected. Total health care expenditure would have been higher if this expenditure had not been omitted. In fact, this expenditure accounts for the majority of government health insurance spending. 48. Private sector employees must be enrolled in the Social Security Fund (CSS). The CSS administers family benefits and workers compensation. Contributions are paid by employers (7% for family benefits and 1% 5% for workers compensation). Family benefits include maternity leave and child benefits. Maternity leave is 14 weeks, and maternity pay is based on daily wages. Family benefits cover: - Prenatal benefits: Quarterly cash transfers to households during pregnancy (CFAF 4,500 9,000); - Maternity benefits: Quarterly cash transfers from the child s birth until age 24 months (CFAF 6,750 13,000); - Child benefits: Cash transfers paid quarterly for each child in the household aged 2 21 years of age (up to 6 children) (CFAF 7,800 per child). - Health and social services include healthcare for mothers and children at two facilities (Guédiawaye and Kolda). The cost of these services is negligible, as is the number of beneficiaries. 20

30 Chapter Social Protection Expenditure I Overview of Social Protection Sector 50. For the purpose of this review, total expenditure consists of expenditure funded directly by the finance bill, social security fund expenditure funded in full by contributions or financially independent, and expenditure funded from external resources. Unless otherwise indicated, all expenditure are given in nominal value. 51. Social protection expenditure 12 increased for the period. Excluding 2011, total expenditure on social protection rose every year. Total expenditure in 2015 was 38% higher than in Among social protection expenditure, employment programs saw the largest gains between 2010 and (Note, however, that the percentage increase is not informative since expenditure levels were highly uneven.) Social assistance expenditure increased by 38%, and social insurance expenditure increased by 31% over the period. TABLE 9: TOTAL EXPENDITURE ON SOCIAL PROTECTION, IN MILLIONS OF CFAF change (% in nominal terms) Social assistance 58,749 52,991 60,686 79,105 90,996 80,916 38% Crisis response 10,016 7,381 20,673 14,308 14,346 11,715 17% Employment programs 5,287 4,885 5,998 11,583 8,385 18, % Social insurance 129, , , , , ,806 31% Total social protection expenditure 203, , , , , ,709 38% Total social protection expenditure, in real terms 155, , , , , ,364 32% Source: Author, compilation of collected data 52. Spending on social protection accounted for a small share of government expenditure compared to spending on other social services (Figure 2). Total expenditure on education varied between 20% and 25% of total government spending, whereas expenditure on healthcare varied between 10% and 12%. On average, total expenditure on social protection was only half the amount (52%) spent on education over the same period. 13 As a share of total government spending, expenditure on social insurance (7.2%) 12 Figures for expenditure on IPRES pensions and social security (CSS) were not available for Expenditure from 2014 were used to keep the same order of magnitude. The number of beneficiaries was also taken from 2014 for the same reasons. 13 Expenditure on social protection programs overseen by the Ministry of Health and the Ministry of Education, such as the CMU and school meals were extracted from total expenditure on healthcare and education. 21

31 exceeded expenditure on healthcare (6.1%) on average for the period. However, expenditure on healthcare exceeded expenditure on social assistance, crisis response, and employment combined, which accounted for only 4.4% of total government spending on average for the period. Spending on various social services and programs trended together, except in 2014, when spending on education grew sharply. FIGURE 2: EXPENDITURE ON SOCIAL SERVICES SECTORS, IN MILLIONS OF CFAF AND AS SHARE OF GOVERNMENT S TOTAL SOCIAL EXPENDITURE Source: Author, based on collected data 22

32 53. In comparison to cash transfer programs made within families (remittances), 14 expenditure on social protection was negligible (Figure 3). with total social protection expenditure equivalent to 30% of total remittances between 2010 and According to ESPS data, in 2011, remittances totaled CFAF 346 billion, 16 or 170% more than total government expenditure on social protection. Government expenditure on social assistance was equal to around 10% of total remittances. FIGURE 3: COMPARING REMITTANCES, CASH TRANSFER PROGRAMS, AND TOTAL SOCIAL PROTECTION EXPENDITURE Source: Author, compilation of data from the Bilateral Remittances Matrix, World Bank, using ESPS II and data collected for the review 54. Over the period, expenditure on various categories of social protection as a share of total expenditure remained stable (Figure 4). Social insurance programs made up the largest share of social protection spending, accounting for 59 67% for the entire period. Social assistance programs share of expenditure varied between 25% and 32% of total social protection expenditure, while the share of employment programs was rather small and uneven over the period. This was also true of crisis response programs and food security programs, which varied between 3% and 8% of total expenditure for the period. FIGURE 4: EXPENDITURE AS SHARE OF TOTAL, BY TYPE OF SOCIAL PROTECTION Source: Author, compilation of collected data 14 Remittances and national transfer payments. 15 Calculations of remittances are based on methods developed by Ratha and Shaw, 2007, South-South Migration and Remittances, Development Prospects Group, World Bank. Figures for transfer payments were disaggregated by analyzing flows of recipient and remitting countries as well as the estimated number of migrants. 16 It is likely that undervaluation of these data occurred as a result of constraints in data collection. 23

33 55. Expenditure on social protection as a share of GDP grew little during the period (Figure 5). Depending on the year, increases totaled % of GDP. However, if expenditure related to social insurance programs is subtracted, expenditure on social protection spending accounted for no more than 1.5% of GDP. FIGURE 5: TOTAL EXPENDITURE ON SOCIAL PROTECTION AS SHARE OF GDP Source: Author, compilation of collected data 56. On average, Senegal outspends other African countries on social protection relative to GDP, mainly as a result of high expenditure levels on social insurance (Figure 6). Expenditure on social protection considered in this review consists of spending on social safety nets (including crisis response programs) and social insurance and employment programs. Data for these types of expenditure are generally incomplete. A review of those countries with sufficient data revealed that their governments spent 1 8% of their GDP on social protection, averaging at 2.75%. Allocating 3.4% of its GDP to social protection, Senegal FIGURE 6: AVERAGE EXPENDITURE ON THREE TYPES OF SOCIAL PROTECTION AS SHARE OF GDP spends somewhat more BY COUNTRY than the average. This is mainly due to Senegal s expenditure on social insurance, which, measured at 2.08% as a share of GDP, is relatively high compared with other countries, which spend on average 1.07%. Only Mauritius spends more relative to GDP than Senegal. As for expenditure on social assistance, Senegal spent less than the average (1.62% of GDP), with expenditure of just Source: World Bank (2017a) 1.14% of GDP. 24

34 57. Social protection expenditure was funded from three sources: the government s budget, external resources (from development partners DP), and beneficiary contributions. Beneficiary contributions include all contributions from employers and employees as well as the credit balance of all private sector insurance providers. The share of expenditure funded through the contribution system fell slightly over the period, from almost 37% to 32% of total expenditure on social protection. The share of expenditure funded from external resources varied over the period (between 6.4% and 10.4%) but fell to its lowest level in 2015 to 6.4% (Figure 7). 58. However, the share of funding from beneficiary contributions and external resources did not shrink because of lower funding levels but rather as a result of an overall increase in social protection expenditure that was paid for by the government. The low level of expenditure funded from external resources can also be explained in part by a number of difficulties in data collection. 17 Additionally, the review focused specifically on programs managed directly by the government. FIGURE 7: TOTAL SOCIAL PROTECTION EXPENDITURE, BY FUNDING SOURCE Source: Author, compilation of collected data 59. It should be stressed that considerable funds are spent by international and non-governmental organizations (NGOs) on directly implementing actions that meet the government s social protection objectives. However, since the scope of these programs is too broad, this review did not include this funding. For information, OCHA estimates that an average of USD 25 million (approximately CFAF 12 billion) was spent on humanitarian assistance in Senegal for the period under review (with estimates ranging from USD 5 million to USD 40 million per year). Additionally, over CFAF 2.3 billion was spent by NGOs under the National Food Insecurity Response Plan (PRNIA) in 2015 (Groupe sectoriel sécurité alimentaire, 2015). 60. External resources were directed mainly at professional training programs, food insecurity response, and school feeding programs (Figure 8). External resources paid for % of these programs total expenditure, making them highly funding-dependent. Meanwhile, the share of external resources directed at subsidy programs promoting access to healthcare and at nutrition programs fell during the period. This is explained mainly by additional government funding counterbalanced by a slight reduction in external resources. 18 Programs with large budgets received little or no external resources. 17 Funding from development partners (DP) does not always appear in the annual reports of agencies and ministries and is not included in SIGFIP. These funds are managed separately by the Investment Department of the Ministry of Economy, Finance, and Planning (MEFP). 18 It should be noted that data collected on external resources directed at the CMU were incomplete. 25

35 This included cash transfer programs 19 and student grants programs. FIGURE 8: CHANGE IN EXTERNAL FUNDING AS SHARE OF TOTAL EXPENDITURE, BY PROGRAM TYPE Source: Author, compilation of collected data II Rising Expenditure on Social Assistance Programs 61. Since 2013, expenditure on social assistance has accounted for 30% of total spending on social services and for almost 1% of GDP. Total expenditure on social assistance programs rose sharply in 2013 and Although expenditure fell in 2015, it remained 38% higher than in Expenditure then contracted in 2015 as a result of a sharp decline in expenditure on student grants (-32% compared with 2014 spending). 62. However, these figures should be qualified by the generous share of student grants as a proportion of total expenditure. A review of expenditure by program type (Figure 9) reveals that three program types accounted for the bulk of these expenditure, with the lion s share allocated to student grants, which accounted for more than half of all social assistance expenditure. Healthcare subsidies and cash cash transfer programs each accounted for 20% of total expenditure on social assistance. The share of healthcare subsidies and cash transfer programs expanded considerably during the period (Ministry of Health, 2016a, 2016b). This matched the government s new policy goal of implementing a national cash transfer program (National Family Security Grants Program PNBSF and universal health coverage). The remaining programs received only 5% of total expenditure on social assistance. 19 The PNBSF is supported by funding from the World Bank. However, since funding began only in October 2014, total funding for 2015 was relatively low. 26

36 FIGURE 9: TOTAL EXPENDITURE ON SOCIAL ASSISTANCE PROGRAMS Source: Author, compilation of collected data 63. Upon closer inspection, if expenditure on student grants is excluded, expenditure on social assistance has only grown since 2013 and more than doubled since Expenditure on nutrition programs and children s social services have remained constant, while expenditure on cash transfer programs and targeted subsidies to improve healthcare access rose sharply. In 2015, cash transfer programs accounted for 37% of total social assistance expenditure (excluding student grants) compared with 4.7% in Although expenditure on targeted subsidies aiming to improve healthcare access increased during the period, its growth of total social assistance share of expenditure (excluding student grants) was far less impressive, varying between 36% and 48% for the period. Meanwhile, total spending on school feeding programs fell markedly during the period. This was related in part to lower World Food Program (WFP) funding resulting from a transfer of all activities to the government. This transfer has not yet been accompanied by any large increases in budget allocated to this program by the government. 27

37 FIGURE 10: TOTAL EXPENDITURE ON SOCIAL ASSISTANCE PROGRAMS, EXCLUDING STUDENT GRANTS Source: Author, compilation of collected data 64. Compared with other African countries, Senegal spends the most on student grants and targeted healthcare subsidies relative to total social assistance expenditure (Figure 11). The data reveal that countries generally spend the bulk of social assistance expenditure on two or three types of programs only. Among all countries, the most common social assistance programs are school grants, targeted subsidies, cash transfer programs, social pensions, and labor-intensive employment programs. Senegal does not have a social pension program or a labor-intensive employment program in its social assistance portfolio. 28

38 FIGURE 11: EXPENDITURE ON SOCIAL ASSISTANCE PROGRAMS AS SHARE OF TOTAL SOCIAL ASSISTANCE SPENDING IN AFRICA Source: Author, compilation of ASPIRE data based on most recent data available (data collected between 2010 and 2016) FIGURE 12: TOTAL SOCIAL ASSISTANCE EXPENDITURE BY FUNDING SOURCE 65. Based on collected data, external funding for social assistance expenditure fell as a share of total funding (Figure 12). This is explained in part by the introduction of two major social assistance programs: the insurance branch of the CMU, and the PNBSF (both funded by the government), and in part by a decrease in external resources for the period. 20 Source: Author, based on collected data 20 Note once again the limitations of data collection on external funding. For example, data on external financing for the insurance branch of the CMU were incomplete and represented several billion CFAF. 29

39 Expenditures on Subsidies Were Higher than Expenditures on Social Assistance Programs 66. It is interesting to note that while there was an increase in total social assistance expenditure (excluding university grants) between 2011 and 2015, expenditure on subsidies shrank considerably. In this review, subsidies include: 1) subsidies directed at consumers, such as subsidies to agricultural inputs and equipment, university student services, 21 fuel, and food; and 2) subsidies directed at producers, including liquefied petroleum gas (Société Africaine de Raffinage SAR) and electricity (SENELEC). FIGURE 13: COMPARISON OF EXPENDITURE ON SUBSIDIES AND SOCIAL ASSISTANCE, EXCLUDING HIGHER EDUCATION GRANTS Source: Author, compilation of collected data and International Monetary Fund (2012, 2016, 2017) data 67. The drastic reduction in subsidies to SENELEC was not due to a change in policy but rather to market changes, specifically the steep drop in the price of oil. Currently, 90% of SENELEC s power is generated from oil. Although Senegal is committed to pursuing reform to diversify its energy sources and to reduce its dependence on fossil fuels by 2030, an increase in the price of oil in the coming years will necessarily imply a large increase in subsidies. This could have a negative impact on the budget available for social assistance programs. 68. Subsidies for agricultural inputs and equipment accounted for the major share of total subsidies in 2015 and nearly equaled the amount spent on social assistance programs, excluding student grants. These were the only subsidies whose value did not fall during the period. III Funding Sources and Expenditure on Crisis Response Programs Most Spending on Crisis Response Programs Was Directed toward Drought and Rainfall Shortages 69. Total expenditure on crisis response programs varied between 4% and 8% of total expenditure on social protection for the period. Total spending was highly variable and depended on the severity of the crisis. Total expenditure in 2012 almost tripled in comparison with 2011 as 2012 was an especially difficult year. 21 University student services include subsides directed at universities to provide students with reasonable housing near the university, balanced meals, and access to healthcare. 30

40 70. Since 2013, food insecurity response programs 22 have accounted for around three quarters of spending on crisis response (Figure 14). Despite the fact that these programs accounted for the bulk of total expenditure, the government allocated less than 15% of the total estimated cost to food insecurity crisis response in It should also be noted that a large share of expenditure on food insecurity response programs is covered by NGOs, which reported having transferred over CFAF 2.3 billion 24 to foodinsecure households in 2015 alone (Groupe Sectoriel Sécurité Alimentaire Sénégal, 2016). FIGURE 14: TOTAL EXPENDITURE ON CRISIS RESPONSE PROGRAMS Source: Author, based on collected data 71. The response to agricultural production shocks appears to be very poorly funded. While shocks to agricultural production were not quantified in this review, over the six years under review, agricultural disaster funds received only an infusion from the government in It was not possible to evaluate the extent to which expenditure on responding to other crises (floods, fire, and others) covered needs since the number of people affected by these shocks was not estimated. It is nevertheless possible to consider the reduction in flood risk (and therefore the corresponding cost of damage and response) resulting from infrastructure-building policies that have been in place since 2012 to improve storm runoff management and prevent flooding in Dakar. African Risk Capacity: A New Funding Mechanism for Responding to Drought 73. Between 2010 and 2013, the finance bill included an expenditure line labeled Emergency Relief Funds for the Ministry of the Economy, Finance, and Planning (MEFP) and which was intended for emergency responses and for activating the emergency services management plan (ORSEC). The amount allocated to the Emergency Relief Funds varied between CFAF 1 billion and 5 billion. Funds were used only in For other years, the supplementary finance bill reallocated these funds to other expenditure. 22 Spending related to actions aiming to save livestock was not included in this total. 23 The total cost of food aid was estimated at USD 32 million in This figure is taken from the 2015 Group Sectoriel Sécurité Alimentaire Report. The transfers, which were funded by the WFP, were deducted from the total amount since this expenditure was already accounted for in this review. 31

41 Funds may be left unused as a result of lack of an emergency but also of lack of clear trigger mechanism for releasing these funds. 74. Starting in 2014, this expenditure line no longer appears in the budget. However, the government signed an agreement to become a member of the African Risk Capacity (ARC) initiative. In 2014, Japan paid for Senegal s insurance premium, at a cost of CFAF 1.8 billion. 25 Following poor crop yields in 2014, ARC indemnified Senegal for CFAF billion. These funds were used as part of the national food insecurity relief response and in operations designed to save livestock. Since 2015, Senegal s finance bill law has included an expenditure line for the payment of this insurance premium. 75. In addition to ARC, the MEFP has at its disposal limited resources with which to fund responses to exceptional emergencies and disasters. These resources include: - Food security stocks; - A management reserve totaling 5% of the government s budget but which is not specifically earmarked for emergency and disaster response; - Budgetary reallocation. The MEFP can cancel certain budget lines that have not yet been spent and reallocate the funds to emergency and disaster response. However, this process is lengthy and disruptive and requires approval from the President. IV Very Low Expenditure on Employment Programs Total Expenditures on Employment Programs Varied Considerably during the Period but Were Uniformly Low, Accounting for 2%-7% of Total Social Protection Expenditure 76. The majority of employment programs had very small budgets. In 2015, only 6 of the 13 programs reviewed had total annual expenditure greater than or equal to CFAF 1 billion. Four of these programs had a credit balance. All the other programs had annual expenditure of CFAF million per program on average. 25 USD 3.6 million. 32

42 77. Total expenditure increased markedly during the years in which a credit line was opened for use with entrepreneurship programs (for example, the PLASEPRI program in 2013) and for financing entrepreneurship in Total expenditure on programs designed to develop income-generating activities (IGA) among vulnerable groups varied over the period but never exceeded CFAF 2.5 billion per year. FIGURE 25: TOTAL EXPENDITURE ON EMPLOYMENT PROGRAMS, IN MILLIONS OF CFAF Source: Author, based on collected data V. Expenditure on Social Insurance Programs 78. Total expenditure FIGURE 16: EXPENDITURE ON SOCIAL INSURANCE, IN MILLIONS OF CFAF on social insurance programs accounted for over 60% of total social protection spending for the period. Expenditure on the pensions system accounted for nearly 90% of total expenditure on social insurance programs. This was constant over the period. Between 2010 and 2015, Source: Author, based on collected data expenditure on social insurance programs increased by approximately 23% for the period. The FNR saw the largest increases (+39%), while maternity leave expenditure saw the largest decrease (-20%). 33

43 79. Expenditure on the pension system alone absorbed 2% of FIGURE 17: EXPENDITURE ON PENSIONS AS SHARE OF GDP GDP. This amount was relatively stable during the past five years, and expenditure on private sector pensions was nearly equivalent to that for government employees. Funding of the national pension fund was indirectly subsidized by the government through a hefty contribution rate (23%) to the system. Additionally, it is important to note that for the period, neither FNR nor IPRES received direct funding Source: Author, based on collected data from the government for pensions. However, this situation, which is detailed in Chapter 5, may change very soon. 34

44 Chapter 4 Analysis of the Coverage of Social Benefits I A Life Cycle Approach to Social Protection Programs 80. Programs targeting only a few hundred beneficiaries have very low coverage and are not included in this section. Preference was given to programs that reach a critical mass of beneficiaries. Only programs operational in 2015 were considered. Subsidies for agriculture were included since they account for a large proportion of social benefits. 81. The coverage of social protection programs is assessed through the life cycle approach. Table 10 categorizes the various programs by social protection objective and target group. 35

45 TABLE 10: SOCIAL PROTECTION PROGRAMS THROUGH THE LIFE CYCLE (Source: Author, collected data) Social protection objectives Pregnant women and children under 5 Free healthcare for children under 5 School-age children People of working age Elderly people Free cesarean sections School feeding programs University grants SESAME Plan Social assistance Emergency programs Employment programs Social insurance Community nutrition program Early childhood education centers Maternity leave Social services for university students National Family Security Grants Program (PNBSF) Equal Opportunity Cards (CEC) HIV/AIDS Program Universal health coverage Disaster fund National Food Insecurity Response Plan (WFP and FSC) Flood fund National Solidarity Fund (FSN) Agricultural insurance Subsidies for inputs and farm equipment PRODES ANPEJ Support for entrepreneurs PRODAC Financing for young entrepreneurs Program for women entrepreneurs Workers' compensation (accidents and illness) PAPA Program National Retirement Fund Family benefits Family benefits IPM IPRES 38

46 II Cross-Cutting Programs Have Highly Unequal Coverage 82. Some social assistance programs and nearly all programs for responding to shocks target households or individuals on the basis of specific need. The three-main cross-cutting social assistance programs are: (a) the National Family Security Grants (PNBSF); (b) the insurance arm of the CMU; and (c) Equal Opportunity Cards (CEC). 83. The National Family Security Grants Program (PNBSF) is unique in that it responds to a wide range of needs. This program targets all households living in extreme poverty in Senegal (roughly 14% of the population). The program s target population is around 3.5 million. In 2015, the PNBSF nearly reached 2 million. It reached its final target at the end of 2016 with households enrolled. It is now working on addressing some of the inclusion and exclusion errors inherent to the targeting process. 84. As of 2015, coverage of the insurance arm of the CMU (subsidy for healthcare plan enrolment) remained quite low at roughly 7% of the population (excluding children under 5 and people over 60). However, this is due to the program starting only recently and to efforts being focused on creating community-based healthcare plans, a process that needs to occur prior to enrolling new beneficiaries. In December 2016, there were 671 healthcare plans throughout the country (or more than one per local administrative area) and beneficiaries of community based healthcare plans. 85. The coverage objectives of the Equal Opportunity Cards (CEC) remain well below identified needs. Started only recently, this program targets all people with disabilities in Senegal. While coverage remains low (with only 2,021 beneficiaries in 2015), the aim is to speed up enrolment to reach 50,000 people in However, this number still falls short of needs, accounting for only 21% of the entire targeted population. In the 2013 general census, 230,000 individuals in Senegal, or 1.7% of the population, reported suffering from a serious disability. 86. While multiple shock-response mechanisms exist, responses rarely achieve adequate coverage in practice. Generally speaking, programs for responding to shocks and emergencies are assessed based on the amount of material or food distributed rather than the number of people covered. This is the case of the FSC, which only takes into account the number of tons of food distributed. Until recently, this was also the case of the FSN. This makes it difficult to assess either the coverage of needs or the efficiency of the response. 87. The National Food Insecurity Response Program (PRNIA) relies heavily on external actors to reach beneficiaries. In 2015, the total number of individuals identified throughout the country as needing support was 927,416. NGOs and development partners (DP) reached 263,645 people under the portion of the PRNIA program for which they are responsible (Groupe Sectoriel Sécurité Alimentaire Sénégal, 2016). Data on the number of beneficiaries of the government response in 2015 are not available.

47 TABLE 11: COVERAGE OF CROSS-CUTTING PROGRAMS 26 Age group / risk group Social protection program Target group Total population 13,508,715 Total population aged 6 59) CMU insurance branch 10,748,315 NA People living in extreme poverty PNBSF 1,958,764 NA People with HIV/AIDS HIV/AIDS program 39,000 People with serious disability Number of people facing food insecurity CEC National Food Insecurity Response Program (PRNIA) Food Insecurity Response Plan (WFP) 21,028 15,771 13,716 18,436 54% 40% 35% 47% 792,985 7% 490, ,000 1,977,510 25% 48% 101% 235,965 2, ,416 (2015) NA NA 0.86% 54,574 39, , , ,690 85,855 People affected by floods Flood fund NA NA NA 9% People affected by shocks (all kinds) National Solidarity Fund (NSF) NA NA 1,167 2,665 Source: Author, compilation of data from Chapter 1 and data collected about programs 26 The target group column in Tables is based as far as possible on data taken from the 2013 general census of the population (ANSD, 2013) 40

48 III Healthcare Coverage Is High among Children under 5 and Pregnant Women While Other Groups Are Neglected 88. The number of beneficiaries of healthcare programs for children under 5 has been increasing steadily since Healthcare coverage is high among children under 5. According to the statistics calculated based on the data collected, nearly 100% of children are covered. However, this figure does not reflect the fact that some children receive four types of services (vaccination, consultation, medication, and hospitalization) and are thus counted four times. The report on CMU performance in 2015 cited a coverage rate for children under 5 of 59%; however, the calculation method was not specified. The community nutrition program also has very good coverage, with 74% of children under 5 receiving community nutrition services, and 68% of malnourished children received treatment under the PRNIA. 89. The number of pregnant women with access to healthcare has been rising steadily since 2010, which has lowered maternal mortality rates, even though the rate remains high (392 deaths per 100,000 births). The coverage of the program offering free cesarean sections has increased sharply in recent years, reaching 80% in Coverage of specialized services for children under 5 is very low. The community daycare program was started in 2013 as a way to provide childcare services to working women on low incomes. The first four community daycares opened in While this program meets a real need, its implementation has been slow and its coverage remains very low despite receiving substantial resources. While the Early Childhood Education (Case des Tout-Petits) program has been in operation since 2002 and its beneficiaries have increased by 50% since 2010, its coverage remains low (roughly 15% of children aged 3 5) despite substantial needs in the area of early childhood education. 91. Maternity leave coverage remains far too low. Maternity leave only targets employed women who contribute to the Social Security Fund (CSS), which constitutes a very low percentage of the population. Fewer than 3,500 women working in the private sector benefit from maternity leave coverage, while women working in the informal sector receive no social protection during pregnancy.

49 TABLE 12: COVERAGE OF PROGRAMS TARGETING CHILDREN UNDER 5 AND PREGNANT WOMEN Target population Number of beneficiaries Pregnant women 147, Pregnant women in the workforce Maternity leave 56,583 Women needing a cesarean section Free cesareans 22,152 Children under 5 Children under 5 suffering from chronic malnourishment Children aged 3 to 6 Children under 5 1,992,857 Free healthcare PRNIA community nutrition service PRNIA treatment for chronic malnourishment Early Childhood Education (Case des tout-petits) 1,992,857 2,936 3,099 3,376 3,416 5% 5% 6% 6% 589 NA 6,705 12,066 17,961 3% NA 30% 54% 81% ,269,059 2,016,765 64% 101% 888,205 1,136,405 1,364,819 1,542,219 1,602,726 1,475,147 45% 57% 68% 77% 80% 74% 117, , , , , ,793 80,475 88% 107% 133% 140% 104% 68% 996, , , , , , ,760 10% 11% 11% 13% 14% 15% Primary and middle schoolchildren Children aged ,766,962 School feeding programs (DCAs) School feeding programs (WFP) 2,223,149 Children under 19 Family benefits 7,272, , , , , , ,706 11% 12% 12% 13% 14% 16% 564, , , , , ,738 25% 25% 22% 22% 22% 14% 439, , , , ,545 6% 7% 7% 8% 7% ND Source: Author, compilation of data from Chapter 1 and data collected about programs 42

50 IV With the Exception of Coverage Provided by Cross-Cutting Programs, the Coverage of Programs Targeting School-Age Children Is Low and Decreasing 92. Very few programs target children between 5 and 15 years of age. School-age children are the group least targeted by large scale social protection programs as this population benefits only from food assistance at school, healthcare insurance through their parents plans, and family benefits. While many families receive family benefits, coverage is in fact low because only families with parents contributing to the Social Security Fund (CSS) are eligible to receive them. While family benefits cover all children up to 21 years of age in targeted households, fewer than 10% of all children in Senegal are covered by them A limited amount of data was collected on the number of beneficiaries of assistance programs aimed at vulnerable children. However, the low level of spending on such programs suggests that they have very low coverage. Given the risks and the degree of vulnerability facing children in Senegal (see Chapter 1), the shortage of benefits for this group is a major weakness in the government s social protection system. 94. The coverage of the school feeding programs is declining. The school feeding program is financed and implemented jointly by the government and external actors FIGURE 18: BENEFICIARIES OF THE SCHOOL FEEDING PROGRAM (one being the WFP). However, the number of beneficiaries covered by the WFP has steadily declined since 2010 (Figure 18), while the number of beneficiaries of governmentfunded school feeding programs has increased. The total number of beneficiaries of the school feeding programs program was 648,444 in 2015 (or 30% of children between the ages of 5 and 15). This is one of Source: Author, based on collected data the only social assistance programs whose coverage dropped significantly during the period, probably due to funding cuts. FR EN Cantines scolaires (Dcas) School feeding programs (DCAS, government funded) Cantines scolaires (PAM) School feeding programs (WFP) TABLE 13: COVERAGE OF PROGRAMS BENEFITING CHILDREN AGED 5 15 Number of beneficiaries Age group/ risk group Social protection program Target group Children between 5 and 14 3,766,962 Children School feeding 254, , , , , ,706 enrolled in programs (DCAS, 2,223,149 11% 12% 12% 13% 14% 16% 27 See Table 11 in Section II for the exact number of beneficiaries.

51 primary and secondary school government funded) School feeding programs (WFP) Children up to the age of 19 Family benefits 7,272, , , , , , ,738 25% 25% 25% 22% 22% 14% 439, , , , ,818 6% 7% 7% 7% 8% ND ND: No data; Source: author, compilation of data from Chapter 1 and data collected about programs V Numerous Programs Target People of Working Age, but Many Needs Remain Unmet 95. Numerous programs target people of working age (15 59). In assessing the extent to which these programs meet the needs of this age group, it is important to distinguish them from programs targeting students and young people in general, agricultural workers, formal sector workers, the unemployed, and informal sector workers. These categories are not mutually exclusive. It is difficult to assess the coverage of these programs since data are not available for some of them. 96. The only social assistance programs specifically designed for working age people benefit students. Nearly all university students receive social support in the form of housing, food, and healthcare. In addition, in 2014, over 70% of students received a grant. 97. Despite efforts to streamline and strengthen youth employment programs, improvements have yet to be made in terms of the coverage of needs. Created in 2014, the National Agency for Youth Employment (ANPEJ) launched several programs designed to improve youth employment. However, the number of beneficiaries remains extremely low, with only 1,655 projects funded and 281 young people receiving professional insertion support. Coverage is virtually nil since the total number of unemployed young people is 840, While the agency was created recently, it did not materialize from scratch as many agencies and funds for youth employment were established in the 2000s. Yet no program has managed to achieve substantial coverage of the target population. 98. The coverage of programs for farmers is difficult to assess. For example, the coverage of the disaster fund cannot be calculated precisely but appears to be low based on what data are available. Several reports point to the lack of data on the number of beneficiaries of this fund, yet such data have yet to be collected. The disaster fund has been used 7 times since 1998 in response to shocks (droughts or floods affecting rice, grain, or peanut farmers, droughts affecting banana growers, etc.). While this fund was set up to protect crop and livestock farmers alike, it has never been used for livestock farmers. Lastly, the majority of CNCAS customers are commercial or semi-commercial farmers, with very few subsistence farmers covered. As a result, the fund has hardly benefited the poorest farmers. 99. Despite the growing number of beneficiaries of the National Agricultural Insurance Fund of Senegal (CNAAS), coverage remains modest. Created in 2009, CNAAS is the only institution in Senegal offering agricultural insurance. Its membership has risen over the years. However, coverage remains too 28 Given that unemployed people account for 25.7% of the potential workforce and that nearly all unemployed individuals (or 90%) are first-time job seekers, the number of young people in need of professional insertion amounts to 840,000.

52 low, with only about 3% of agricultural households being covered by insurance. In addition, 90% of CNAAS customers are commercial farmers (with average farm size greater than 1 hectare), and subsistence farmers are not covered (with the exception of those benefitting from the WFP s R4 project) People working in the formal sector benefit from social insurance programs, but the coverage of some of these programs remains very limited. Since these individuals constitute the wealthiest group in Senegal, it makes sense that they do not benefit from social assistance programs. However, with regard to the Social Security Fund (CSS), only 1% of formal sector workers are covered by workers compensation insurance. In addition, this proportion has declined steadily since This raises questions about the handling of work-related accidents Employment programs for unemployed people or those living in extreme poverty do exist but fall well short of needs. While programs such as economic grants have an ambitious objective (26,000 beneficiaries in three regions by 2017), they still have a long way to go. Substantial funds have been set aside for microfinance, but data on how these funds have been used are unavailable. It is therefore difficult to assess the coverage of these programs or how effectively they target the poorest populations. TABLE 14: COVERAGE OF PROGRAMS TARGETING PEOPLE OF WORKING AGE Target population Number of beneficiaries Working age population (15 to 59) 3,704, Number of students University grants 143,939 62,000 78,000 80,000 85, , ,632 NA 72% NA Number of women aged in workforce Agricultural households Unemployed population Program for women entrepreneurs 1,257,405 ND Subsidies for inputs and farm equipment Agricultural insurance fund PRODAC PRODES Program to support entrepreneurs 755, ,269 ND ND ND 1,511 2,257 9,751 13,559 19, % 0.3% 1.3% 1.8% 2.6% NA ND ND 1,023 1,094 0% 0% Unemployed women National fund for women entrepreneurs 514,237 ND Unemployed youth Financing program for 844,513 NA 1,655

53 Formal sector workers young entrepreneurs 0.2% Youth insertion program Program for youth training and professional insertion Workers' compensation (accidents and illness) NA ND 295,000 2,660 2,333 2,506 2,191 1, % 0.8% 0.8% 0.7% 0.6% % ND Source: Author, compilation of data from Chapter 1 and data collected about programs 46

54 VI One Quarter of People over 60 (the Target Group) Receive Pension and Health Benefits 102. People over the age of 60 have access to two types of programs: retirement pensions, and healthcare In 2014, 18% of the FIGURE 19: RECIPIENTS OF RETIREMENT AND SURVIVORS PENSIONS population over the age of 60 received retirement benefits from IPRES. IPRES paid the pensions of 219,363 people in Of this group, 135,466 received retirement pensions and 83,897 received survivors pensions. Over the last decade, survivors pensions have accounted for roughly 40% of all IPRES pensions. In terms of coverage of the workforce, Source: Author, based on data collected around 4% of the population of working age (people aged 15 59) contributed to IPRES in Coverage of the elderly population aligns with regional trends. While coverage of the elderly population (in terms of the proportion of people past retirement age receiving pensions) might appear low since only one fifth of the population over 60 is covered, this coverage aligns with regional trends. In fact, considering only national pension systems, coverage in Senegal is at the upper end of the regional spectrum. 29 FIGURE 20: PERCENTAGE OF THE POPULATION OVER 60 RECEIVING BENEFITS UNDER NATIONAL PENSION SYSTEMS Source: Author, based on data about pension systems in Africa; World Bank 29 The national pension system encompasses pension systems that cover private sector employees and, in some cases, government employees not covered under another system. 45

55 105. 5% of the population over the age of 60 receives FNR pensions. In 2014, around 1% of the working age population contributed to the FNR pension scheme. It is important to emphasize this number given FNR s cost, which was roughly 1% of GDP in In the same year, IPRES expenditure amounted to 0.9% of GDP and covered four times as many people over 60. Table 21 indicates the coverage and expenditures of the two schemes. FIGURE 21: FNR COVERAGE AND COST (% OF GDP) 106. Senegal has not reached its goal of ensuring that all people over the age of 60 have health insurance. Individuals in this age group without medical coverage under FNR or IPRES schemes are eligible to receive benefits under the SESAME plan. SESAME covers between 12% and 28% of the population over the age of Source: Author, based on data collected 60. Data are only available for the years 2014 and In addition, as highlighted by the Universal Health Coverage Agency (ACMU), in 2015, many people receiving healthcare benefits under pension schemes also use the SESAME plan, which is perceived as more convenient. A better estimate of SESAME s actual coverage cannot be calculated by adding up pension system members as the benefits of pension scheme members are different from those of non-members. TABLE 15: COVERAGE OF PROGRAMS TARGETING PEOPLE OVER THE AGE OF 60 Target population Number of beneficiaries People over 60 People over 60 without pension People over , Recipients of FNR retirement benefits Recipients of IPRES retirement benefits SESAME Plan PAPA 767, ,108 26,461 28,444 30,382 32,178 34,134 35, % 3.7% 4.0% 4.2% 4.4% 4.7% 119, , , , , % 16.7% 17.3% 16.8% 17.6% ND ND ND 215,000 94, % 15.8% 880 2,050 7,160 8, % 0.3% 1.2% 1.4% Source: Author, compilation of data from Chapter 1 and data collected about the programs 46

56 Chapter 5 Program Efficiency I Social Protection Programs Should Target Poor People as a Matter of Priority Prevalence of Targeting by Category Compared to Poverty-Based Targeting 107. The social protection system primarily uses the category-based targeting method to identify beneficiaries. The targeting methods used by the social protection programs that were active in 2015 were analyzed. 30 This included the following: FIGURE 22: PERCENTAGE OF PROGRAMS, BY TARGETING METHOD geographic targeting, targeting by category (demographic), targeting based on income or level of food insecurity, and a combination of these methods. Income-based targeting includes a range of methods including community targeting, income level, the proxy means test, and calculating food insecurity scores. Approximately 45% of these programs use a combination of at least two targeting methods. The most Source: Author, based on data collected frequently used targeting methodology is category-based targeting as 50% of programs use this methodology alone, and 21% use a combination of geographic and category-based targeting. Only 24% of programs use targeting to evaluate the income or food insecurity level of beneficiaries. Almost all shock response programs use geographic targeting in order to prioritize the targeting of zones where the shock occurred Only a quarter of social protection expenditure (excluding social insurance) targets the poorest populations or those facing food insecurity. An analysis of the financial volume of the programs using the various targeting methods used reveals that three main methods are used: (a) targeting by category; (b) a geographic and category-based targeting combination; and (c) a geographic and income-based targeting combination. The other methods are insignificant in terms of the volume of expenditure. About three quarters of total expenditure by the programs under study target categories of persons (by itself or combined with geographic targeting), and only a quarter target the poorest households or those facing food insecurity. 30 Social security programs were not included as they are not targeted. All contributors to the system are beneficiaries. 47

57 109. Exclusive use of the category-based targeting method is problematic because not all households with vulnerable members are poor. Although the poverty rate is higher among vulnerable groups, it does not exceed 50% of the targeted group. For example, close to half of Senegalese individuals with a disability live below the national poverty line (Échevin, 2012). This is also the case of elderly people or children under five old. Therefore, targeting all persons with disabilities or all the elderly means targeting a group in which 50% of people are not poor. FIGURE 23: PERCENTAGE OF EXPENDITURE, BY TARGETING TYPE (2015) Source: Author, based on data collected TABLE 16: BREAKDOWN OF VULNERABLE INDIVIDUALS AND HOUSEHOLDS, BY POVERTY LEVEL Poor Not poor Groups Age Number % Number % Under five years old 0 4 1,077, % 1,045, % School age ,767, % 1,859, % School age but has never attended school 751, % 606, % School age but not enrolled in 2010/ ,590 45% 67,332 55% Over 60 years old , % 420, % Source: ESPS II 110. National programs that use geographic targeting cover the country evenly by poverty level, food insecurity, or malnutrition in the zone in question. For example, the National Family Security Grants Program (PNBSF) uses the results of the poverty map at the community level as well as population density to set quotas of beneficiaries to be enrolled. Similarly, the National Food Security Response Program uses data from the harmonized framework to target geographic areas to carry out interventions after identifying the number of households to be targeted. Finally, the anti-malnutrition program uses the findings of SMART and DHS studies. Determining the extent to which programs that do not use the geographic targeting method allow the country to be covered in an equitable manner is more challenging. 48

58 Most Services Benefit the Richest Populations 111. Most services benefit the richest populations, not necessarily because of poor targeting but because of the design of programs. Some programs are not aimed at combatting poverty but instead strive to tackle claims from certain population categories, including students, farmers, persons over 60, and so forth. Where this is the case, the use of the targeting by category method is not open to discussion. However, these programs lack of relevance and equity within the social protection system raises questions since a significant portion of total social protection expenditure is dedicated to these programs An analysis of the beneficiaries of the costliest programs shows that the majority of services benefit the wealthiest. For example, ESPS II data show that 70% of beneficiaries of higher education grants fall within the two wealthiest quintiles. This should be borne in mind as this program costs approximately 0.5% of GDP. FIGURE 24: BREAKDOWN OF STUDENTS RECEIVING UNIVERSITY SCHOLARSHIPS, BY WEALTH QUINTILE Source: ESPS II, ANSD Similarly, over 60% of those over 60 receiving a pension through the National Retirement Fund (FNR) or the Retirement Savings Fund of Senegal (IPRES) belong to the two wealthiest quintiles. However, only 30% of those beneficiaries are women. Almost 2% of GDP is spent for these programs, half of which is directly financed by the state (FNR). While these data are from 2011, they are still relevant t given that there has been no structural reform in the distribution of higher education grants or the retirement system. FIGURE 25: PENSION RECIPIENTS OVER 60, BY WEALTH QUINTILE Source: ESPS II, ANSD

59 114. In 2011, healthcare FIGURE 26: PERCENTAGE OF ACTIVE POPULATION WITH ACCESS TO A beneficiaries belonged primarily to HEALTHCARE SYSTEM, BY WEALTH QUINTILE the wealthiest quintile. However, the trend could be reversed depending on the success of the CMU s insurance arm. In 2011, medical coverage was limited to the active population in the richest quintile and residing in Dakar, that is to say most government employees and those working in the formal sector. Other elderly people (and those working in agriculture) between 15 and 59 Source: ESPS II, ANSD 2011 years of age had no insurance healthcare coverage until Universal health coverage was set up at that time to allow for the population without medical coverage to receive some protection. After two years of implementation, the CMU s insurance arm is said to cover 792,985 people. However, profile data for these beneficiaries are not yet available. The CMU s insurance arm, for which the priority target is households in the PNBSF, should allow the medical coverage system to be extended to the poorest and alter the profile of the population with access to healthcare coverage, which goes back to However, mutual healthcare associations must function effectively and offer quality services if the system promoted by the CMU s insurance arm is to allow the populations to have medical coverage Similarly, the subsidy to farming inputs and equipment, which totals approximately 0.5% of GDP, favors farmers who are not poor. According to the recently published study by the Agricultural and Rural Forecasting Initiative on farm input subsidies in Senegal (IPAR, 2015), 59.7% of farmers who are not poor received a subsidy as opposed to only 49.4% of poor farmers. The IPAR-led analysis shows that the first quintile of the farmer population consumes about 22% of the subsidized inputs, while the fifth quintile (or the richest farmers) consumes about 26% of subsidized inputs. The study also shows that subsidies on seeds and fertilizer theoretically have the most distributive effects on the first and second quintiles as opposed to equipment, which benefits the richest quintiles to a greater extent. Finally, the study notes that many farmers complain about the fact that input distribution does not take farmers level of need into consideration However, with the implementation of the PNBSF, the government has tried to rectify this imbalance. An analysis of the findings of the PNBSF reference survey conducted in November 2016 on the fourth-generation beneficiaries (Ferre, 2017) reveals that poverty rates observed within the PNBSF beneficiary group are high (93%) and serve as testimony to the program s successful targeting. The proportion of households in extreme poverty is 58%, with substantial variation from one environment to another: 24.8% of households in the PNBSF are extremely impoverished in Dakar, 54.8% in other cities, and 80.4% in rural zones. Thus, the majority of households living on less than USD 1.90 per person per day are found in a rural setting (61% of PNBSF households, while only 44% of PNBSF households live in a rural area). 50

60 II Matching Benefits to Relevant Objectives Amounts Allocated to Recipients 117. The sums spent per beneficiary under the school feeding program are very low and cast doubt on the quality of the service offered to students. First, expenditure per beneficiary under for the Directorate of School feeding programs (DCAS) program fell by almost 50% over the period and by 15% for beneficiaries of the WFP except in 2012 and 2013, when expenditure per beneficiary doubled. Second, these amounts are not sufficient to provide one meal per day per child for the entire school year. In 2015, the total spent per recipient per year was CFAF 2,419 for the DCAS program and CFAF 5,448 for the WTF program. For the program managed by the DCAS, the amount allocated to schools is calculated as part of the total school budget (16%) rather than in terms of needs or the cost of a meal per day per child. Given such low amounts, it is likely that the number of beneficiaries has been overestimated or that school feeding programs are not able to provide each child with one meal per school day. DCAS WTF TABLE 17: TOTAL SCHOOL FEEDING PROGRAMMES EXPENDITURE, PER BENEFICIARY (IN CFAF MILLIONS) Total expenditure 1,061,000, ,000,000 1,166,000,000 1,210,000, ,000, ,000,000 Number of 254, , , , , ,706 beneficiaries Expenditure per 4,171 2,700 4,207 4,182 1,262 2,419 beneficiary per year Total expenditure 3,512,328,500 2,744,805,000 6,670,841,500 7,927,525,500 2,633,216,500 1,654,764,000 Number of 564, , , , , ,738 beneficiaries Expenditure per 6,226 4,852 13,625 15,883 5,276 5,448 beneficiaries per year Source: Author, compilation of data collected 118. The same analysis could be conducted for numerous programs and will reveal the same limitations. For example, the preschooler category includes infrastructure construction cost under total expenditure while the annual amount spent per beneficiary (excluding managerial staff salaries) is around CFAF 10,000, a seemingly small amount to cover the awakening and wellbeing needs of children in these categories. TABLE 18: TOTAL EXPENDITURE PER BENEFICIARY FROM PRESCHOOLER ITEM (IN CFAF MILLIONS) Total expenditure 1,200,000,000 1,050,000, ,000,000 1,150,000,000 1,100,000,000 1,605,000,000 Number of 103, , , , , ,760 beneficiaries Expenditure per beneficiary per year 11,586 9,471 7,494 8,664 7,819 11,087 Source: Author, compilation of data collected 119. From another perspective, the amounts allocated per beneficiary for shock response programs are very high. The National Food Insecurity Response Program (PRNIA) allocates up to CFAF 45,000 per household over three months. This is equivalent to 100% of the cost of a basket of food for a nine-person household. Given that the Senegalese households benefitting from this program are not in an extreme situation where they would have no resources from which to feed themselves, it is not unreasonable to question this figure. Similarly, the National Solidarity Fund (FSN) may allocate up to CFAF 280, per 31 This consists of a CFAF 100,000 transfer, an in-kind transfer equivalent to CFAF 80,000, and an additional CFAF 51

61 beneficiary household affected by a shock regardless of the household s standard of living or the damage incurred. Sustainability of No-Charge or Insurance Healthcare Programs 120. The various no-charge or insurance programs are used inappropriately. First, excessive use of nocharge services is made by individuals who do not belong to the targeted groups. Users take advantage of how difficult it is to prove a child s age because of the lack of a registry of births, deaths, and marriages or to prove the lack of membership of the retirement system for people over 60. Additionally, excessive prescriptions from healthcare professionals significantly inflate total healthcare costs because of the failure to conform to the health pyramid and the failure of healthcare institutions to adhere to the billing system (Dieng, 2016). Moreover, when medications are not available in healthcare centers, this causes some households to register their children under five years of age with a mutual healthcare association so that they can receive partial coverage for medications purchased in private pharmacies. These practices trigger the financing of two subsidies for the same item: the free under-five service, and the subsidy for mutual association membership. BOX 1: CHOOSING POLITICALLY APPROPRIATE PROGRAM PARAMETERS Several types of programs can be implemented as social safety nets, and various parameters can be included in their design. This includes public works, financial transfers with or without conditionalities, support measures, program duration, and criteria for graduating from the program. The decision to include these parameters is first and foremost a technical one aimed at the greatest expected impact. However, political considerations often come into play if program ownership is to be maximized. In fact, politics plays a significant role in program design. Program parameters must take preferences, incentives, and perceptions into consideration. The best designs are those that are technically solid, administratively feasible, and politically sound because they increase ownership while maximizing impact. The technical and administrative feasibility elements of programs are often addressed during their design, but the political aspect is often underestimated or reluctantly addressed (Pritchett, 2005). At the extreme, a perfect technical design that ignores the social safety net support policy may well be the worst option for those it serves. Political obstacles can be overcome in many ways. For example, characteristics and parameters can be chosen in such a way that political preferences are taken into account, and the targeting can be adapted so that it is compatible with given political incentives. However, political adjustments should be introduced as a backstop and minimized and mitigated by paying special attention to program inclusiveness and transparency in order to avoid capture risks. Source: Realizing the full potential of social safety nets in Africa (World Bank, 2017a) 121. The lack of a clear definition of the service package affected by free healthcare and by the CMU s insurance arm poses a significant financial risk to the national budget as well as to mutual healthcare associations. Despite a 2009 ministerial decree excluding certain services from the SESAME plan, the 100,000 transfer should a household member die. 52

62 package of services offered under this program remains fluid and subject to extensive interpretation by individuals and medical personnel. At the end of 2013, the SESAME plan owed CFAF 4.7 billion to healthcare organizations. The service package underwritten by the CMU as currently defined is very broad and quite generous. It covers 80% of the cost of all services (from consultation to surgery) and 50% of medications purchased in private pharmacies, while the annual contribution is low (CFAF 7,000). For persons receiving family security grants, who for the time being constitute most members, the package of services covered by the CMU s insurance arm covers 100% of costs, including healthcare and medications even when purchased at private pharmacies (Ministry of the family and Women s Organizations, This system is subject to abuse The financial viability of the system of mutual healthcare associations is in doubt. Community healthcare association systems can only work if the number of persons registered is much higher than the number of users of the association. This is because the healthcare costs of some are financed only by the membership fees of others. Senegal decided to not make its mutual healthcare association system membership mandatory but rather to encourage membership through partial or total subsidies to costs. The related financial risk to healthcare mutual associations is that they will not have enough members to function. An Abt Associates study cited in a report by Dieng (2016) shows that while the registration rate in the 78 mutual associations targeted by the study increased significantly between 2013 and 2015, payment for healthcare is still low and has not changed over the period. Only 20% of mutual healthcare associations in the sample under study are financially viable. Further, totally free healthcare and medication for persons receiving a family security grant and significant healthcare use by such persons constitute a very heavy financial burden for mutual healthcare associations. Delays in the Distribution of Services in Kind or Money 123. After observing the recurrence of shocks and their impact on their population, a small majority of countries in the region have begun to use their social safety nets as a shock response mechanism. The use of existing social safety nets allows for a rapid response to shocks, thus diminishing their impact on the most vulnerable populations and reducing the total cost of the response. However, Senegal has not yet embarked on this path and is one of approximately 20 countries (for which data are available) that do not use social safety nets as a response mechanism. FIGURE 27: USE OF SOCIAL SAFETY NETS AS SHOCK RESPONSE MECHANISM IN AFRICA 124. Inefficiency of in-kind distribution as a response to shocks. In-kind distribution is more costly than cash transfers (Cunha, 2014; Margolies and Hoddinott, 2014), particularly because of the significant logistical costs entailed. It is true that despite higher costs, certain contexts require in-kind distributions, 53

63 particularly when products are not available on local markets. However, because Senegal s local markets are supplied mainly with imported rice, this commodity (the main staple) is widely available in markets. Further, recent impact assessments comparing in-kind distribution and cash transfers in 12 different countries show that the impact on certain indicators varies according to the operational method. These studies show that the impact of money transfers on food consumption is more significant than the impact of distributing provisions. This also true of the food diversity index (Gentilini, 2014) Rice distribution under food insecurity plans often reaches beneficiaries after the first weeks or months of the beginning of the infertile period. For example, when the government obtained African Risk Capacity (ARC) financing to address food insecurity in the country in 2015, cash flow problems prevented governmental implementation actors from obtaining the funds needed to purchase and distribute the rice in a timely manner. Some rice was distributed late (September 2015), and the remainder was distributed as part of the 2016 response plan (ARC, 2015) While 2015 saw improved consistency in PNBSF cash transfers, the process involved in getting payments to beneficiaries in certain regions remains inadequate. Owing to the limited number of representatives of the payment operator in some departments and the size of the territory to be covered, beneficiaries are forced to travel long distances and to wait up to two days at the counter to be paid. A new payment operator has been recruited in five departments, and mobile teams are now deployed in each community. Although this mechanism does not resolve all the problems encountered by beneficiaries, it is a vast improvement for those living in areas that are particularly to access, such as Medina Yoro Fulla or Dagana. 54

64 PNBSF Cash Transfers Will Not Be Enough to Break the Intergenerational Cycle of Poverty BOX 2: IMPACT OF CASH TRANSFER PROGRAMS IN AFRICA Nine African programs have documented their results for household consumption. Using the metaanalysis approach, the impacts of seven programs are combined into a composite indicator of the effect of social safety nets on household consumption. On average, household consumption increases by USD 0.74 per US dollar transferred. Most programs see an increase in household consumption. However, there is considerable variation between countries. Among them, five programs lead to significant increases. In Malawi, recipient households see an increase in consumption of 179 % of the value of transfers. To an extent, this may be partially explained by the fact that most respondents lived outside of the monetary economy prior to entering the program. The cash transfer program for children in Zambia also shows very large positive effects on total consumption by consumption sub-categories, with 76% of transfers used for food followed by healthcare and hygiene (7%), clothing (6%), and communication and transportation (6%). These programs highlight the transformative potential of social safety nets, which lead to increased consumption exceeding the total cash transfer received. Three programs (the productive social safety net in Ethiopia, the cash transfer program for children in Lesotho, and the Niger social safety net project) have shown a limited consumption increase associated with the cash transfer from the social safety nets program. Overall, the impact as a portion of the consumption of households before the intervention (reference study) is between 0% and 33%. It should be noted that households do not typically spend all of their cash transfer on consumables, but are more likely to allocate it to other areas such as productive investments, staggered expenses such as school fees, or savings. Source: Realizing the full potential of social safety nets in Africa (World Bank, 2017a) 127. The PNBSF amount has an impact on reducing extreme poverty. The program transfers CFAF 25,000 to recipient households on a quarterly basis, or CFAF 100,000 per year for five years. This amount amounts to % of the extreme poverty line for a six-child household in Dakar and in a rural zone, respectively. 32 This is within the average of the amount of money transfers in developing countries, which is between 10 and 20% of household consumption before the transfer (World Bank, 2008). A recent analysis of the distributive effects of Senegal s budgetary policy (World Bank, 2016b) shows that the allocated amounts have an impact on extreme poverty but that they are too small to have a significant impact on poverty. 33 This is not surprising and does not contradict the PNBSF objective of strengthening the development of children s human capital in order to halt the process of intergenerational poverty transmission Yet the mechanism that aims to encourage PNBSF recipient households to invest in their 32 The extreme poverty line for a 6-child household with parents is CFAF 67,637 per month in Dakar and CFAF 57,742 per month in rural areas. The monetary poverty line for a household of the same size is CFAF 154,145 per month in Dakar and CFAF 92,171 per month in rural areas (ANSD, 2011) % of the extremely poor and 100% of the moderately poor who receive family subsidies remain at the same level of poverty after the transfers (World Bank, 2016a). 55

65 children s human capital remains limited. Three years after the first cash transfer, awareness building activities for behavior change in PNBSF households in education, births, deaths, marriages, and health had not begun. This is partly due to the lack of an available budget. For the time being, all behavioral change in households and communication support is financed by development partners (DP). Another explanation is the challenge of inter-ministerial collaboration. Ministries providing social services do not yet perceive the PNBSF as a sufficient opportunity for them to improve their own results and affect those who are most excluded Although the PNBSF can stimulate demand for social services, these services must be available and of good quality to be able to break the cycle of intergenerational poverty. Yet the quality of social services remains poor. For example, the findings of the most recent Program for the Analysis of the Educational Systems of CONFEMEN Countries (PASEC, 2015) reveal that 40% of children cannot read at the end of elementary school. In addition, healthcare entities lack personnel. The analysis of health personnel coverage (ANSD, 2016) revealed in 2013 that Senegal had one doctor per 12,373 persons, one nurse per 4,320 persons, and one midwife per 2,426 women of reproductive age. Further, coverage of the population by healthcare personnel varies enormously between regions, to the detriment of the poorest regions (ANSD, 2016). III Lack of Financial Viability of the Retirement System Doubts Surrounding IPRES Financial Viability 130. While the IPRES system is financed in a context where contributions from current workers are used to pay the pensions of current retirees, the system has a surplus that generates investment income. In 2014, investment income represented approximately 11% 34 of contribution income. A more detailed analysis of the amount of assets and of their utilization rate over the past decade could help explain the financial performance of the IPRES system and anticipate its future financial needs. It is not generally unreasonable to spend 0.9% of GDP on approximately 20% of the elderly population. However, more information is needed on the level of benefits, expenditure by category of retirement, and sources of financing for a more thorough evaluation of the fiscal position of the system and the equity of its benefits The Audit Office pointed the finger at several aspects of IPRES financial management of inflows and investments and of its technical management, which is a possible indication of poor risk management by the institution (Cour des Comptes, 2014). The Audit Office specifically underscored the following: (a) a gap in inflow of contributions; issues with the management of inflows; (b) non-lucrative equity participation; (c) ineffective management of income properties; (d) a lack of planning in terms of property investment; (e) failure to optimize the recovery of contributions; (f) poor control activity; (g) failure to incorporate years of unpaid contributions into the pension equation; (h) prolonged pension settlement waiting times; and (i) pension fraud, all of which were underscored at the technical level. Without major reform of IPRES financial and technical management, these issues are harbingers of a gradual decline in the system financial position. 34 The value of accumulated IPRES assets is unknown. 56

66 132. IPRES has an 80% dependency ratio, which means FIGURE 28: IPRES DEMOGRAPHIC PROFILE that it has 295,000 contributors, or 3.8% of the working age population, and 219,363 pension recipients. This raises critical questions about the sustainability of the retirement system even if this rate is trending downward for pensions at the current replacement rate with no additional sources of revenue. This may also be an indication that the system is Source: Author, based on collected data coming to maturity, when the number of elderly people exceeds (or is close to) the number of contributors. Greater coverage of elderly people could also indicate that the system has undergone a contraction in coverage, that is to say that preceding generations acquired rights to retirement that were higher than the current working age population. The dependency ratio could also be explained by the fact that a former contributor acquired rights to a pension that were transferred to his or her survivors over a very long period. However, more data are needed to be able to draw conclusions on the fiscal sustainability of the IPRES system Finally, Senegal will face significant demographic aging in the decades to come. By mid-century, the number of over 60-year-olds is set to double in proportion to the total population. Currently, the population aged over 60 represents 4% of the total population. However, it is predicted that by the middle of the century, this figure will reach 9% of the total population. The doubling of the elderly population as part of the total population does not necessarily mean that there will be a doubling of the population covered and therefore an increase in retirement spending, but there may be serious repercussions for the retirement system and the wellbeing of the elderly. FIGURE 29: POPULATION BREAKDOWN BY AGE GROUP AND PROJECTION Source: UN demographic projections 57

67 The FNR is Not Sustainable 134. The FNR s 2014 expenditure was 1% of GDP and significantly increased between 2010 and 2014 despite only minimal change in the total number of beneficiaries. The two most important points to be highlighted are: (1) the relatively high cost of the absolute number of persons benefitting from retirement funds; and (2) the relatively high portion of public inflows going to the retirement system when compared to other sectors such as health and education A very generous formula is used to calculate FNR retirement amounts. In fact, when the amount spent per beneficiary is analyzed, the average monthly pension 35 stands at double per capita GDP and about 80% of the insured person's salary. The average monthly pension benefit from the FNR is more than three times the average pension paid by IPRES The state s employer contribution is very costly. In the case of the public sector retirement system (FNR), as the employer is the government, the employer's portion of the retirement contribution rate (23%) is a direct cost to the government. Even if the retirement system were in budgetary balance, with adequate contribution inflows covering retirement expenditure, the relatively high employer contribution rate may be considered problematic from both a budgetary and ethical standpoint. Senegal s retirement contribution rate for the public sector is the highest in the region (Figure 30). FIGURE 30: RATE OF CONTRIBUTION TO PUBLIC SECTOR RETIREMENT SYSTEMS 35 Calculated by dividing retirement expenditure by the total number of beneficiaries. 36 Calculated based on total retirement expenditure divided by the total number of beneficiaries. 58

68 Source: Author, based on retirement systems data, World Bank 137. The National Retirement Fund (FNR) is in deficit and will have to be directly financed from the national budget in the near future if no reform takes place. According to an FNR actuarial study (Ndiaye and Diabate, 2012), various reforms were adopted in 2002 following a large deficit at the beginning of the 2000s, leading to an increase in the retirement age, a reduction of the annuity rate, and taking the last three years of salary into account in calculating pensions. These reforms allowed cash deficits to be reabsorbed, with surpluses appearing within the first year of their implementation. However, the surplus funds did not last long as a deficit already existed in The 2014 public report of the Audit Office illustrated the problem. In 2013, the FNR collected CFAF 64.5 billion in inflows but spent CFAF 68.6 billion on pension payments. While it still had a positive balance left over from the reforms at the beginning of the millennium, its credit balance fell to 14% in 2013, or from CFAF 29.4 billion to CFAF 25.1 billion. 59

69 Chapter 6 Effectiveness of the Social Welfare System I The RNU: A Tool with High Potential Recent Efforts to Improve Targeting 138. Based on the interviews conducted, poverty does not appear to be a decisive criterion in designing assistance and empowerment programs. Such programs largely use the language of vulnerability in their targeting criteria. A great number of them associate the adjective vulnerable with a category, such as vulnerable women, vulnerable young people, vulnerable children, etc. In practice, every program measures vulnerability differently. In some cases, a social survey is carried out to verify vulnerability. However, when vulnerability is the sole criterion, there is no prioritization of support based on degree of vulnerability, and this is in a context where the majority of the population is poor and resources are limited. In such a case, benefits are allocated on a first come, first served basis. Given that very often, the most vulnerable people are those with the least access to information, aid is unlikely to reach those most excluded. BOX 3: HOW BRAZIL, COLOMBIA, PAKISTAN, AND SENEGAL CREATED A SOCIAL REGISTER FOR BETTER EFFICIENCY AND GREATER IMPACT Many countries use social registers to coordinate the identification, recording, and assessment of household eligibility. These registers enable programs to be more efficient through improved targeting, reduce the cost of such targeting, and encourage coordination, which permits households to be supported comprehensively. These registers were initially launched in Brazil and Colombia, where they form the core of the social welfare system. They are used by many programs that target poor households in terms of health, education, water, electricity, social nets, transportation, etc. In all, over 30 programs use Brazil s Cadastro Único and Colombia s SISBEN. Many countries around the world have now adopted social registers. In Africa, Ghana's National Register serves to identify the beneficiaries of LEAP (a cash transfer program) as well as destitute households under the National Health Insurance System. Similarly, in Senegal, the Single National Register (RNU) is used by the National Family Security Grants Program (PNBSF), the Universal Health Coverage Program (CMU), and certain productive programs. TABLE 1: EXAMPLES OF REGISTERS AROUND THE WORLD Country Name of register Number of households (% of population) Number of programs using the register Brazil Cadastro Único 27.2 million (40%) 30+ Chile Social Register of Households 12.3 million (70%) 65+ Colombia SISBEN 10.4 million (73%) 31 Georgia TSA Register 1.2 million (36%) 16 Pakistan National Socio- Economic Register 27 million (90%) 30 (NSER) Philippines Listahanan 15.3 million (77%) 52 Senegal Registre National Unique (RNU) 450,000 (30%) 3 Source: Social Registries for Social Assistance and Beyond. Working paper. World Bank (2017b) 139. However, this has been changing since 2014, the year Senegal set out to construct a Single National Register (RNU). The goal of 60

70 the register is to target the 450,000 poorest households in the country, or about 30% of the total population. As the rate of extreme poverty is around 14.5%, the RNU should be able to identify all of this population despite potential inclusion errors. TABLE 19: RNU POPULATION AND PERCENTAGE IN POVERTY Percentage of the total population Estimated number Individuals in extreme poverty 14.5% 2,012,643 Poor individuals 46.7% 6,482,098 Households in RNU 28% 450,000 Estimated number of individuals in RNU (on the basis of 10 persons per household) 32% 4,500,000 Source: Author, based on ESPS II (ANSD, 2011), and collected data 140. By the end of 2016, 450,000 households had been registered in the RNU. The purpose of this targeting tool is to: 1) enable various social programs to select their beneficiaries quickly by conducting a simple analysis of the database; ii) reduce the costs associated with targeting and selecting the beneficiaries of the various social programs; and iii) enable social programs to be implemented quickly. The PNBSF uses the RNU to target its beneficiaries However, if the RNU is to fulfill its objectives, the identification of the poorest households will have to improve over time in order to minimize exclusion or inclusion errors. A study (Sadoulet and Adrianarimana, 2014) conducted in four municipalities revealed that community targeting, the first step in the RNU targeting process, was not very effective in identifying the poorest households in these communities. In fact, the study showed that the households pre-identified by each community were not significantly poorer than those not pre-identified. Moreover, the RNU should quickly put in place an efficient process for updating its data along with an effective mechanism for processing claims if it is to be useful in implementing social programs. Efficiency Improvements in the Social Welfare System Will Invariably Come with Increasing Use of the RNU 142. Having gone through rapid development, the RNU remains in an institutionally and financially precarious position. The register is highly dependent on external financing to conduct its campaigns aimed at identifying the poorest households. It does not have its own budget line, and no decree has as yet endorsed its existence. The cost of targeting a household for RNU purposes varies between CFAF 3,500 and 4,000 (not counting salaries or the department's running costs). Given the per-household cost of targeting, the RNU could be an effective mechanism for targeting social welfare programs provided it is used by a majority of them The underutilization of the RNU by government players holds back efficiency improvements in the social welfare implementation mechanism. Only a few social welfare programs managed by the government have started using the RNU, including the PNBSF, the insurance arm of the CMU, and PRODES. There are three main reasons for this. First, the RNU is a recent tool that remains under construction and is relatively unknown by sectoral players. Second, the RNU data and household identification process still 61

71 need to be improved. Last, institutional infighting and the dynamics of power limit cooperation between players and harmonization of the methods and tools used by the system. Systematic use of the RNU should permit increased synergies between programs and offer an integrated service package to the poorest people Use of the RNU to target the beneficiaries of the higher-education grants program or subsidies for agricultural inputs or machinery would limit the unfairness of the system. With the RNU as the basis, benefits under such programs could be allocated transparently to the poorest and thereby boost the impact of those programs instead of targeting the wealthiest, as is currently the case In addition, using the RNU to target food insecurity response programs would achieve substantial savings that could be redistributed to beneficiaries. In 2014, the government spent CFAF 600 million on targeting the food insecurity response program and CFAF 100 million in At the same time, NGOs also targeted their beneficiaries and spent several million CFA francs in doing so. A number of studies are underway to determine the extent to which the programs targeting food insecurity could use the RNU as a targeting tool. TABLE 20: FILTERS THAT COULD BE USED IN THE RNU Program Potential filters Food Insecurity Response HEA scoring Program (PRNIA) Subsidies for Agricultural Agricultural households living in poverty Inputs and Machinery PRODAC Young people from households living in poverty in a rural area PAPA People over 60 belonging to a household living in poverty Program to Combat Poor households with at-risk children Vulnerable Children Support for Wards of the Orphaned children living in a poor State household Medical Care Program for Persons living in a poor household the destitute SESAME Plan People over 60 belonging to a poor household School Cafeterias Schoolchildren living in a poor household Agricultural Disaster Fund Poor households living from agriculture who suffered a shock National Solidarity Fund (FSN) Poor households who suffered a shock Equal Opportunity Cards Person with a disability living in a poor (CEC) household Childhood Deprivation Poor households with children under 12 Program Source: Author, based on collected data 146. The RNU constitutes an opportunity for low-budget social welfare programs to reach the poorest and most excluded populations. As we saw earlier, many programs have limited budgets and opt for the least expensive targeting method (category-based and "first come, first served") to allocate the maximum possible resources to the benefits on offer. The RNU can enable these programs to ease this targeting shortfall and allocate their resources to the neediest by making available a database of the poorest individuals and households. II A Complex Institutional Mechanism 62

72 Difficult Sector Coordination 147. The government of Senegal has been engaged in the construction of a national social safety nets system since It is one of some 20 African countries that have embarked on this path or that already have a national safety nets system. A similar number of countries, for which data are FIGURE 31: CONSTRUCTION OF SOCIAL NET SYSTEMS IN AFRICA unavailable, have no strategy for developing a national social safety nets system (Figure 31). Such a system includes: (i) a common mechanism that enables each program to target and record its beneficiaries; (ii) an institutional platform that serves as a coordinating device at central, regional, and local level; and (iii) a set of interventions that enable appropriate coverage of needs. As we saw in earlier sections, Senegal has made great advances in putting in place a common targeting mechanism and implementing programs with wide coverage. However, progress in sectoral coordination is less evident Coordination efforts are focused on the sector as a whole and not specifically on safety nets. It becomes all the more complex, as can be seen in Senegal's social welfare sector, when multiple players are involved in its implementation. We saw earlier that there are so many programs that they are difficult to capture exhaustively. At the same time, the institutional structure of the implementation of these programs is complex. The programs in progress in 2015 are implemented by at least 9 oversight ministries and 12 agencies. Additionally, the institutions in charge of social welfare programs regularly suffer a degree of institutional instability. For example, the FSC and FSN came under the Ministry of Family Services until 2012 and the SESAME Plan under the General Directorate of Social Action (DGAS) until This setup does not always meet the demands of efficiency: for example, at least three institutions (DGAS, CMU, FSN) are tasked with providing healthcare for the destitute. Similarly, at least three institutions are tasked with helping highly vulnerable groups to develop income-generating activities (IGA): DGAS, the Poverty Monitoring Committee, and the Women's Entrepreneurship Fund A sector coordination mechanism was set up in 2013 but seems ill-equipped to manage this multi-sectorality. An interministerial steering committee for the National Social Protection Strategy (SNPS) and a technical support committee were created by decree. These committees are tasked with ensuring better coordination of social welfare interventions. They are chaired by the DGPSN and the Ministry of the Economy, Finance, and Planning (MEFP). Much effort has gone into energizing these committees and improving sectoral coordination. However, the results of the coordination remain limited. The multiplicity of players makes coordination particularly difficult. The DGPSN, which reports to the President's Office, and the MEFP, which reports to the Prime Minister's Office, do not have direct 63